ENGAGING PATIENTS IN THEIR OWN PAIN MANAGEMENT: AN ACTION RESEARCH STUDY A thesis submitted for the degree of Doctor of Philosophy By Dee Burrows Faculty of Health Studies, Buckinghamshire Chilterns University College Brunel University May 2000 ABSTRACT Traditionally analgesics have been the main stay of postoperative pain management. Taught non-pharmacological strategies have also been used. The possibility that patients may have a repertoire of self-generated strategies has tended to be overlooked. The aim of this study was to identify whether patients possess and utilise self-generated strategies for acute pain, and if so, to ascertain the effect of engaging patients in their postoperative pain management by supporting their strategies. To empower nurses to deliver effective pain management, and to promote evidence based practice, a second aim was to actively engage nursing staff in the process. A collaborative action research design based on critical theory was used to facilitate practitioner participation, enable practice development and generate a theoretical understanding of the issues. Three action research cycles were identified, each containing various methods and involving both patients and staff. The first cycle, entitled “telling the situation as it is”, described current practice from the ward staff’s perspective and identified 10 strategy themes through an outpatient survey. Cycle two focused upon “testing the proposed changes”. In contrast to many action research studies, a randomised controlled trial was undertaken to test the effect of incorporating patients’ self-generated strategies into postoperative pain management. The nursing staff were pivotal in both the decision to undertake the trial and in its management. The results indicated that identifying and supporting patients’ self-generated strategies significantly reduced postoperative anxiety, pain intensity and distress. Fewer experimental subjects required opiates, although non-steroidal anti-inflammatory consumption was increased in this group. Staff understanding and awareness of pain and their ability and confidence to manage pain increased. Certain staff also gained research skills. The third and final cycle, “creating change in practice: implementation and implications”, is being led by the nurses. The potential of action research to engage nurses in practice developments related to patient-centred care has been demonstrated. TABLE OF CONTENTS TITLE 1 ABSTRACT 2 TABLE OF CONTENTS 3 LIST OF ACRONYMS 15 ACKNOWLEDGEMENTS 16 INTRODUCTION 17 CHAPTER ONE: PAIN: A MULTIDIMENSIONAL PHENOMENON 1.1 Introduction 19 1.2 The Phenomenon of Pain 19 1.3 Pain Perception and The Gate Control Theory 21 1.4 Pain As a Multidimensional Phenomenon 23 1.5 Strategies for Chronic Pain 28 1.6 Teaching Strategies for Acute Pain Management 32 1.6.1 The Evidence 33 1.6.2 From Research to Practice 36 1.7 Self-Generated Strategies in Acute Pain Management 38 1.8 Pain Coping: Engaging Patients and Nurses in Postoperative Pain Management 43 1.9 Conclusion 45 CHAPTER TWO: PATIENT-CENTRED PAIN MANAGEMENT 2.1 Introduction 46 2.2 Clarity or Confusion? 46 2.3 Patient-Centred Care 47 2.4 Empowerment 51 2.5 Patient Empowerment 55 2.6 Nurse Empowerment 57 2.7 Patient-Centred Care and Empowerment in Acute Pain Management 61 2.8 Conclusion 65 CHAPTER THREE: METHODOLOGY: CRITICAL THEORY AND ACTION RESEARCH 3.1 Introduction 66 3.2 Positivism and Naturalism: The Traditional Research Paradigms 66 3.3 Action Research: A Design for Practice Based Disciplines 71 3.3.1 In Defence of Action Research 73 3.4 Critical Theory: What Is It and What Does It Contribute To Action Research? 74 3.4.1 Emancipation 75 3.4.2 Reflexivity 75 3.4.3 Praxis 77 3.4.4 Ideology 78 3.4.5 Knowledge Development 80 3.4.6 Critical Theory and Action Research 81 3.5 Action Research: Characteristics and Definitions 82 3.5.1 Identifying Situation Specific, Practical Problems 83 3.5.2 Creating Change In Practice 86 3.5.3 Collaboration 88 3.5.4 Developing Theory 91 3.5.5 Action Research Defined 93 3.6 Action Research Approaches 95 3.6.1 The Technical Collaborative Approach 95 3.6.2 The Mutual Collaborative Approach 96 3.6.3 The Enhancement Approach 97 3.6.4 Alternative Typologies 99 3.7 Action Research: The Issue of Rigour 102 3.7.1 Truth-Value: Internal Validity Versus Credibility 102 3.7.2 Applicability: External Validity Versus Fittingness 103 3.7.3 Consistency: Reliability Versus Auditability 104 3.7.4 Neutrality: Bias Versus Rigour and Reflection 105 3.8 Action Research: The Issue of Ethics 106 3.8.1 Participant Vulnerability 106 3.8.2 Voluntary Consent 107 3.8.3 Informed Consent 108 3.8.4 Anonymity and Confidentiality 109 3.9 Action Research: Design Issues 109 3.10 Conclusion 110 CHAPTER FOUR: METHODS 4.1 Introduction 111 4.2 The Research Objectives and Questions 111 4.3 Developing The Action Research Cycles 112 4.4 Cycle One: Telling The Situation As It Is 114 4.4.1 Negotiating Access 115 4.4.2 The Setting 116 4.4.3 The Collaborative Framework 119 4.4.3.1 Co-action Researchers 120 4.4.3.2 Collaborative Group 121 4.4.3.3 Collaborative Informants 124 4.4.4 From Planning To Acting and Observing 125 4.4.5 The Focus Groups 126 4.4.5.1 The Process 127 4.4.5.2 Analysis 129 4.4.5.3 The Process Adopted 130 4.4.5.4 Scientific Rigour 131 4.4.6 The Survey 132 4.4.6.1 The Method 133 4.4.6.2 The Sample 135 4.4.6.3 Ethical Issues 135 4.4.6.4 Scientific Rigour 137 4.4.6.5 Analysis 139 4.4.7 Reflecting Upon The First Cycle 141 4.4.7.1 Staff Workshops 141 4.4.8 Summary of Cycle One 145 4.5 Cycle Two: Testing The Proposed Changes 145 4.5.1 From Planning to Acting, Observing and Reflecting 145 4.5.2 The Randomised Controlled Trial 146 4.5.2.1 The Research Hypothesis 149 4.5.2.2 Determining The Sample Size 150 4.5.2.3 Inclusion and Exclusion Criteria 150 4.5.2.4 Data Collection Methods 151 4.5.2.5 The Trial Protocol 156 4.5.2.6 The Procedure for Recruiting Patients 157 4.5.2.7 The Analysis Strategy 158 4.5.2.8 Summary 161 4.5.3 The Staff Questionnaires and Researchers’ Critical Discussion 161 4.5.3.1 Data Collected During the RCT - The First Staff Questionnaire 161 4.5.3.2 Data Collected on Completion of the RCT - The Second Questionnaire 163 4.5.3.3 The Researchers’ Critical Discussion 164 4.5.4 The Field Notes 164 4.5.4.1 Structured Observation 165 4.5.4.2 Unstructured Observation 167 4.5.4.3 Field Note Analysis 170 4.5.5 Summary of Cycle Two: From Reflecting to Planning 171 4.6 Cycle Three: Creating Change In Practice - Implementation and Implications 172 4.7 Conclusion 175 CHAPTER FIVE: SELF-GENERATED PAIN COPING STRATEGIES: ANALYSIS OF THE PATIENT DATA 5.1 Introduction 176 5.2 The Survey Findings 176 5.2.1 The Sample 177 5.2.2 Questionnaire Sample Characteristics 178 5.2.3 Pain Relieving Strategies 180 5.2.3.1 Patients’ Self-Generated Strategies 180 5.2.3.2 Family and Friends Generated Strategies 181 5.2.3.3 Patients’ Perceptions of Practitioner Generated Strategies 182 5.2.3.4 Summary 183 5.2.4 The Efficacy and Use of Pain Relieving Strategies 185 5.2.4.1 Perceived Efficacy of Self-Generated Strategies 185 5.2.4.2 Differences in Strategy Use According to Pain Intensity 187 5.2.4.3 Strategy Use in Hospital 189 5.2.4.4 Being Empowered 190 5.2.5 Summary 192 5.3 Findings From The Randomised Controlled Trial 193 5.3.1 The Sample 194 5.3.2 Sample Characteristics 196 5.3.3 Outcome Measures 201 5.3.3.1 Pain Intensity 201 5.3.3.2 Pain Distress 203 5.3.3.3 Analgesic Consumption 204 5.3.3.4 Antiemetic Consumption 208 5.3.3.5 Anxiety 209 5.3.3.6 Length of Stay 210 5.3.3.7 Patient Satisfaction 211 5.3.3.8 Patients’ Perceptions of Control Over Pain Management 212 5.3.3.9 Summary 213 5.3.4 Strategy Use 214 5.3.5 Researchers’ Reflections 219 5.3.5.1 Pain Perception 220 5.3.5.2 Pain Relieving Strategies 221 5.3.5.3 The Trial Design 223 5.4 Summary and Initial Interpretations 224 5.5 Conclusion 230 CHAPTER SIX: DEVELOPING NEW PRAXIS: ANALYSIS OF THE NURSES DATA 6.1 Introduction 231 6.2 Nature of the Practice Problem: Analysis of the Focus Group Data and Related Field Notes 232 6.2.1 Pain Management Goal 233 6.2.2 Pain Assessment 236 6.2.3 Communication and Teamwork 238 6.2.4 Preoperative Information Giving 241 6.2.5 Patients’ Attitudes, Perceptions and Responses to Pain 242 6.2.6 Pharmacological Pain Management 245 6.2.7 Non-Pharmacological Pain Management 250 6.2.8 Self-Generated Strategies 256 6.2.9 Multidisciplinary Working and Organisational Issues 260 6.2.9.1 Physiotherapists 261 6.2.9.2 Pharmacists 262 6.2.9.3 Doctors 263 6.2.9.4 Nurses 267 6.2.10 Recent Improvements 269 6.2.11 Summary 270 6.3 Creating Change In Practice: Pain and Its Management 270 6.3.1 Pain Assessment 271 6.3.2 Pain Management 272 6.3.2.1 Patient Generated Strategies 274 6.3.2.2 Pharmacological Management 279 6.3.3 Multidisciplinary Working 280 6.3.4 Summary 282 6.4 Collaboration: The Action Research Process 282 6.4.1 The Co-action Researchers’ Experiences 283 6.4.2 The Collaborative Groups’ Experiences 289 6.4.3 The Collaborative Informants’ Experiences 297 6.4.4 Becoming Empowered 302 6.5 Conclusion 303 CHAPTER SEVEN: A CRITICAL THEORY FOR ACUTE PAIN MANAGEMENT: DISCUSSION AND CONCLUSIONS 7.1 Introduction 305 7.2 Self-Generated Strategies: An Approach to Coping with Pain 305 7.3 The Role of Self-Efficacy 310 7.4 Individuality and Partnerships, Emancipation and Empowerment 316 7.5 Acute Pain Management: The Beginnings of a Critical Theory 319 7.6 The Pain Strategies Questionnaire 321 7.7 Emancipatory Praxis: Furthering the Critical Theory Perspective 323 7.8 Evidence Based Praxis: Methodological Issues 328 7.9 Recommendations and Summary 331 CONCLUSION 333 REFERENCES 335 LIST OF APPENDICES: Appendix 1 Burrows, D. and Kendall, S. (1997) Focus groups: what are they and how can they be used in nursing and health care research. Appendix 2 Burrows, D. (1998) Using focus groups in nursing research: a personal reflection. Appendix 3 Information Sheet and Consent to Participate in Focus Groups Appendix 4 Patient Information Sheets and Consent for Out-Patient Survey Appendix 5 Survey Questionnaire and Interview Schedule Appendix 6 Information Sheet and Invitation to Cycle One Staff Workshops Appendix 7 The Power Calculations Appendix 8 The Data Collection Instruments Used During the Randomised Controlled Trial 8.1 The Preoperative Patient Questionnaire 8.2 The Strategies Questionnaires (Pre- and Postoperative) 8.3 The Self-Report Pain Intensity and Distress Scales 8.4 The Analgesic and Antiemetic Consumption Record 8.5 The Postoperative Patient Questionnaire Appendix 9 The Strategies Booklet - A4 Version Appendix 10 Patient Information Sheet and Consent for Participation in the Randomised Controlled Trial Appendix 11 Covering Letter and Staff Questionnaire, Administered During the RCT Appendix 12 Covering Letter and Staff Questionnaire, Administered Following Completion of the RCT Appendix 13 The Pain Strategies Questionnaire Appendix 14 The Demographic and Clinical Characteristics of the Survey Interviewee Sub-Sample Appendix 15 Strategies Listed by Interviewees on Their Questionnaires Appendix 16 Sample Interview Transcript from the Outpatient Survey Appendix 17 Letter and Handouts for Staff Unable to Attend the Workshops CHART Chart 1 The Research Process and Roles of Those Involved in the Study 21 DIAGRAMS Diagram 1 The Three Action Research Cycles 117 TABLES Table 1 The First Cycle: Telling the Situation As It Is 118 Table 2 The Second Cycle: Testing the Proposed Changes 148 Table 3 Summary of the Data Collection Instruments Used During the RCT 154 Table 4 Variables Identified for Baseline Analysis 162 Table 5 Variables Identified for Hypothesis Testing 163 Table 6 The Range of Field Notes Collected During the Study 168 Table 7 The Framework Used for Recording Observational Field Notes (Spradley, 1980) 169 Table 8 The Codes Identified for Analysing the Field Notes 173 Table 9 The Third Cycle: Creating Change In Practice - Implementation and Implications 176 Table 10 Summary of Reduction From Population To Sample 180 Table 11 The Demographic Profile of the Survey Sample 181 Table 12 The Clinical Profile of the Survey Sample 182 Table 13 Classification of Self-Generated Pain Relieving Strategies 183 Table 14 Frequency of Patients Self-Generated Strategies by Theme and in Ranked Order 184 Table 15 Frequency of Family and Friends Generated Strategies by Theme and in Ranked Order 185 Table 16 Frequency of Practitioner Generated Strategies by Theme and in Ranked Order 186 Table 17 Overall Frequency of Strategy Use by Theme and in Ranked Order 187 Table 18 Interviewees Comments Regarding the Efficacy of Strategies for Relieving Pain 189 Table 19 Number of Times Strategy Theme Recorded for Everyday Versus Acute Pain 191 Table 20 Strategy Themes Identified, by Interviewees , For Use in Hospital 192 Table 21 Summary of Reduction From Admission to Consent 198 Table 22 Reasons for Withdrawal of Subjects from RCT 198 Table 23 The Demographic Profile of the RCT Sample 199 Table 24 The Clinical Profile of the RCT Sample 201 Table 25 The Pain Profile of the RCT Sample 202 Table 26 Overall Pain Intensity 204 Table 27 Daily Pain Intensity 205 Table 28 Overall Pain Distress 206 Table 29 Daily Pain Distress 207 Table 30 Analgesic Administration by Category and Group 208 Table 31 Amount of Morphine via PCAs/Epidural 208 Table 32 Number of IM/Oral Opiates 209 Table 33 Number of Subjects Receiving Opiates 209 Table 34 Non-Opiate Consumption 210 Table 35 NSAID Consumption 210 Table 36 Postoperative Antiemetic Consumption 211 Table 37 STAI Scores 212 Table 38 Comparison of Entry and Exit STAI Scores 213 Table 39 Length of Stay 214 Table 40 Patient Satisfaction 215 Table 41 Patient in Control of Pain Management 215 Table 42 Nurse/Doctor in Control of Pain Management 216 Table 43 Summary of the RCT Findings 217 Table 44 Experimental Subjects’ Strategy Use On Admission to Hospital and Preferences for Postoperative Use 218 Table 45 Number of Strategies Used by Individuals Postoperatively 219 Table 46 Subjects Postoperative Strategy Use, by Group and Theme 219 Table 47 Mean Score of Perceived Strategy Efficacy by Group and Strategy 221 Table 48 The Themes Generated from the Focus Group Data and Related Field Note 236 List of Acronyms AR Action Research APS Acute Pain Service Bpac Burrows pain assessment chart CCSI Cognitive Control Strategies Inventory CNS Clinical Nurse Specialist CSQ Coping Strategies Questionnaire DNA Did Not Attend HCAs Health Care Assistants IASP International Association for the Study of Pain IM Intramuscular MPQ McGill Pain Questionnaire MST Morphine Sulphate Tablet NHS National Health Service NHSE National Health Service Executive NHSME National Health Service Management Executive NSAIDs Non-steroidal anti-inflammatory drugs PCAs Patient Controlled Analgesics PONV Postoperative nausea and vomiting PSQ Pain Strategies Questionnaire PSEQ Pain Self-Efficacy Questionnaire RCT Randomised controlled trial SD Standard deviation SPSS Statistical Package for Social Science STAI State Trait Anxiety Inventory TENS Transcutaneous electrical nerve stimulation UKCC United Kingdom Central Council WDR Wide dynamic range acknowledgements I would like to thank the following people without whom this study would not have been possible. Firstly, Professor Sally Kendall and Dr Kate Seers, for their consistent support and understanding. Their insight, challenges and academic encouragement, provided many opportunities to enable me to critically appraise and reflect upon my research. I thank you for enabling me to grow. To the nurses involved at Stoke Mandeville NHS Trust, I offer my heartfelt thanks. To Chris, who encouraged the approach. To Ann, who accepted the challenge, provided access and produced a motivated team, I thank you for your calmness and on-going support. To all the ward staff who were unfailing in their welcome and who made me feel so at home, your motivation has been out-standing despite some difficult times. To Marie-Louise and the staff in outpatients, thank you for your interest and help and the never-ending refreshments. To the action group members, for their hard work and support: Kim, Vanessa and Jackie who were active members early on, Pauline for maintaining interest and involvement despite moving wards (and to Carol for allowing her to do so), to Margot who took on the responsibility for ensuring that the night staff were involved, and to Margot, Louise and Tessa, without whom the trial would simply not have been completed. My special thanks must be extended to Sue and Elaine. To Sue for her commitment to learning and practice development, for all the out of hours work, for her honesty, ideas and insight, my thanks. And to Elaine, Senior Sister for the Acute Pain Service and my co-researcher, who maintained my motivation, helped me to focus and ensured that my feet were kept on the ground. Penultimately, my thanks to Becky whose last minute running around was invaluable. Lastly, to my husband Mick, who kept me going, showed unfailing interest even when charged with not doing so, loved, helped and supported me through-out. INTRODUCTION The contemporary political and health care climate is placing increasing emphasis upon patient choice and participation in decision making (DH, 1989; 1998; Entwistle et al, 1998). In acute pain management, current opportunities for patients to participate actively are primarily focused upon the use of patient controlled intravenous analgesics and patient initiated requests for oral analgesics. Despite advances in analgesic management over the past decade, many patients continue to experience moderate to severe pain following surgery (Bruster et al, 1994). Previous research by the author suggested that patients in surgical settings may use self-generated strategies to help them cope with postoperative pain. However, nurses and other health care practitioners are largely unaware of this activity and are thus unable to support the patient in their actions (Burrows, 1990, 1997b). Furthermore, there is no evidence regarding the effectiveness of encouraging patients to use their own strategies following surgery. The White Papers Working for Patients: The Health Service Caring for the 1990s and A First Class Service: Quality in the new NHS (DH, 1989; 1998) exhort health care professionals to take account of patients’ views and wishes and to develop stronger partnerships. According to Wilkinson (1999), an appreciation of patients’ beliefs and values about health care is crucial to enable patients to become more involved in health care decisions. However, Cahill (1998) argues that patient participation demands a radical change in the nurses role, a view which is reflected within the recent White Paper Making a Difference (DH, 1999). Fulton (1997) suggests that nurses need to be empowered before they can begin to involve patients in their own care. If nurses are to actively engage acute pain patients in their own care, they need to move from directional forms of care management, to consultative and collaborative approaches which place the patient at the centre. This requires a consciousness shift that goes beyond providing individualised pain management, to a patient-centred approach which proactively involves and empowers patients to make decisions-about their health care (Cahill, 1998; Entwistle et al, 1998). However, as Cahill (1998) points out, there have been few empirical studies which address whether such a change in approach is beneficial to either the patient or health care professional. This study sets out to empower nurses to engage patients in their own pain management through the identification and support of patients self-generated, acute pain relieving strategies. Both process and outcome are investigated in addressing the implications of the study for practice. Chapter one examines definitions of pain and presents an overview of the Gate Control theory (Melzack and Wall, 1965, 1996). The idea of pain as a multidimensional phenomenon is explored and used as a theoretical framework for the subsequent discussion on pain relieving strategies. Both taught and self-generated pain coping strategies are commonly described in the chronic pain literature as being of benefit in pain management. However, the focus in acute pain management has tended to be upon taught strategies. It is postulated that self-generated strategies may interfere with the efficacy of taught strategies. It is proposed that patients’ own strategies should be identified upon admission for surgery and that nurses should adopt a patient-centred approach to the management of postoperative pain by engaging patients in their own pain management. Chapter two explores the concept of patient-centred care and proposes that if nurses are to use this approach successfully, they need to consider transferring power for decision making to the patient. The concept of empowerment is discussed and it is argued that only the emancipatory approach (Friere, 1972; Habermas, 1979, 1988) has the potential to liberate patients and practitioners from the traditional biomedical beliefs surrounding pain management. It is suggested that if patients are to be engaged in their own pain management, nurses need to reflect critically upon current practices and free themselves to approach pain management creatively and from the patient’s perspective. In chapter three, it is argued that the two main philosophical approaches to research and knowledge development, positivism and naturalism, are unequal to the challenge of generating functional knowledge for patient-centred care. Rather, “new paradigm research” (Reason & Rowan, 1981), with its emphasis upon collaborative inquiry, critical reflection and social action, is advocated. Linked to critical theory, new paradigm designs, such as action research (Carr and Kemmis, 1986), lend themselves to considering issues of empowerment and practice based problems. The need to empower nurses to engage patients in their own postoperative pain management, through the identification and employment of self-generated pain coping strategies, is viewed as a practice based issue within patient-centred care. The action research design, its characteristics and approaches are discussed, together with a consideration of some of the issues confronted by action researchers. Chapter four details the methods used within the three action research cycles and offers rationale and a critical discussion for their choice. In the first cycle, entitled “telling the situation as it is”, field notes and focus groups were used to obtain insight into the ward staff’s perspectives on current practice. A survey of 200 out-patients, and interviews with 20 patients, provided data regarding the existence of self-generated acute pain relieving strategies. Cycle two focused upon “testing the proposed changes”. In contrast to many action research studies, a decision was taken to undertake a randomised controlled trial to test the effect of incorporating patients self-generated strategies into postoperative pain management. The nursing staff were pivotal in making this decision and in managing the trial. Cycle three, “creating change in practice: implementation and implications”, is discussed briefly as it is being managed by the ward nurses and Acute Pain Service. The chart, on page 21, highlights the roles adopted by those involved in the study during the three action research cycles. Chapters five and six present the results, with chapter five focusing upon data and findings from the patients’ perspective and chapter six from the nurses’ perspective. Each appears very different, with the focus upon quantitative, outcome orientated data from patients, and qualitative, process orientated data from staff. Thus in chapter five, the results of the survey, patient interviews and randomised controlled trial are presented, together with some qualitative data taken from the researchers’ reflective field notes. Overall, the survey indicated that self-generated strategies are used to relieve acute pain. Some 43 strategies were identified and collapsed into 10 themes which were generated from the data, rather than the literature. The results of the randomised controlled trial indicated that identifying and supporting patients’ self-generated strategies significantly reduces postoperative anxiety, pain intensity and distress. Fewer experimental subjects required opiates, although non-steroidal anti-inflammatory consumption was increased in this group. The field note data suggested that patients require permission and facilitation if they are to be empowered to engage in their pain management. The staff results, presented in chapter six, highlight the nature of the practice problem, the manner by which changes were made and the way in which the nurses collaborated in both the research and action. Ideologies were exposed and, through the raising of collective consciousness, beliefs and practices were challenged and transformed. The focus moved from practitioner-centred analgesic management, to a patient-centred approach in which pharmacological techniques were combined with patients’ self-generated pain relieving strategies. Staff awareness and understanding of pain, and their ability and confidence to manage pain, increased. Certain staff also gained research skills. The fact that the third and final cycle of the study is being managed by ward staff reflects the potential of action research to empower those involved. Chapter seven presents the discussion. Coping and self-efficacy (Bandura, 1987) are used to explain some of the differences found between the control and experimental groups in relation to the use of self-generated pain coping strategies. The contribution of the study to patient-centred care and to a critical theory for acute pain management are explored from the perspective of both patients and nurses. The discussion concludes by suggesting that engaging patients in their own pain management has the power to improve postoperative outcomes. Chart 1: The Research Process and Roles of Those Involved in the Study Research Process Authors’ RoleRole of Acute Pain SisterRole of Collaborative Group*Role of Collaborative Informants*CYCLE ONE Problem identification & literature review ( Problem refinement ( (  Initial research design and identification of cycle one methods (   Refinement of methods ( ( ( Submission to Ethics Committee ( Advise Focus groups Co-facilitator Co-facilitator Participants Participants Patient survey ( ( ( 1 nurse Field notes and reflective diaries ( ( ( 1 nurse Data analysis ( Involved in coding 1 nurse involved in coding Reflection on & refinement of results ( ( ( ( at staff workshopsCYCLE TWO Planning & identification of methods ( ( ( ( at staff workshops Submission to Ethics Committee ( Advise Advise RCT ( ( ( First staff questionnaire ( ( 1 nurse Participants Second staff questionnaire ( Participants Participants Researchers’ critical discussion Participant Facilitator Participants Field notes and reflective diaries ( ( ( Data analysis ( 2nd questionnaire Reflection on & refinement of results ( ( ( Theory development ( CYCLE THREE Discussion on practice transformation and identification of action plan ( ( ( Implementation of plan. Field notes and audit. Analysis and reflection Support & withdrawal ( ( ( * The collaborative group and collaborative informants are described in sections 4.4.3.2 and 4.4.3.3. In essence, the “group” were representatives from the nursing team and the “informants” the remainder of the nurses. CHAPTER ONE Pain: A Multidimensional phenomenon 1.1 Introduction This chapter considers the concept of pain as a complex, subjective, multi-dimensional phenomenon. Starting with a brief historical overview and critical examination of selected definitions of pain, the chapter moves on to examine the Gate Control theory (Melzack and Wall, 1965; 1996). The multi-dimensional perspective of pain described by McGuire (1992) and the need to reflect the complexity of pain through a combination of pharmacological and non-pharmacological strategies are considered. Literature from chronic and then acute pain management on the efficacy of both self-generated and taught non-pharmacological strategies is discussed. Finally, the concept of engaging surgical patients in their own pain management is introduced. 1.2 The Phenomenon of Pain The manner in which pain is interpreted and defined by society is contextually bound (Merskey, 1980). The Ancient Greeks, Aristotle (Smith and Rose, 1931, Translation) and Plato (Jowett, 1871, Translation), viewed pain as a unidimensional experience linked to emotion and passion. As scientific knowledge advanced, increasing importance was given to pain as a physical sensation arising from tissue damage (Descartes, 1644). However, it was not until the groundbreaking work of Melzack and Wall (1965) that the mind-body perspective, combining physical and psychological components, began to become inextricably linked in Western medicine. The increasing awareness of a mind-body link is reflected in definitions of pain. Writing in 1968, Sternbach described pain as a complex, private experience of hurt, which may arise as a consequence of tissue damage. Although the reference to the experience of hurt acknowledges that pain may be more than a biological phenomenon, Sternbach’s (1968) description emphasises pain as a physiological concept. In contrast, the International Association for the Study of Pain (IASP) recognised the mind-body link more overtly, when they defined pain as: “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage” (IASP Subcommittee on Taxonomy, 1979, p.250)” According to the IASP (1979), descriptions of pain are always subjective. Anand and Craig (1996) point out that this assumes that pain can and will be self-reported. They argue that the absence of self-report does not mean the absence of pain and suggest that the subjective emphasis has led society to assume that those who cannot verbally and fluently express pain, experience little or no pain. Anand and Craig (1996) propose that the communication abilities of an individual should be considered in determining the balance of importance between verbal self-reports and behavioural responses. In a counter argument, Merskey (1996) comments that while pain behaviour may be a useful indication of pain, it is not by nature, subjective and can not therefore be used to define the existence of pain. Derbyshire (1996) further argues that both views miss the complexity of pain, which incorporates a number of different elements. The following statement by Montes-Sandoval (1999) clearly reflects the current view of pain as a complex: “neurophysiological, psychological, socio-cultural response to noxious stimuli” (p.938). These stimuli may be physical and/or psychological, but are definitely subjective and unique. It should be noted however, that the above quotation is one of eight listed critical attributes in what is a rather limited and weak concept analysis of pain. Of the 23 references included, just eight are specific to pain, with the remainder being psychology textbooks or encyclopaedias. Although an analysis framework was used, no explanation is given of the way in which critical attributes were generated. As a result, the conclusions appear arbitrary. Superficially, Montes-Sandoval’s (1999) work may be further criticised for its inability to offer a succinct definition of pain as a consequence of the analysis. However, as the author herself states: “it would be misleading to represent pain as having an exclusive definition” (p.940). Indeed, as a subjective phenomenon, it is difficult to define pain from any perspective other than that of the individual who is experiencing it (McCaffery, 1999). Sofaer (1998) adds that attempts to define pain prove problematical as a result of the complexity of factors involved in perceiving and responding to pain stimuli. The intricacies of pain perception are outlined in the section below. 1.3 Pain Perception and The Gate Control Theory The mechanisms of pain perception are complex and continually being updated as scientific knowledge in the area increases. Detailed descriptions may be found in Price et al (1994), Ekblom and Rydh-Rinder (1998) and Cousins and Power (1999). A basic overview is presented in the paragraph below to set the scene for the subsequent explanation of the Gate Control theory. Pain is perceived through a variety of neurophysiological mechanisms. Essentially, pain stimuli activate nociceptors in the skin, mucous membranes, or viscera, resulting in the transmission of impulses by A delta and C sensory nerve fibres. The former conducts “fast pain” which is well defined, of short duration and frequently described as sharp, pricking and localised. The latter transmit “slow pain” which tends to be diffuse and described in terms of aching or throbbing. The pain impulses are conducted to the dorsal horn of the spinal cord where they cross the neural synapse and are transmitted via ascending nerve pathways to synapse at the level of the thalamus. They are then conducted on tertiary afferents to the cerebral cortex where the stimulus is perceived by the brain as pain. Other areas of the brain involved in pain perception include the reticular activating system (which is concerned with arousal and awareness), the limbic system (which controls the emotional aspects of behaviour) and the motor pathways (Serpell et al, 1998). The above description provides a simplified account of the physiological mechanisms of pain perception. It was as a result of an increasing need to acknowledge the complexity of pain, including the psychological influences, that Melzack and Wall first postulated the Gate Control theory in 1965. Subsequent basic and clinical research, by both Melzack and Wall (1996) and other pain specialists, continues to add to the theory which is now accepted in Western medicine as the most influential current explanation of pain perception (Rawal, 1998). In brief, the Gate Control theory proposes that pain impulses are regulated by activity in the central nervous system which provides a “gating” mechanism to pain perception. An area of activity was identified in the substantia gelatinosa that comprises layers (laminae) of specialised neurones in the dorsal horn extending the length of the spinal column. Two main mechanisms are described - peripheral action and central action. In peripheral mechanisms, sensory input through large A beta touch fibre stimulation activates inhibitory interneurones within the substantia gelatinosa. As a consequence, pain impulses, carried by the A delta and C fibres, are blocked from reaching the WDR (wide dynamic range) neurones (previously termed transmitter, or T, cells) located deep within the laminae (Serpell, 1998). As a result, the “gate closes”. It is thought that pain relieving techniques such as ice massage (DeCrosta, 1984), heat and cold (Mobily et al, 1994), touch (Wilkinson, 1996) and transcutaneous electrical nerve stimulation (TENS) (Stevenson, 1995), work through these peripheral actions. Central mechanisms involve psychological and cognitive processing, which through descending impulses activate the inhibitory inter-neurones within the substantia gelatinosa, thus, again preventing, or modifying, pain impulse transmission across the gate. For example, several studies have indicated the benefits of providing preoperative information to help reduce patients postoperative pain experiences (Mogan et al, 1985; Wilder-Smith and Schuler, 1992; Abude et al, 1999). Information giving activates cognitive processing and may also reduce anxiety through stimulating the limbic system (Cousins and Power, 1999; Craig, 1999; Weisenberg, 1999). The limbic system is connected, via the reticular activating system, to the nociceptive neurones in the dorsal horn and together with cognitive influences can inhibit pain transmission (Cousins and Power, 1999). The influence of central descending impulses helps to explain the mind-body link. Further explanation of the Gate Control theory can be found in the seminal texts of Melzack and Wall (1965, 1996) and the fourth edition of the Textbook of Pain (Wall and Melzack, 1999). The Gate Control theory has laid the foundations for a new understanding of pain perception as a multi-dimensional experience, containing physical and psychological components. With the mind-body link being explicated in 1965, Melzack and Casey went on to propose a three dimensional model of pain in 1968. This and other multi-dimensional frameworks are presented in the section below. 1.4 Pain As a Multidimensional Phenomenon In an early extension of the Gate Control theory, Melzack and Casey (1968) and Melzack (1973), proposed three dimensions to the pain experience. The first of these is the sensory-discriminative determinant, which is concerned with the recording and analysis of the sensory features of pain such as intensity and quality. The motivational-affective component focuses upon the registering of emotion and behavioural responses. Occurring within the limbic and reticular activating systems, this dimension deals with, for instance, fear, anxiety and pleasure. Finally, Melzack (1973) proposed a cognitive-evaluative dimension concerned with evaluating, interpreting and conceptualising pain in the higher centres of the cerebral cortex. Psychological components such as expectation, attention and memory are acknowledged within this dimension. The three determinants are seen to be interactive and integrated in any single pain experience. In a further expansion of the model, Price (1988) suggested that a causal relationship might exist between the three dimensions. He hypothesised that the manner by which sensory features of pain are cognitively appraised, affects emotional responses to pain and the ability to take action, seek help, or employ pain coping strategies. The three dimensional model, underpinned by the Gate Control theory, has continued to prompt researchers to consider the multidimensionality of pain. Ahles et al (1983) described a five dimensional conceptualisation for cancer pain, which includes physiological, sensory, affective, cognitive and behavioural components. The model was tested and confirmed with 40 cancer patients in a study that focused upon the ability to measure the different dimensions in clinical practice. As a result of the work of Ahles et al (1983), McGuire (1992) developed a multidimensional framework for the assessment and nursing management of pain. According to McGuire (1992), pain incorporates six dimensions. The physiological dimension encompasses physical, neurological and biochemical aspects, while the sensory component addresses the location, intensity and quality of pain. Emotional responses are considered in the affective dimension, and the manner in which pain influences individuals’ thought processes, attitudes, beliefs, and preferred strategies for dealing with pain, are addressed by the cognitive element. The behavioural dimension includes a variety of behaviours associated with pain perception and attempts by the individual to control pain, while the sociocultural component deals with the cultural and environmental influences on pain. McGuire’s (1992) framework extends the three dimensions proposed by Melzack (1973) and the five identified by Ahles et al (1983) through the addition of a sociocultural dimension. There is also a less physiological feel to McGuire’s (1992) proposal and a greater sense of a person-centred approach. Although the elements of each dimension may overlap, the multidimensional framework is a useful way of conceptualising the pain experience. McGuire’s (1992) framework was adopted, in the United States, by the National Institute of Nursing Research (1994) in their report on acute pain management. It is discussed further below in relation to the factors that need to be taken into account for the proposed study: The physiological dimension encompasses the basic anatomical structures and physiological mechanisms involved in pain perception and the gating mechanisms described previously. In acute postoperative pain, the stimulus arises from surgical tissue damage to the superficial, deep somatic, and visceral layers (Heye, 1991). Peripherally, nociceptive stimulation activates the release of substance P and agents such as bradykinin and noradrenaline that increase sensitisation to painful stimuli. Centrally, noxious input to the spinal cord increases sympathetic activity resulting in raised blood pressure, heart rate, oxygen consumption, muscle tension and sphincter tone in the bladder and bowels. Stimulation of the cardiovascular, respiratory and hypothalamic centres in the brain, further increase activity and give rise to the signs and symptoms commonly associated with acute pain (Cousins and Power, 1999). The sensory dimension addresses the location of pain, together with its intensity, quality and pattern. Acute postoperative pain is usually well localised at the incision site, although it may be referred elsewhere. It should also be noted that the age and disease aetiology of postoperative populations means that there may be other pain foci to consider following surgery. The intensity of postoperative pain tends to reduce with time, however research continues to suggest that 31 - 65% of patients experience moderate to severe pain 3 - 7 days following surgery (Seers, 1987; Royal College of Surgeons of England and the College of Anaesthetists, 1990; Bruster et al, 1994; Hall-Lord et al, 1998). This duration may be extended in older populations (Blomqvist and Hallberg, 1999). Pain intensity is influenced by the effectiveness of pain management and factors such as emotional state, fatigue and environment, which in turn impact upon the individual’s ability to tolerate and cope with pain. The quality of postoperative pain is frequently defined through descriptor words which suggest that the pattern of perception is one of A delta stimulation, resulting in sharp, stabbing pain, gradually replaced by perceptions of C fibre stimulation associated with dull and aching pain (Carr, 1990). In older people, A delta stimulation may be compromised through demyelination of the nerves as a result of the ageing process. Consequently, the pain descriptors used by the elderly may be those of C fibre stimulation (Chakour et al, 1996). McGuire (1992) has described the affective dimension as encompassing emotional states such as distress, anxiety and helplessness. In acute postoperative pain, it has long been known that anxiety and pain are associated. The classical nursing studies of Hayward (1973), Boore (1978) and Mogan et al (1985), have demonstrated that by providing preoperative temporal and sensory information, patients' appraisal of their surgery as an unknown and threatening event is reduced. Consequent reductions in anxiety are matched by reductions in reported pain and the amount of analgesics required. More recently, Walding (1991) has suggested that there may be a link between pain, anxiety and perceived powerlessness. She hypothesised that if perceptions of powerlessness are decreased, personal control may be enhanced, coping promoted and pain perception reduced. However, this hypothesis has not been tested. The cognitive dimension comprises aspects such as the meaning of pain, cognitive processes, knowledge and coping strategies. In surgical populations, motivation may be high if the surgery is viewed as a means of reducing, or even negating, pain, such as following total hip and knee replacements. Enhancing knowledge through preoperative information giving, has, as mentioned previously, been shown to reduce postoperative pain intensity and distress. The use of self-generated strategies which modify pain perception through peripheral and central mechanisms, has been shown to help chronic pain patients decrease or control their pain experiences (Jensen et al, 1991a; Keefe et al, 1997a). In acute pain management, the focus has tended to be upon taught strategies (McCaffery, 1999). The use of coping strategies for acute and chronic pain is returned to later in this chapter. The behavioural dimension encompasses both behaviours arising as a response to pain and those initiated in attempts to control pain. Texts on pain management frequently provide lists of overt signs of acute pain, such as facial grimacing, moaning, supporting painful body parts, rigid posture, or restlessness (Cousins and Power, 1999). This has given rise to assumptions that patients who appear to be relaxed, laughing with visitors, or sleeping, are not experiencing acute pain. However, lack of pain behaviour, even with postoperative pain, may be more indicative of physiological and behavioural adaptation than it is of lack of pain (McCaffery, 1999). Attempts to control pain, through for instance the use of distraction techniques, relaxation, or resting peacefully, may emphasise the lack of pain behaviour. The sociocultural dimension focuses upon the impact of ethnicity, culture, family norms, work and leisure. The ethnic status of the population in the proposed study is, according to the 1991 local census data, predominantly Caucasian (97%) and skewed towards the higher social classes. McDonald (1994) describes how, in a sample of 180 surgical patients, Caucasians received more analgesics than other ethnic groups despite a stated wariness of medication. The characteristics of coping strategies in the proposed study population are unknown, as most of the work carried out to date is North American in origin. The goal of postoperative pain management is to ensure sufficient pain control to reduce psychophysiological distress, enhance mobility and promote recovery (Carr, 1990; Graffam, 1990; Closs, 1992). Despite the multidimensional nature of pain, analgesics are frequently the sole relief measure offered to patients in postoperative pain. Several authors comment upon their limitations in terms of both pharmacological action and ability to treat a multidimensional phenomenon (Donovan et al, 1987; Wells, 1987; Walker and Campbell, 1988; Melzack and Wall, 1996; Sofaer, 1998; McCaffery, 1999). They recommend that decisions about pain management should be based upon Melzack and Wall's (1996) Gate Control theory and a multidimensional approach. By using a variety of non-pharmacological strategies that prompt both peripheral and central action on the gate, it is possible that analgesic effect may be enhanced and holistic pain management improved. In advocating such an approach for acute pain management, it is important to note that this premise has not previously been tested in acute pain settings. However, the use of non-pharmacological strategies has been examined in chronic pain management and the next section considers some of the relevant literature. 1.5 Strategies for Chronic Pain Specialists in chronic pain management have argued that patients need to develop strategies to enable them to cope with their pain and re-gain control over their lives (Flor and Turk, 1988; Williams et al, 1994; Williams, 1995; Arnstein et al, 1999). The literature indicates that adults with persistent pain develop strategies through both self-generated and taught means. A brief review of this literature follows. Studies with chronic and cancer pain patients have attempted to identify and promote individuals’ self-generated strategies in an effort to enhance perceptions of control and reduce pain intensity (Jensen et al, 1991a; Sofaer and Walker, 1994; Williams et al, 1994; Arnstein et al, 1999). However, not all self-generated strategies are viewed in the literature as positive techniques for controlling pain. For instance, catastrophising (negative thinking) has been found to result in higher clinical reports of pain and poorer outcomes (Keefe et al, 1997b). Other strategies, such as distraction and positive thinking, when viewed as adaptive, are linked to lower pain reports (Jensen et al 1991a; Craig, 1999; Weisenberg, 1999). The Coping Strategy Questionnaire (CSQ), developed by Rosential and Keefe (1983), was evolved to measure patients’ coping strategies in seven different categories: coping self-statements, praying/hoping, ignoring pain sensations, reinterpreting pain sensations, increasing behavioural activities, catastrophising, and diverting attention. Patients are asked to rate each item on a 0 (never) to 6 (always) scale according to how often they use the strategy when in pain. Various studies have found the CSQ to be a reliable and valid instrument for assessing cognitive coping strategies and predicting chronic pain, psychological distress, disability and pain behaviour (Parker et al, 1989; Beckham et al, 1991; Keefe et al, 1992). Four selected examples of studies on pain coping (three of which used the CSQ) are reviewed below: Schanberg et al (1997) found that a modified version of the CSQ predicted ability to cope with pain in a group of 56 adolescents and young adults with juvenile arthritis. Subjects, who rated their ability to control and decrease pain as high, had significantly lower pain intensity ratings. With demographic and medical status controlled for, coping explained a significant element of variance in the pain intensity measures. However, Schanberg et al’s (1997) study should be regarded with care in view of the relatively small sample. They suggest that a further limitation relates to the predominance of Caucasians (95%) in the sample, however this figure is similar to that of the local population in the proposed study. Of potentially greater concern, is the question of whether a UK population demonstrates similar results to those found in North America, given that coping strategies are learnt and may therefore be affected by sociocultural norms. Schanberg et al (1997) support the findings of Gil et al (1993, 1997) that coping style tends to stabilise as adulthood is reached and that change in style during adolescence give rise to increased catastrophising. Schanberg et al (1997) rather summarily conclude that health care interventions should focus upon teaching effective coping strategies in order to prevent the development of negative self-generated strategies and promote pain self-efficacy. Bandura (1977, 1986) defines self-efficacy as the belief that an individual has in their ability to engage in a particular action, such that a desired outcome is achieved. Such beliefs are said to significantly influence the initiation and continuation of specified behaviours. According to Bandura (1997), self-efficacy beliefs can be refined through the development of skill mastery, the sharing of vicarious experiences, verbal persuasion and the provision of information related to the individual’s physiological and affective state. A number of studies in both laboratories and chronic pain settings have demonstrated the utility of self-efficacy theory in relation to pain perception and control (for example: Bandura et al, 1987; Jenson et al, 1991a, 1991b; Williams and Kinney, 1991; Williams et al, 1993; Strong, 1995; Barlow et al, 1996). These indicate that patients consistently use strategies that they believe are effective in relieving pain and that the more efficacious people believe their strategies are the greater their pain tolerance. Certain studies have also shown that pain self-efficacy impacts positively on activity, quality of life and perceived disability (Manne and Zautra, 1992; Taal et al, 1993; Williams et al, 1993; Arnstein et al, 1999). Others have demonstrated relationships with reduced perceptions of powerlessness and enhanced control (Buckelow et al, 1992; Keefe et al, 1992, 1996, 1997a; Arnstein et al, 1999). One study, which started from the premise that individuals with similar degrees of arthritis differ greatly in their pain reports and confidence to cope with pain, examined the relationship of pain coping strategies and self-efficacy in 130 patients with osteoarthritis of the knee (Keefe et al, 1997a). The most frequently used coping strategies, as measured by the CSQ, were coping self-statements, increasing behavioural activities and ignoring pain sensations. In a subsequent regression analysis, the authors found that patients scoring high on ignoring pain sensations had higher pain self-efficacy (as measured by the Arthritis Self-Efficacy Scale), while those with high catastrophising scores had lower self-efficacy for pain. High levels of pain intensity were also related to lower self-efficacy as might be expected. After controlling for pain intensity, pain coping strategies were still found to be associated with self-efficacy. Keefe et al (1997a) concluded that the significance of their study was in identifying adaptive and maladaptive strategies. They suggested that health care professionals might consider focusing interventions on identifying patients’ coping styles and actively intervening to reduce negative, and enhance positive, strategies. They point out however, that both their own and similar work is as yet unable to specify whether successful pain coping leads to a belief that the strategy works, or whether it is the belief in the strategy that results in successful coping. Furthermore, questionnaire based studies, such as Keefe et al’s (1997a), do not address individual perspectives on the qualitative nature of pain coping strategies. A study by Large and Strong (1997) attempted to investigate the qualitative nature of pain coping using focus groups and a repertory grid. Results from the focus groups concluded that a wide range of strategies was used and that strategy use is determined by the specific situation. Data from the repertory grids indicated that participants’ view coping as “a necessary evil” for those in chronic pain, and on the whole, regarded themselves as copers. The authors submit that there is a suggestion in the data that level of pain and ability to cope are related. However, some care should be taken with the findings from this study. The sample was small (n = 19) and participants were self-selected following a media advertisement requesting volunteers for a study exploring how people with chronic low back pain manage their pain. According to Large and Strong (1997), the mean age of 54 years and the average number of years in pain, at approximately 17 years, is somewhat higher than found in studies involving clinical samples. Furthermore, the authors comment that participants had lower pain intensities in comparison to clinical samples. Nevertheless, the study adds to the literature and acknowledges the potential for exploring coping strategies in non-patient pain groups with a view to enhancing professional understanding of the implications for teaching pain relieving strategies. Ter Kuile et al (1995) studied 144 people with chronic headaches who participated in a pain management programme. They hypothesised that cognitive training, which explicitly attempts to alter pain appraisal and develop cognitive strategies, would be more effective than relaxation. Both interventions resulted in patients using more coping strategies, as measured by the CSQ, with the exception of praying/hoping, which remained consistent, and catastrophising which was reduced. However, cognitive training was found to be more effective in promoting change, particularly in respect of strategies aimed at diverting attention and promoting coping self-statements. Pain self-efficacy was also enhanced in both groups. Unfortunately, there was no comparative control group against which the effectiveness of the interventions could be tested. Furthermore, subjects were a mix of outpatient attendees, university students and a self-selected group who responded to a newspaper advertisement. No mention was made of how they were allocated to the intervention groups. According to McQuay and Moore (1998), failure to randomly allocate subjects, such that each has an equal chance of being included in the different study groups, increases the risk of selection bias resulting in a possible overestimation of treatment effect by as much as 41%. Despite its rather questionable quality, it is noteworthy that Ter Kuile et al’s (1995) work is one of the few experimental studies found in the literature on chronic pain coping strategies. The studies discussed above have all been undertaken since a comprehensive review of the literature in this area by Jensen et al (1991a). The reviewers concluded that evidence exists to support the theory that believing one can employ strategies to enable adaptation to chronic pain is positively related to adjustment. In other words there is evidence for the utility of self-efficacy in chronic pain management. However, while it appears that cognitive coping strategies impact upon the chronic pain experience, evidence of a positive relationship with measures of psychological and physical functioning remain inconclusive (Williams, 1995). As a consequence, much of the work to date has focused upon teaching positive cognitive and behavioural strategies in an attempt to enhance self-efficacy and reduce perceptions of pain intensity and distress (McCracken, 1991; Keefe et al, 1992; Lester and Keefe, 1998; Turner and Keefe, 1999). Whether adaptive coping results in lower pain intensity and distress, or whether less pain enables more positive coping, is unclear. Equally, it is evident that research designs are compromised by the current inability to specify measurable outcomes that reflect the totality of the multidimensional conceptualisation of pain. Williams (1995) suggests that this situation is beginning to be addressed as researchers develop and test new tools which attempt to reflect the different dimensions of pain. The propensity for teaching coping strategies, and the assumption that this is “a good thing to do”, has been picked up in acute pain management where the literature focuses (mainly through experimental studies in surgical settings) on assessing the efficacy of taught strategies against a range of measurable postoperative outcomes. The next section considers these types of strategies and the evidence for their effectiveness. 1.6 Teaching Strategies for Acute Pain Management There appears to be an acceptance in the nursing literature of the benefits of teaching non-pharmacological strategies to patients undergoing surgery. Indeed the classical texts, such as those by McCaffery and Beebe (1994), Fordham and Dunn (1994), and Sofaer (1998), which are often used as a basis for nurse education on pain management in the UK, contain chapters exhorting the use of such strategies. However, there are two issues of interest: firstly whether the suggested techniques do relieve acute pain, and secondly, whether they are used in practice. Each of these issues is addressed in turn. 1.6.1 The Evidence The evidence-based practice movement asserts that the best sources for questions about effectiveness are obtained from systematic reviews of multiple well-designed randomised controlled trials (RCTs) (Sackett et al, 1996). Failing this, single quality RCTs are advocated as providing the best opportunity to determine the effectiveness of interventions. The quality indicators by which a trial is judged are based upon whether or not the study includes an appropriate randomisation process, is double-blind and provides commentary on the number and reasons for subject withdrawal (McQuay and Moore, 1998). Two systematic reviews and a meta-analysis related to taught non-pharmacological strategies are reviewed below. The Critical Appraisal Skills Programme guidelines were used to assist in this review (CASP, 1997). The systematic review by Seers and Carroll (1998), on relaxation techniques for acute pain management, identified 40 published studies through a comprehensive search of electronic databases combined with secondary searching from reference lists. Reports were included if they were focused upon relaxation alone, were RCTs with samples of 10 or more, and measured pain related outcomes. Given the nature of the intervention, double blinding was not enforced as an inclusion criterion. After applying Jadad et al’s (1996) quality scoring system, thirty-three studies were excluded from the review. Eleven were not RCTs, seven did not record pain outcomes and the remainder combined relaxation with other interventions or had samples of 10 or less. Of note, 15 of the 22 excluded RCTs were broadly positive about the impact of relaxation on pain. The seven studies covered by the review involved a total of 362 patients. Outcome measures included pain intensity and distress, anxiety, and analgesic consumption. Three studies reported generally positive results in the intervention groups, suggesting that relaxation is effective in reducing postoperative pain. However, the remaining four demonstrated no significant differences in the outcome measures. Seers and Carroll (1998) noted that most of the studies contained design limitations related to sample size, differences in relaxation intervention techniques, the validity of outcome measurements and the clinical significance of the findings. They conclude that the evidence for relaxation being effective in the management of postoperative pain is weak and that there is a need for quality RCTs in this area. Seers and Carroll’s (1998) review may have been limited by not including unpublished studies and by over-stringent application of the quality criteria. Nevertheless, the questionable quality of the excluded studies is apparent. It must be concluded that the evidence for nurses spending time teaching relaxation techniques to surgical patients is currently weak. McCaffery (1999) advocates the use of music as a distraction and emotional relaxation technique to reduce postoperative pain perception through cognitive and affective mechanisms. For a recent systematic review of the effect of music on postoperative pain and anxiety, Grenside (1997) retrieved 57 articles via an electronic search of both health care and music databases from 1979 onwards, and a further 56 articles through secondary searching and hand searches. The initial criteria for inclusion in the systematic review were described as clinical studies using qualitative or quantitative approaches, which had been published from 1987 onwards. Rationale for the decision to include qualitative studies was based upon Grenside’s (1997) knowledge of music research and a suggestion by Good (1996) in her review of the effects of relaxation and music on postoperative pain. However, following the initial literature review, the inclusion criteria were revised to randomised studies of 20 or more subjects. A survey instrument was developed to enable critical review of the eight studies (n = 370) identified for analysis. On closer examination, two of the studies while claiming to be randomised appeared to have used a matched pairs design and one of these produced results on only 10 of the original 34 in the sample. Six of the eight studies did not report how their sample sizes were calculated, an issue that Grenside (1997) suggests may affect validity as the samples may be too small to determine statistical significance. A further criticism offered by Grenside (1997) relates to the disparate nature of the samples in terms of the type of surgical experience that subjects underwent. In systematic reviews, wide differences between samples and other study characteristics are known as heterogeneity. The greater the heterogeneity, the less reliable the findings (Chalmers and Altman, 1995). Pain intensity, as measured by visual or numerical analogue scales, was a common outcome measure as was analgesic consumption. However, there were differences in how the latter was calculated. Qualitative data were also collected in five of the studies. Grenside (1997) concluded that while the qualitative data suggested improvements in patient satisfaction, pain coping and affective issues such as reassurance, relief of boredom and perceived relaxation, statistical analysis of pain intensity measures were inconclusive regarding the effectiveness of music on postoperative pain perception. Thus it appears that evidence is again lacking in relation to nurses spending time teaching surgical patients non-pharmacological strategies, such as the use of music for postoperative pain management. It should be noted that Grenside (1997) believes that the internal validity of the reviewed studies may be flawed given the difficulty of controlling and measuring a variable as subjective as music enjoyment. Indeed, she concludes her review by advocating that health care researchers need an understanding of music to address confounding variables such as the characteristics of music and music preference. Grenside’s (1997) review was limited by the lack of quality RCTs conducted on the effect of music on pain which prevented her from adopting stringent quality indicators to determine inclusion criteria. In a meta-analysis of non-pharmacological nursing interventions for acute pain, Sindhu (1996) analysed 49 studies examining a variety of techniques including biofeedback, relaxation, music, imagery, coping strategies and self-care education. Study reports were obtained through electronic health care database searches from 1960 onwards, manual searches of key journals, secondary searching, and contact with experts in the field to identify unpublished work. The quality inclusion criteria were not fully described in the article, although it was stated that only RCTs were included and that the main outcome was pain intensity as measured on a visual analogue scale. The 49 trials involved a total sample of 3,357 subjects. The pooled effectiveness of the studies suggested that there is little current evidence to support the effectiveness of taught non-pharmacological strategies. However, Sindhu’s (1996) review can be criticised on the basis of the large number of different interventions included. Furthermore, there were also wide differences in sample characteristics and outcome measurements. Thus, heterogeneity proved a considerable problem in this review. As a result, Sindhu’s (1996) conclusions must be regarded with care. However, this review again suffered from a lack of quality RCTs on the non-pharmacological nursing management of pain. The systematic reviews by Seers and Carroll (1998), Grenside (1997), and Sindhu (1996), are all limited by the quality of available research on non-pharmacological pain relieving strategies. Studies in this area are frequently not randomised, suffer from small samples, combine interventions, or contain other design flaws. Consequently, reviewers are restricted to small numbers of papers that tend to be heterogeneous. The absence of evidence for the effectiveness of taught non-pharmacological strategies is not the same as saying that there is no effect. As the authors of the above reviews point out, there is a need for high quality RCTs to establish what, if any, effect taught non-pharmacological strategies have on acute postoperative pain. It should also be noted that although systematic reviews are helpful in identifying effective interventions, they cannot capture the full, subjective, experience of pain (McQuay & Moore, 1998) and therefore need to be supported by rigorous studies which address the qualitative aspects of pain and pain management. 1.6.2 From Research to Practice Having considered the effectiveness of taught non-pharmacological strategies, the evidence related to their use in practice is now addressed. Astley (1990) suggests that prior to the introduction of analgesics a range of non-pharmacological strategies were used to relieve pain. However, most of these disappeared from everyday use earlier this century. It is difficult to assess whether or not nursing and medical staff are currently teaching non-pharmacological techniques. To the author’s knowledge, there are no papers available in the literature specifically addressing this issue. However, several researchers do provide some indication of practice in this area. Carr and Thomas’ (1997) study, on patients’ expectations and experiences of pain following major abdominal surgery, found some evidence of nurses employing non-pharmacological strategies. Patients provided narrative examples of receiving emotional support and comforting touch when in pain. A second study by Nagy (1999) investigated the techniques used by nurses to cope with the pain experienced by burns patients. One of the methods described was to “engage with the patient’s pain”. This was operationalised through the provision of sensory and temporal information, sharing control, employing all the non-pharmacological measures known to the nurse, emotional comfort and touch. In a qualitative analysis of postoperative nursing documentation, Briggs and Dean (1998) found little evidence for the use of non-pharmacological strategies when compared with analgesics. Of the 65 care plans reviewed, only 22 identified pain as a problem. This was despite the fact that 59 of the patients had informed the researchers that they experienced postoperative pain. Eighteen of the 22 care plans which acknowledged pain included statements related to nursing the patient in a comfortable position. A further eight suggested the use of distraction techniques, while two included the provision of sensory and temporal information and one specified using pillows and warmth. The lack of information regarding non-pharmacological strategies in the remaining 37 care plans may have been a documentation oversight, however the authors believe that the omission is in fact a reflection of practice rather than being related to the completion of documentation. In a previous study undertaken by Burrows (1997b), there was evidence that nurses knew about and employed a range of taught non-pharmacological strategies. Part of the study involved an educational programme in which participants shared their knowledge with each other. Subsequent interviews with the nursing team indicated that nurses were actively using strategies such as imagery, touch, relaxation, talking, positioning, distraction, warm baths, and ice to assist with patients postoperative pain. Non-pharmacological strategies, such as those discussed above, can be regarded as patient-centred and are thought to enhance perceptions of control. However the emphasis in the literature is very much one of teaching strategies for acute postoperative pain management, rather than identifying and building upon what the individual already knows and uses. In an RCT focused upon teaching non-pharmacological strategies to patients undergoing dental surgery, Chaves and Brown (1987) discovered that 28 of the 63 subjects used a variety of spontaneous strategies in addition to those being taught. They concluded that: “there are indications that administered cognitive strategies may interact in complex ways with spontaneous self-generated strategies, that are already within the patient’s repertoire” (p.276). The importance of Chaves and Brown’s (1987) work is that it provides evidence for the use of self-generated pain relieving strategies in acute pain episodes. Furthermore, both intervention and control subjects were employing self-generated strategies, thus adding an unanticipated variable to the clinical trial. It is possible that studies on taught strategies have been compromised by a lack of awareness regarding the existence of spontaneous self-generated strategies. For instance, subjects enrolled on a trial on taught relaxation may use their own relaxation strategy in addition to, or in place of, the taught technique. The following section of this chapter considers the evidence for the use and effectiveness of self-generated acute pain relieving strategies. 1.7 Self-Generated Strategies in Acute Pain Management The existence and use of self-generated strategies in acute pain management can be informed by publications on both taught acute and chronic pain strategies and on self-generated chronic pain strategies. Two questions from the chronic pain literature arise: firstly, do patients develop strategies when initially faced with their chronic pain problem? Secondly, do they already have experience of using strategies to modulate “everyday” pain, such as headaches, dysmenorrhoea and mild sports injuries? Laboratory research indicates that individuals do have experience of pain coping strategies and also have preferences in relation to the strategy used. Studies by both Worthington (1978) and Stevens et al (1987) found that subjects demonstrate flexibility in coping with experimental pain and that those offered a choice of strategies show a greater reduction in pain than those not given a choice. Rokke and al’Absi’s (1992) research also demonstrated that subjects who were offered a choice doubled their tolerance time to experimental pain, as did those assigned a strategy matched against a high score on the Cognitive Control Strategy Inventory (CCSI). Conversely, those assigned strategies matched against a low score, which would suggest poor coping, dealt less well with the laboratory inflicted pain stimulus. Rokke and al’Absi (1992) concluded that offering subjects a choice is more cost effective and more efficient than administering the CCSI and teaching matching strategies for pain relief. In a community-based study, Broome (1986) surveyed 109 students to identify the strategies that they used to relieve pain at home. Thirteen cognitive and behavioural strategies were identified as being used by individuals to enhance everyday pain relief. Some differences were found between the methods most frequently used and those perceived as most effective. However, the most consistently beneficial method, as identified through subjects’ self-reports, was their own relaxation technique. Broome (1986) concluded that if health care professionals identified individuals’ strategies it might be possible to continue their use in hospital. Prompted by an interest in the variability of pain reports during mammography, Kashikar-Zuck et al (1997) used the CSQ to examine pain self-efficacy and pain coping in 125 female outpatients. They hypothesised that women using strategies developed through coping with everyday pain, may report less pain during mammography. The results indicated that women, who believed that their strategies would decrease pain, did in fact report less pain. However, although they used a wide range of pain coping strategies there was no correlation between the type of strategy employed by the subjects and the intensity and quality of pain reported. Interestingly, Kashikar-Zuck et al’s (1997) results indicated that subjects used considerably less catastrophising than many chronic pain patients. They concluded that the relatively high catastrophising scores reported by individuals with low back and myofacial pain result from the persistence of the pain, whereas patients who are more likely to experience episodic pain, such as those with rheumatoid and osteoarthritis, tend to have lower catastrophising scores. The authors did not link their own findings to the duration of the mammography stimuli, however it is possible that the briefness of mammography pain is relevant to pain self-efficacy. Kashikar-Zuck et al’s (1997) discussion raises a potentially interesting point: individuals who have developed strategies to cope with everyday pain may demonstrate more adaptive strategies and a greater self-efficacy for decreasing pain postoperatively. Alternatively, the intensity and uniqueness of postoperative pain may result in lowered pain self-efficacy. In a qualitative clinical study with 148 postoperative inpatients, Copp (1974) conducted interviews to identify the way in which people interpret acute pain and the strategies they use to cope with it. A number of strategies were mentioned, including focusing, counting, visualisation, prayer and imagery. Many patients stated that they preferred these strategies to medication, but felt that health care staff might disapprove of their use in hospital. Consequently, patients were not employing their strategies. It may have been interesting to assess whether this influenced pain self-efficacy. If the Gate Control theory and McGuire’s (1992) multidimensional framework apply to acute as well as chronic pain, it could be suggested that patients experiencing acute pain will demonstrate emotional responses in the affective domain and strategy use in the cognitive and behavioural dimensions. Strategy use may include the methods referred to above and in previous sections of this chapter, taking painkillers, or other less common techniques. They may initially be employed in response to everyday pain and subsequently developed by individuals when faced with more substantial acute or chronic pain episodes. In an attempt to identify the range of acute pain coping styles, Butler et al (1989) developed the CCSI and administered it to a group of postoperative patients in North America. Subjects were also interviewed to ascertain whether they used any additional strategies not assessed by the CCSI. Those identified were then included on the inventory. Following several alterations, the tool was re-tested. Results indicated that the inventory appears to reliably assess cognitive coping strategies in postoperative patients. Butler et al (1989) further noted that analgesic consumption and self-reports of pain perception were reduced, while pain tolerance was enhanced in patients employing cognitive coping strategies. As pain responses are influenced by cultural considerations, the inventory needs testing for external validity with patients in the United Kingdom. Furthermore, pain and its management involve more than cognitive issues, thus a variety of cognitive and behavioural strategies may be needed to enhance tolerance and achieve effective pain control (Weis et al, 1983). Coping strategy questionnaires need to reflect this perspective. In the UK, a survey by Pick et al (1990) found that postoperative patients used a variety of self-generated pain coping strategies and that pain intensity and distress were reduced in those using positive techniques. The authors concluded that there is a need to identify the range of strategies used by patients in the UK and to explore ways of incorporating positive strategies into postoperative pain management. In their more recent qualitative study, Carr and Thomas (1997) employed interviews to elicit 10 patients’ expectations and experiences of pain following major abdominal surgery. Data indicated that participants used self-initiated distraction techniques and actively sought emotional support, empathy and touch in their attempts to reduce pain. However, there were few narrative examples of nurses directly supporting this activity. Although small and not directly focused upon non-pharmacological strategy use, this study does support the view that patients may use self-generated strategies for postoperative pain management. The Burrows pain assessment chart (Bpac) was developed by the author and subsequently tested for utility in a small experimental study (Burrows, 1990, 1997b). The Bpac incorporates assessment of pain and patients’ self generated pain relieving strategies, planning and intervention based on the use of these strategies and evaluation of any resulting pain relief. Although the initial study provided some evidence for the possible benefits of identifying and using patients' strategies, nurses’ knowledge of, and therefore ability to support interventions other than analgesics, was found to be limited. This resulted in an inability to fully incorporate patients’ self generated strategies into the plan of care. A second study, using an action research design, was subsequently undertaken to evaluate the effectiveness of a course of preparation to enable staff to assess patients’ strategies using the Bpac and to support these strategies postoperatively. Although the study was successful in achieving its aims, no data were collected directly from patients. Thus, although recordings on the Bpac demonstrated that patients employed a variety of strategies for everyday pain, and continued to use these postoperatively, there were no data indicating the range, or type of strategies, or the efficacy of these strategies in decreasing postoperative pain (Burrows, 1997b). In summary, evidence exists to support the premise that individuals employ self-generated strategies in response to everyday pain experiences. These strategies may then be developed further during acute and chronic pain episodes. Strategy use appears to be correlated to pain self-efficacy and to decreases in pain perception. In the acute pain setting, there is a need to develop nurses’ understanding of the relevance of adopting a multidimensional approach to pain management in order to promote pain control. One potential mechanism would be to combine analgesic management with identifying patients’ pain coping strategies and supporting them postoperatively (Broome, 1986; Burrows, 1997). 1.8 Pain Coping: Engaging Patients and Nurses in Postoperative Pain Management The terms “pain coping” and “coping strategies” have been used in the preceding discussion without acknowledgement of the literature upon coping, or a definition of coping being offered. This section briefly reviews the literature. Collins Paperback English Dictionary (1999) defines coping, firstly, as an ability to tolerate or endure and secondly, as an ability to deal successfully with situations, such as stress or pain. Historically, coping has been viewed as having trait (personality dependent) and state (situation dependent) elements (Monat and Lazarus, 1991; Shaw, 1999). However, Lazarus and Folkman (1984) highlight the dynamic nature of coping, which they describe as the cognitive and behavioural efforts that are continually altered to enable the individual to manage the internal and external demands made upon them. These demands are interpreted as troublesome to the individual or exceeding their resources. In contrast to Collins Paperback English Dictionary (1999), Lazarus and Folkman (1984) further suggest that the cognitive and behavioural efforts may be either positive or passive. Thus, they would view catastrophising as a coping strategy for pain. Coping strategies are developed in order to manage, or alter, the external or internal demand and to regulate and reinterpret emotional responses. According to Folkman and Lazarus (1980, 1985), this is achieved through problem-focused coping which involves efforts to alter the situation and thereby eliminate the demand and by emotion-focused coping which involves reinterpreting the situation so that it is not perceived as a threat. Thus individuals need to appraise the situation (primary appraisal) and then ask themselves what they can do about it (secondary appraisal) (Folkman et al, 1986). Numbers (1986) suggests that determining what strategies to employ is highly personal and generally self-taught. The overall aim is to deal with the demand while maintaining self-esteem and promoting self-management. In Lazarus and Folkman’s (1984) model, coping is viewed as a dynamic process that changes over time as the individual is faced with new and different situations. Thus, each appraisal adds to the individual’s repertoire of both positive and negative coping strategies. It would therefore appear reasonable to suggest that adults admitted to hospital, have, through their appraisal of everyday pain experiences, developed a repertoire of pain coping strategies. Faced with a surgical episode, these strategies may be inadequate and be abandoned, or they may be adapted and employed together with analgesics to relieve postoperative pain. Although Lazarus and Folkman’s (1984) process model is helpful in furthering understanding of coping, it appears that trait elements may also be important. For instance, Schwengkmezger et al’s (1990) study on haemodialysis patients found that coping styles were highly stable over time. It is possible that individuals have a range of coping strategies which can be matched to different situations, although each has preferred choices based upon their personality (Miro and Raich, 1999; Shaw, 1999). From a multidimensional pain perspective, a number of factors are thought to influence the ability of an individual to cope with the demand of pain stimuli. These include physiological pain experiences, sensory perception, the emotional state of the individual, attitudes and beliefs about pain and sociocultural influences upon pain behaviour. In combination, these dimensions are important in initiating pain coping and determining strategy choice. Melzack and Wall (1996) maintain that the greater the suggestion that pain will be relieved, the greater the relief obtained by the patient. In pharmacological pain management this statement contributes to our understanding of the placebo response. In non-pharmacological management, self-efficacy is equally important. Wilkinson (1999) argues that if patients believe that they have a repertoire of strategies that work to relieve pain, health care professionals should seek to understand this belief and enable patients to become increasingly participative in their own care. This can only be achieved if the patient is regarded as an individual with a unique and expert perspective on their pain experience. Seers (1994) suggests that: “if you are not alongside the person with pain, you run the risk of managing their pain out of context” (p.viii). If the above perspective is followed, nurses will need to consult with patients to identify their strategies and engage them in a collaborative, patient-centred approach to postoperative pain management. These concepts are considered in greater depth in chapter two. 1.9 Conclusions Pain is a complex phenomenon the management of which requires sufficient diversity to take account of the multidimensional nature of pain. A variety of taught non-pharmacological strategies are described in the acute pain literature. However, both the chronic pain literature and the concept of coping suggest that individuals develop a repertoire of self-generated strategies in response to everyday pain. Some evidence exists to suggest that patients employ these strategies when faced with acute pain episodes, such as surgery. It is proposed that there is a need to identify the range of strategies used by a UK sample and to examine whether mechanisms can be developed to identify patients’ strategies upon admission to hospital and to support these strategies postoperatively. In an effort to achieve this, patients need to be engaged in their own pain management through nurses working within a collaborative, patient-centred approach to care. Chapter Two: Patient-Centred Pain Management 2.1 Introduction In the previous chapter, the idea of adopting a patient-centred approach to acute pain management was introduced. This chapter addresses patient-centred care in greater depth and considers its relationship to participation, collaboration, involvement, partnership and empowerment. The concept of empowerment is discussed and the issue of whether or not it is possible to empower others is deliberated. It is suggested that engaging patients in their own pain management requires a shift in thinking so that both patients and nurses are freed from existing beliefs. These beliefs have tended to place patients at the bottom of the power ladder in traditional pain management approaches. 2.2 Clarity or Confusion? The term patient-centred care is often used interchangeably with patient-collaboration, participation, involvement and partnerships (Cahill, 1998; Campbell, 1998). The lack of consensus regarding this terminology has resulted in an explosion of papers which aim to enhance understanding for practice by attempting to define the concepts as independent entities (for example: Biley, 1992; Stower, 1992; Heinneman et al, 1995; Cahill, 1998). However, Paley (1996) argues that conceptual clarification is frequently arbitrary, as the manner in which defining attributes are identified is often not explained. Paley (1996) suggests that nothing is: “ever classifiable in just one way, to the exclusion of other forms of classification” (p.574). Yet the literature continues to attempt to define the concepts, referred to above, in a mutually exclusive manner (for instance: Chadderton, 1995; Heineman et al, 1995). The outcome is ambiguity and confusion. Given that the terms are commonly used in every-day language, Paley (1996) contends that individuals will, in any case, confer meanings according to their own personal experiences, values and beliefs. Campbell (1998) supports this view when she suggests that the confusion surrounding certain concepts relate to attempts to extend definitions beyond a simple understanding of the word labels. The suggestion is that conceptual clarification of commonly used words cannot be achieved, as ambiguity is endemic in general language use. Defining terms within specified contexts may be useful. For instance, coping was defined within the context of pain coping in chapter one, while a definition of collaboration, related to the research design, is offered in chapter three. In the context of this chapter however, it is the author’s intention to minimise the number of definitions offered in order to reduce potential confusion and prevent a somewhat staid and repetitive discussion of meanings. Rather, the focus will be upon patient-centred care and the manner by which the related concepts collaboration, participation, involvement and partnership are incorporated within this approach. 2.3 Patient-Centred Care The contemporary political and health care climate is placing increasing emphasis upon patient-centred care. This emphasis arises from the drive towards consumerism in the NHS (Melville, 1997), which is based upon the messages contained within the White Papers: Working for Patients: The Health Service Caring for the 1990s and A First Class Service: Quality in the New NHS (DH, 1989; 1998). Both stress that the needs of patients must be the main focus of health care and that patient choice and participation in care must be extended. This perspective has been underlined in a recent Government communication which identifies five Modernisation Challenges to be incorporated into the new National Plan for the NHS (National Health Service Executive, 2000). These challenges include “prevention, partnership, patient care, performance and professions”. The patient care challenge is focused upon empowering and informing patients so that they can be involved in their own care. Patient-centred care demands that nurses work with patients to identify and meet individual biopsychosocial and spiritual needs (Titchen, 1996). Wilkinson (1999) believes that if needs are to be identified accurately, patients’ beliefs and values about health care must be elicited and understood. Stower (1992) takes this further by suggesting that nurses must view patients as equals and develop a partnership based upon negotiation between the nurse as an expert in health care on the one hand, and the patient as an expert on their individual needs on the other. Negotiation should include an acknowledgement of the patient’s normal routines and health care actions (Macleod Clark and Latter, 1990; Nethercott, 1993). It is only through such a process that the patient’s right to make choices can be exercised. Exercising the right to make a choice is also dependent upon being aware of and assimilating relevant information, as without knowledge choices cannot be determined (Avis, 1994). Stower (1992) and Campbell (1998) argue that historically health care professionals have tended to withhold information from patients, thus retaining power for decision making, reducing or controlling patient choice and preventing participation in care. As a consequence, patients have, on the whole, been passive recipients of care and are said to lack free will and self-determination (Stower, 1992; Avis, 1994; Campbell, 1998). Cahill (1998) suggests that involving patients in care decisions requires a radical alteration in the role of both patients and practitioners. This premise is supported by the Department of Health (1999) publication Making a Difference which includes discussion on the education and training required by nurses for the future. At a minimum, nurses aiming to develop patient-centred care approaches will require sound questioning and listening skills in order to identify needs, explanatory and educative skills to enable patient understanding, and negotiation skills to ensure that choice and thereby active participation are a reality. Active participation involves practitioners collaborating with patients in both decision making and care management, with the patient taking responsibility for specified clinical or daily living skills (Saunders, 1995). The White Papers and many of the published opinion articles tend to assume that patients want to be involved, however, the empirical literature is somewhat ambiguous on this issue. Several studies indicate that the traditional, passive role remains popular with patients who gain security from assumptions that the practitioner “knows best” (Biley, 1995; Caress, 1997). Waterworth and Luker (1990), in a small in-depth qualitative study, posited that this extends to certain patients reluctantly collaborating with practitioners who adopt patient-centred care approaches. Conversely, other studies suggest that although some patients may prefer practitioners to act as decision-makers, they also want sufficient information to enable them to participate in the process. A number of factors appear to influence whether patients are likely to want an active role. Older patients and those who are acutely ill appear to prefer less active participation than younger, better-educated patients, those who are less seriously ill and patients with greater knowledge of their conditions. There is also evidence that active involvement in care which does not require medical expertise is preferred (Beisecker, 1988; Biley, 1992; Thompson et al, 1992; Beaver et al, 1996; Caress, 1997). Other factors include the social culture (Avis, 1994) and the local health care system (Kim et al, 1993). In North America, for example, private health care has tended to encourage active participation as a means of reducing the length and costs of hospitalisation (Biley, 1992; Kim et al, 1993). In summary, it appears that patient characteristics, such as age and education, and the type of involvement envisaged, for instance clinical versus patient orientated expertise, may be relevant in considering patient-centred care approaches. In a review of the literature, Clayton (1988) concluded that active participation has the potential to enhance the appropriation of power by patients. The aim of the White Papers, and of patient-centred care, has been to alter the status quo that holds patients at the bottom of the decision making ladder. This stance has been taken up by the UKCC (1996) in their Guidelines for Professional Practice which stress the importance of patient-centred communication, and in the Code of Professional Conduct (1992) which requires the nurse to promote the interests of the individual, recognise their uniqueness: “work in an open and co-operative manner with patients..., foster their independence and recognise and respect their involvement in the planning and delivery of care (p.1). The literature suggests that nurses frequently do place patients at the centre of care. Needs are identified and goals agreed, interventions are as individualised as possible and stated philosophies emphasise holistic practice (Cahill, 1998). However, several authors maintain that the approach is often paternalistic and involves the nurse controlling the patient in covert ways, through, for example, avoiding psychosocial issues and focusing upon inconsequential conversation (Wade, 1995). Consequently, the White Papers are viewed by many authors as having promoted rhetoric among practitioners who in practice fail to deliver choice and power to patients (Malin and Teasdale, 1991; Trnobranski, 1994; Glenister, 1994; Hewison, 1995; Cahill, 1998; Campbell, 1998). If patients are to participate in decision-making, issues of power and control must be considered. Campbell (1998) suggests that the ideology behind patient-centred care is the empowerment of patients through the transfer of information and facilitation of choice. As a consequence of information gain, it is implied that those patients who want to accept responsibility for their own health care will be enabled to do so. Thus the ability of the health care professional to consider, understand and utilise strategies to empower patients to be active in making choices, and thereby be involved in decision-making and care management, appears crucial for patient-centred care. 2.4 Empowerment Over the last decade, a considerable amount has been written about empowerment as nurses have attempted to grasp its meaning and relationship to patient-centred care. Empowerment is generally regarded as an enabling activity, which aims to support individuals or groups in deciding upon and implementing their own agendas. This is brought about by recognising deficits in resources, knowledge, or activity, providing information to enhance knowledge and understanding, and enabling action to be taken according to personal beliefs, values and wishes. By implication, this action alters the balance of power and thereby enfranchises the individual or group to take control. Empowered individuals and groups are said to have the freedom to make choices and assume responsibility for their own actions (Biley, 1992; Rodwell, 1996; Campbell, 1998; Ryles, 1999). Empowerment is variously defined as: “a process by which people, organisations and communities gain mastery over their lives” (Rappoport, 1981, p.3) “a social process of recognising, promoting and enhancing people’s abilities to meet their own needs, solve their own problems and mobilise the necessary resources in order to feel in control of their own lives” (Gibson, 1991, p.359) “a process of enabling or imparting power transfer from one individual, or group, to another” (Rodwell, 1996, p.310). Thus an understanding of mastery, control, and power appear central to grasping the meaning of empowerment. It is interesting to note that only Rodwell’s (1996) definition overtly states that a power transfer must take place. Inherent in the idea of transfer is that empowerment needs to involve two parties, one of whom is initially invested with power and the authority to transfer that power. Conversely, the group to whom power is being imparted is viewed at the outset of the process as being disadvantaged in some way and powerless to grasp the advantage without the tacit approval of the more powerful group. One of the difficulties inherent within Rodwell’s (1996) definition relates to the issue of whether it is in fact possible for one person to empower another. She herself concludes that venturing to do so removes choice and that therefore empowerment must rely upon self-determination. By focusing upon the individual’s needs and giving them the authority to act, together with any necessary skills and information, they are then in a position to choose whether or not to accept empowerment and thereby act (Rodwell, 1996; Campbell, 1998). Fulton (1997) describes two approaches to empowerment, with the first being related to management theories whereby empowerment is said to be achieved through delegation. Although there is a sense that the balance of power has been shifted, control frequently remains within the hands of those who have delegated. Conversely, it is suggested that humanistic models, which value individuals, their rights and expertise, are able to produce a transfer of power (Cahill, 1998). However, Ryles (1999) maintains that proponents of Carl Rogers’ (1979, 1983) approach, which emphasises individual empowerment through increasing self-awareness, personal growth and success in interpersonal relationships, will also not necessarily recognise that a transfer of power is needed. This is due to the individual nature of the model, which advocates relationships built upon mutual trust, respect, openness and participation, which are viewed by Burrows (1997a) as characteristics of facilitation rather than empowerment. Adding in education and support, as suggested by Hawks (1992) and Rodwell (1996), does not appear to provide a convincing argument indicative of an ability to change the focus of power. Rather it suggests maintenance of the status quo, albeit from a humanistic perspective (Skelton, 1994). It has been argued that if one believes that a particular group is powerless, the only way to achieve empowerment is through an emancipatory philosophy (Friere, 1972; Foucault, 1988; Meyer, 1995a;. Ryles, 1999). In this approach, accepted patterns are challenged through increasing understanding of the relevant historical context and social structures. Developing critical insight is viewed as important in enabling groups to understand how they came to lack power and in helping them to acquire the means to free themselves from constraints. It is suggested that it is only through this activity that power can be grasped and control retained (Habermas, 1979; Harden, 1996; Fulton, 1997). The emancipatory approach is influenced by the work of Paulo Friere (1972) and the critical social theorists of the Frankfurt school (Craib, 1992) which are briefly outlined below. Friere’s (1972) approach adopts a political model which stresses the importance of people working together to increase resources and raise critical consciousness at both individual and group levels. Critical consciousness refers to the need to understand the complex political, social and economic factors, which influence and shape individuals’ lives, together with the impact of paternalism on the individual’s ability to act. Friere (1972) argues that people need to be aware of these factors in order to gain insight into areas where they are disempowered or oppressed. Awareness of oppression, or a lack of power, is thus viewed as a starting point in the process of emancipating victims of oppression towards a position where they are empowered to take control and act. Thus critical consciousness also demands collective action. Craib (1992) provides a useful analysis of the historical development of critical theory, from the early work of Adorno, Horkheimer, and Marcuse in the 1920’s and 30’s, to the later elaboration by Jurgen Habermas from the 1950’s onwards. Extrapolating the central features of critical theory, especially from the writings of its early proponents, is according to Craib (1992) extremely difficult. He criticises the work for its high level of abstraction, complexity and lack of guiding structural detail. Furthermore, he proposes that in these early works there is an assumption that people are oppressed and wish to change. Indeed Fay (1987) suggests that the broad aim of critical theory is to: “understand in a rationally responsible manner the oppressive features of a society such that this understanding stimulates its audience to transform their society and thereby liberate themselves” (p.4). Thus within Fay’s (1987) interpretation, critical theory has a clear emphasis upon oppression as well as upon understanding what is occurring in a society in order to change it. As Fay (1987) highlights, critical theorists are interested in explaining and criticising their social world, while at the same time freeing people to act. The oppressive features may include the oppression of one group over another, or the domination of a particular way of thinking. Indeed, Craib (1992) points out that the later work of critical theorists, such as Habermas, were as concerned with power imbalances and liberating people from false beliefs, as with actual oppression. Liberation is achieved through critical dialogue and raising consciousness regarding political, scientific and practical interpretations of beliefs. This process and its consequences can be viewed as a valid subject for scientific enquiry. Cahill (1998) suggests that emancipatory designs, such as critical theory, should be the methodology of choice in exploring patient-centred care, as they acknowledge individuals’ experiences and perceptions and place them within the relevant political, social and cultural contexts. As a consequence, she is critical of studies which have adopted positivist philosophies to describe and predict aspects of patient-centred care, rather than explain them (for example: Wagner et al, 1995; Spreitzer, 1996; Morrison et al, 1997). Certainly, Habermas (1979, 1988) and the other critical theorists favour emancipatory rather than positivistic or naturalistic research designs. Empirical work, in which emancipatory approaches have been adopted, indicates that the potential benefit to those involved includes raised self-awareness, understanding of constraining historical, cultural and political factors, skills in critical reflection, the development of action which challenges power balances, perceived personal power and, finally, social change resulting in alterations in the power balance and freedom from oppression (Stevens, 1989; Craib, 1992; Harden, 1996; Kincheloe and McLaren, 1998; Kuokkanen and Leino-Kilpi, 2000). In addition, negative outcomes, such as tension and conflict, are also recorded. Given that human nature tends to avoid tension and conflict it may not be surprising that empowerment, as a process, does not yet appear to have been popularly ascribed to by those working in the NHS (Cahill, 1998; Ryles, 1999). In summary, the patient/practitioner relationship in patient-centred care is characterised by empowerment (Campbell, 1998). Empowerment can only be achieved through understanding the factors that underpin oppression, followed by action to alter the power balance. In health care, the practical goal of empowerment is not only to promote patient choice and participation in care, but also to effect specific changes in perceptions, activity, and outcomes related to health. In management approaches this is achieved predominantly through delegation (Fulton, 1997), in humanistic, but individualised, models, through education (Hawks, 1992; Rodwell, 1996), and in emancipatory models through encouraging critical reflection and discussion regarding both the issue at focus and the surrounding political, social and structural contexts (Friere, 1972; Habermas, 1988). Further information on the philosophical and structural detail of critical theory and emancipatory research is presented in chapter three. 2.5 Patient Empowerment The notion of patient empowerment requires that health care professionals offer choices to patients based upon their individual needs, that they actively encourage patients to participate in decision making and that they use their skills and knowledge to work in partnerships with patients. Chadderton (1995) suggests that as the consumer approach to health care has gained favour, patients have become motivated to collaborate in their care management. She adds that this motivation is further enhanced by reactions against NHS technology, which leave the patient alienated. Alienation is increased by health care practitioners’ narrow definitions of health, focused upon technological outcomes that may not reflect patients’ needs and lifestyles. Chadderton (1995) concludes that patients are empowering themselves through an increasing understanding of health and health care, recognition of the limitations of the NHS and raised awareness of the power of patients as consumers. Her scenario reflects Gibson’s (1991) definition of empowerment and Roger’s (1979, 1983) individualised model. Conversely, Cahill (1998), writing some three years later, argues that if the NHS is to adopt patient-centred approaches the public needs to be re-educated regarding their roles and responsibilities. She bases this view upon an opinion that after decades of knowledge and power resting with practitioners there are few real choices for most patients, empowerment is paid lip service at best and society, as a whole, remains disempowered in terms of health care. Re-educating patients regarding their roles and responsibilities in the new NHS could result in rigid application of patient-centred care. If government or health care professionals determine the approach without reference to patients, the outcome may be further oppression of those who prefer to be passive recipients of care. Indeed, Elliott and Turrell (1996) warn that advocates of empowerment must accept that true empowerment may mean that the patient exercises the right not to be involved. They add that the patient-centred care literature variously describes the patient’s role as advisor, collaborator, equal partner or pre-eminent team member. As each role is dependent upon the style of decision making and the scale of power transferred, the role adopted by the nurse and the willingness and motivation of the patient to actively participate in care will influence the degree of empowerment achieved. The type of role adopted and the degree of power transfer aimed for in this study is addressed further in later chapters. In a study by Caress (1997) involving 405 patients undergoing renal dialysis, the majority of patients viewed their roles as passive despite stating that they preferred active and collaborative roles. Where patients deliberately choose passive roles, their rationale included a diminished desire for participation when the illness was serious (n = 48), trust in and deference to the practitioner (n = 85) and a lack of knowledge (n = 56). The National Health Service Management Executive (NHSME) (1993) stress that information giving is the most important factor in empowering patients. Not only would this increase patients’ knowledge, but it may also reduce the development of inappropriate deference. In practical terms, however, the decision on how much information to give, or indeed seek, lies in the hands of the practitioner. Consequently, patients may be enabled to be involved or they may be controlled in overt or covert ways. In individualised care, nurses are encouraged to seek the patient’s perspective. Elliott and Turrell (1996) believe that by focusing on individuals’ needs and wishes, nurses should gain sufficient insight into the amount of information required by patients to determine whether they wish to participate in their own care. Once this judgement has been made, those patients who prefer active involvement should be supplied with the information needed to empower them to make decisions and actively participate in care management. A variety of positive findings are highlighted in studies that have attempted to describe or test patient outcomes of empowerment. In a survey of 117 general practice patients, Brody et al (1989) found those who perceived that they were involved in decision-making reported fewer concerns regarding their illness, less discomfort and dysfunction, and a greater sense of control. They concluded that involving patients in their care improved outcomes and enabled power transfer. Other studies in chronic care have also demonstrated increased patient satisfaction and improved functional status as a consequence of empowering individuals to be involved in decision-making (Mahler and Kulik, 1990; Coulter, 1997). Empowering patients through the provision of information has been found to influence rates of service utilisation. When men were made aware of the true benefits and disadvantages of investigation and treatment, Wolf et al (1996) noted a decrease in prostate-specific antigen screening, while Wagner et al (1995) found that surgery rates for mild to moderate benign prostatic hyperplasia fell. Despite these findings there is relatively little research which specifically addresses patient outcomes in relation to empowerment. 2.6 Nurse Empowerment Consumerism and patient-centred care challenges the technical and professional supremacy of health care practitioners (Chadderton, 1995). If approaches are used which not only encourage active participation, but also aim to empower the patient, nurses may feel threatened by the loss of their own power. Wade (1995) warns that this is likely to increase stress and further decrease motivation to adopt new approaches to care. In addition, tension and conflict may arise as a consequence of patients asserting their right to make decisions which may differ from those the practitioner would have made. In a study by Kahn and Steeves (1988), the caring relationship between nurse and patient was negatively affected as a result of patients acting in a decisive, autonomous manner. A second study found that the most common reaction to increased patient participation was hostility on the part of the nurses (Campbell, 1990). It should be noted that both studies pre-date much of the emphasis on patient-centred care and it is possible that the findings would be different today. However, May (1995) believes that nurses are frequently reluctant to involve patients if it means they will have to relinquish their own power and control. Nurses are commonly regarded as an oppressed group subjected to institutional structures developed by an elite (managers and doctors) to reduce conflict and maintain power in their organisations (Foucault, 1988; Chavasse, 1992; Street, 1992; Fulton 1997). It has been argued that the control and exploitation of nurses by medical staff in particular, has led nurses in their turn to initiate structures that impose control over patients. On the one hand, exercising control may reduce conflict, a potentially important consideration in busy, high stress environments, on the other hand, it may result in exploitation of patients and the denial of freedom and choice (Harden, 1996; Fulton, 1997). Friere (1972) argues that oppressed groups commonly internalise the values and norms of the dominant group. Two main reasons are suggested for this process: firstly, dominant values and norms come to be regarded as the “right” ones in society; secondly, the subordinate group comes to believe that if they behave like the oppressor they will gain power and control. In health care, doctors are said to oppress nurses, who in turn dominate patients. As a result, patients are encouraged to be passive recipients and are placed firmly at the bottom of the power ladder. Those who attempt to react against nurse domination, by for instance demonstrating their health care knowledge, are frequently regarded as difficult and unpopular patients (Ryles, 1999). It is suggested that nurses need to be empowered before they can hope to empower patients and involve them in their own care (Rodwell, 1996; Fulton, 1997; Cahill, 1998; Campbell, 1998). If nurses are to actively engage patients in their own care, they need to be empowered to move from directional forms of care management towards consultative and collaborative approaches which place the patient at the centre. This requires a consciousness shift, that goes beyond providing individualised care, to an approach which proactively involves and empowers patients to make decisions about their health care (Cahill, 1998; Entwistle et al, 1998). However, as Cahill (1998) and Kuokkanen and Leino-Kilpi (2000) point out, further empirical research is required to assess whether such a change in approach is either possible, or of benefit, to the health care practitioner. Indeed, if nurses enable patient empowerment while at the same time remaining oppressed by medical staff, the consequences could be a break down in any multidisciplinary team working. The skills and qualities required by clinicians to adopt empowering approaches to patient involvement in decision making and care management, receives little attention in the literature. In a paper on empowerment in nurse education, Chally (1992) suggests that those aiming to empower others require a positive self-concept, creativity in terms of generating ideas and offering alternative choices, appropriate resources which include access to up-to-date knowledge, information in relevant technical, clinical and practical areas and an ability to offer support. Harden (1996) expands Chally’s (1992) list of qualities by adding commitment, caring and interactive skills capable of promoting critical dialogue. Several other authors support the need for critical dialogue, maintaining that it is only through raising critical consciousness that empowerment can be enabled in disenfranchised individuals (Harden, 1996; Fulton, 1997; Ryles, 1999). In a study which set out to empower nurses to work within a patient-centred framework, Pill et al (1999) trained staff to use negotiation to enhance diabetic patients’ decision making regarding day-to-day choices. However, despite the training and long-term support, most nurses were unable to alter their consultation style. The researchers commented that practitioners felt threatened by the idea of patient choice and involvement in decision making, as they believed that the patient was at risk of making the “wrong” choice and thus compromising their health. The assumption is that the nurses felt responsible for ensuring that patients acted in the “correct” and conventional biomedical way, rather than providing options which took account of the individual’s lifestyle. This was particularly true where the nurse felt that blood glucose levels were not well controlled. Several subjects did become more critically reflective and demonstrate a willingness to involve patients in their own care. However, the authors concluded that even these nurses failed to internalise the philosophy of patient-centred nursing. Pill et al’s (1999) study involved 18 part-time practice nurses in 15 practices. Training was undertaken through a minimum of two, three-hour workshops, which focused upon how to enable patients to voice their concerns, select topics for discussion and set goals. Support lasted for three years and took the form of quarterly newsletters, two group meetings and visits from the research nurse. Data were predominantly collated through field notes written up by the research nurse after each visit. A social scientist also interviewed the nurses approximately six times during the study to discuss the method used with patients, encourage compliance and, if necessary, provide extra teaching. The failure of the study to empower nurses may be due to the methodological approach. There was no indication that nurses were encouraged to critically reflect upon the status quo, challenge traditional values and beliefs, raise their collective consciousness about the philosophy of patient-centred care, or collaborate in the research as full participants, rather than subjects. These appear to be necessary factors in studies concerned with empowerment and their inclusion might enable nurses to deal with any oppression. Two options appear to be available for studies in empowerment. Firstly, to take a broad view of the oppressed group as being subject to wide political, economic and social constraints which impact upon the totality of individuals’ roles. Secondly, to adopt a more focused approach where the emphasis is upon a particular aspect of role experience and function (Kincheloe and McLaren, 1998). This latter perspective enables groups to focus upon false beliefs and power imbalances in specified areas and is consistent with Habermas’ (1979) approach and that of the post-modern critical theorists. Further discussion is offered in chapter three, however, it is postulated that focusing upon the potential for engaging patients in their postoperative pain management, through the employment of self-generated strategies, may enable consciousness raising and critical reflection in acute pain management. 2.7 Patient-Centred Care and Empowerment in Acute Pain Management The implications of a patient-centred approach to acute pain management may be distinguished through a consideration of the characteristics of patient-centred care. In meeting patients individualised biopsychosocial and spiritual needs, a multidimensional perspective of pain needs to be adopted. Thus the impact of pain in the physiological, sensory, affective, cognitive, behavioural and sociocultural dimensions needs to be assessed. Inevitably, this will require identification of the patient’s pain coping strategies and their values and beliefs regarding acute pain management. Sofaer (1998) advocates the adoption of McCaffery’s (1968) definition of pain, as: “whatever the experiencing person says it is, existing when the experiencing person says it does” (p.95). This classical definition may be somewhat limited by its lack of overt acknowledgement of the factors involved in pain perception, but it is well known and captures the subjective nature of pain appropriate to patient-centred care. It is thus viewed as a relevant definition for a study that aims to engage patients in their own pain management. Operationalising McCaffery’s (1968) definition should require the nurse to negotiate pain management interventions with the patient and to identify the required professional and personal resources to support effective pain relief. Once appraised of the relevant information, those patients who wish to can participate in decision making about their pain management. Traditionally, the power for decision making in postoperative pain management has been invested in health care practitioners. The dominant pattern has been that of the medical model focused upon pharmacological management. Paternalistic control is evident through the conventional activity in which medical staff prescribe the analgesics of their choice, while nurses administer according to their own decisions regarding which of the prescribed medications to use and at what dose. The patient frequently accepts the analgesic without knowing or questioning what the drug, its effects or side effects are. As a consequence, the ability of the patient to actively participate in decision making is negated by the withholding of information. Timed drug rounds and delays in analgesic administration following patient requests, further compound the situation. Perceptions that control lies in the domain of practitioners, rather than patients, are thus enhanced. As such, patients are unable to exercise the right to make choices and are powerless to become involved in this aspect of their care. It is possible that in the first few days following surgery, patients, and particularly those in older age groups, would prefer doctors and nurses to act as decision-makers in relation to postoperative pain management. The expertise required to understand and make informed decisions regarding pharmacological management may also deter those who prefer to confine their involvement to areas where no medical understanding is required. However, the success of patient controlled analgesia (PCA), which aims to enhance control through the delivery of a controlled dose of intravenous morphine in response to patient demand, may limit this argument. Once PCAs are discontinued, control for medication tends to return to the hands of health care professionals. Although focusing interventions upon pharmacological management results in an unidimensional approach and control being retained by practitioners, it is not to say that analgesics are not useful. However, the focus on analgesics may intimidate patients, force them to conform to a “nurse knows best” attitude and prevent them from exercising their preferred pain coping mechanisms. The lack of active patient participation in acute pain settings is evident in Copp’s (1974) study. She found that many patients prefer non-pharmacological self-generated strategies to medication, but are reluctant to use them in hospital fearing the disapproval of health care staff. Her findings indicate that patients had neither the power nor choice to feel enabled to participate in their own care. It is possible that the ensuing loss of autonomy may lead to frustration and anxiety, which in turn may enhance pain perception (Walding, 1991; Titchen, 1996). To address this situation, health care professionals could combine pharmacological management with techniques which aim to identify and understand the patient’s preferred strategies. In this way multidimensional, therapeutic pain management may be enhanced and individuals may be empowered to participate in their care (Fulford, 1996; Kitson, 1996; Campbell, 1998; Carter, 1998). Patient-centred approaches are recognised as being essential for effective chronic pain management where active participation is encouraged, information shared and collaboration emphasised (Turner and Keefe, 1999). However, this is not the case in acute pain management. As previously mentioned, knowledge is required if choice is to be enabled. In a small exploratory study, Avis (1994) found that patients allow choices to be made for them if they lack knowledge. Although in Copp’s (1974) study, patients used self-generated strategies in attempts to relieve acute pain at home, their lack of knowledge regarding the impact of non-pharmacological strategies on pharmacological management, appeared to disempower them from exercising their preferred strategies in hospital. The information required by the patient wishing to employ self-generated strategies may be as simple as discussing how the strategies may be utilised in a ward setting, and reassuring patients that their use will not interfere with pharmacological management. The issue of whether the use of self-generated strategies is beneficial in postoperative pain management remains open to question. Although the evidence for the effectiveness of taught non-pharmacological strategies is also open to question, it is interesting to note that those practitioners seeking interventions other than analgesics for postoperative pain management, have focused upon taught rather than self-generated pain relieving strategies. Piper and Brown (1998) believe that nurses are concerned with achieving technological and measurable outcomes, which emphasise the rationality of human behaviour and are linked to medical model perspectives. Teaching pain relieving strategies may emphasise patient powerlessness through de-valuing the patient’s own pain coping attempts. Conversely, an approach which emphasises a humanistic, emancipatory philosophy, aimed at raising consciousness for action, control and mastery, may empower patients to use their own self-generated strategies. As mentioned in chapter one, Walding (1991) hypothesised that there may be a causal link between pain, anxiety and powerlessness such that perceptions of powerlessness result in increased pain. She suggests that decreasing perceived powerlessness may enhance personal control, promote pain coping and reduce pain perception. If a patient believes that a particular strategy is useful in relieving pain, they have a right to discuss their preferred strategies and make informed choices with a full understanding of the implications. This right extends across environments, from home to hospital. However, as the hospital environment is frequently alien, nurses may need to enable patients to employ their own strategies. If nurses do not offer this facility, it could be argued that they remove power, control, choice and decision-making from the patient, thus disempowering and indeed potentially oppressing the individual. It is postulated that patients are experts in relation to self-generated pain relieving strategies and should be actively encouraged to participate in their pain management. It would appear that to incorporate patients’ own strategies into the plan of care, consciousness raising is necessary with both patients and nurses. The ensuing enlightenment regarding the political, scientific and practical implications would aim to liberate patients and nurses from their existing values and beliefs and the institutional regimes associated with practitioner led pain management interventions. Through forming an alliance with patients and developing critical consciousness, it is suggested that nurses approach acute postoperative pain management from a multidimensional perspective which is creative, enables patient-centred care and offers choices to both patients and nurses. As Jenkins (1997) argues, the reality of patient-centred care can be strengthened if nurses enable patients to become involved in self-help initiatives - such as the use of self-generated pain relieving strategies. The issue of whether it is actually possible to empower patients and nurses in just one focused aspect of care is considered in chapter seven. 2.8 Conclusion Despite the widespread interest and commitment to patient-centred care amongst opinion leaders, many health care professionals have not been trained or educated to adopt the role change required in practice. The literature is frequently vague on how patient-centred care may be operationalised, focusing upon analysing the pertinent concepts rather than offering empirical studies or practical ideas regarding the way forward. The credibility of patient-centred care may be intertwined with the degree to which nurses and patients are able to express their values and beliefs through a process of empowerment. It is argued that empowering staff to employ patient-centred care in one specific area of practice, could be beneficial in helping nurses to understand and internalise the philosophy overall. Focusing upon the patients’ every-day experiences of, and routines to relieve pain, may offer an aspect of clinical practice in which nurses can engage patients postoperatively. Nurses may also benefit by learning from and with the patient so that both parties are freed from the traditional values and beliefs that currently constrain postoperative pain management. The following chapter considers how the adoption of a critical theory approach could be used in an attempt to empower nurses to reflect critically upon their actions and adopt more patient-centred approaches to postoperative pain management. CHAPTER THREE: METHODOLOGY: CRITICAL THEORY AND ACTION RESEARCH 3.1 Introduction Chapter one suggested that patients should be engaged in their postoperative pain management through the use of self-generated acute pain relieving strategies. In chapter two, it was argued that to achieve this, nurses need to be empowered to collaborate with patients. This chapter discusses the research design for the proposed study. The philosophical approaches to research and knowledge development are briefly and critically examined, followed by the suggestion that practice based disciplines, such as nursing, require a different approach. Having introduced the idea of action research as a potentially valuable design to investigate practice based problems, a more in-depth consideration of critical theory is offered than was possible in chapter two. It is proposed that action research, when combined with critical theory, is an appropriate design to facilitate nurse/patient collaboration and empowerment. The characteristics of action research are discussed and the chapter concludes by critically examining some of the issues that need to be considered when undertaking an action research study. 3.2 Positivism and Naturalism: The Traditional Research Paradigms An understanding of the philosophical and theoretical approaches to research and knowledge development are fundamental to an appreciation of the potential of action research as a valid, scientific and respectable research design. The philosophical approaches, or paradigms (Kuhn, 1962), historically viewed as providing a basis for research, are the positivist, postpositivist and naturalistic perspectives. The advantages and limitations of these paradigms are discussed below. Positivists believe that knowledge exists independently from individuals and their social and historical contexts. The world can therefore be viewed objectively and knowledge can be discovered through observation and measurement in an unbiased and carefully controlled manner. As parts of the total picture are identified, through verifying a priori hypotheses, theorists can begin to predict what will happen as a consequence of what is already known. Because context is not an issue, theories developed in one situation are viewed as being generalisable to similar situations in other contexts (DePoy and Gitlin, 1994; Roe, 1994; Oldham, 1994; Poole and Jones, 1996). Examples of studies on pain that arise from a positivistic stance include research on the physiology, biochemistry and pharmacology of pain and its management. The emphasis upon strict objectivity and careful control within positivism has come to be regarded as overly constraining in examining human phenomena (Poole and Jones, 1996; Forbes et al, 1999). In striving for control, positivists neglect the influence of subjects’ and researchers’ perceptions, values and beliefs upon the research process (Behi and Nolan, 1996; Poole and Jones, 1996). Consequently, these influences may be missed and their potential to affect the results ignored (Playle, 1995). As Poole and Jones (1996) highlight, theories derived from this perspective have the ability to describe and predict, but they cannot explain. In an attempt to respond to these criticisms, the postpositivist movement acknowledges the need for investigations in less strictly controlled human environments and recognises that theory cannot be neutral. Instead it is located within the sensory perceptions and cognitive processes of the researcher and participants (Poole and Jones, 1996). Postpositivists view objectivity as: “a social result of mutual criticism among a community of inquirers” (Poole and Jones, 1996, p.111). This position differs considerably from the black and white positivist view of linear cause and effect observations. Postpositivists recognise the complex nature of causality and aim to advance understanding by attempting to grasp the different dimensions of reality from a relatively broad perspective. The interest of postpositivists is, therefore, one of corroboration rather than theory verification (Forbes et al, 1999). Studies are carried out in natural settings and may involve a variety of methods. Although it may not always be possible to capture reality completely, rigorous data collection, combined with recognition of social and historical contexts, may enable a near representation of reality (Carter, 1998; Guba and Lincoln, 1998; Poole and Jones, 1996). Consequently, postpositivist studies may be more able to capture the realties of clinical pain and its management. Research designs belonging to the postpositivistic paradigm include experiments, quasi-experiments and many descriptive and exploratory studies. Of these, experimental designs are the strongest examples, seeking to corroborate theory by establishing levels of cause and effect. Through exposing one group to a specific treatment (the experimental group), and then comparing the effects with a second group who do not receive the intervention (the control group), the effect of treatment can be assessed (Polit and Hungler, 1999). The findings can provide evidence to explain and predict the effectiveness and efficiency of particular treatments, thereby indicating best practices (Oldham, 1994; Closs and Cheater, 1996). Forbes et al (1999) suggest that, in comparison to positivists, postpositivists adopt a more sceptical approach to generalisation based upon their acknowledgement of human perception. Findings are thus said to be generalisable from one situation to another only if the contextual characteristics are similar (Poole and Jones, 1996). Critics of both positivism and postpositivism argue that given the complexity of human living, knowledge development should not only be grounded within political, social and cultural contexts, but also be investigated from the perspective of the individual (Field and Morse, 1985; Simmons, 1995). As Simmons (1995) asserts: “facts and values cannot be separated” (p.839). From the naturalistic perspective, knowledge is viewed as being grounded within individuals’ experiences and perceptions. Thus individuals create their own subjective view of the world which is interdependent with others’ encounters and understandings. Naturalism is based upon the belief that the complexity of human experience can only be understood if it is examined holistically and within the social setting. Knowledge is subjective and interactive and cannot be directly transferred from one setting to another. The research designs linked to the naturalistic paradigm include grounded theory, ethnography, phenomenology and hermeneutics (DePoy and Gitlin, 1994; Simmons, 1995; Koch, 1996; Ashworth, 1997). Examples from the pain management literature include research into patients’ experiences of living with pain and perceptions of the impact of pain management interventions. According to its proponents, the strengths of naturalistic enquiry lie in its flexibility, recognition of the interaction between researcher and subjects and ability to view phenomena within ever changing, context specific situations (Simmons, 1995). However, it is also these characteristics which leave naturalistic research most open to censure as critics suggest that the approach lacks the traditional rigour of positivist and postpositivist science (Forbes et al, 1999). Supporters of naturalistic research, on the other hand, argue that there is more than one approach to assessing rigour and that, in its own way, naturalistic enquiry can be as rigorous as positivist research (Guba and Lincoln, 1981; Sandelowski, 1986, 1990; Ashworth, 1997). In today’s cost conscious and evidence based world the inability of naturalistic research to produce generalisable findings, in the positivist sense, is, however, viewed by some as a major weakness (Poole and Jones, 1996). (Further discussion on rigour may be found in section 3.7 of this chapter). It is precisely because of its ability to yield generalisable results that research based upon positivism and postpositivism has, in the past, been viewed by many as the only way to develop an understanding of the world (Carr and Kemmis, 1986). Certainly, in health care research these paradigms have dominated as a result of their appeal to medicine. Yet, the interactive elements of naturalism may be more pertinent to understanding human experience. However, naturalism is not directly concerned with practice-based problems and development. Thus a naturalistic study of patients self-generated strategies might focus upon identifying the strategies that patients use, gaining insight into how they are used and what this means to the individual, rather than determining the efficacy of strategy use. Conversely, a postpositivist approach may test the use of one strategy against another, confining the patient to agreed strategies in a carefully controlled manner. Examples of such studies were offered in chapter two. Greenwood (1984) argues that, as a practice based discipline, nursing requires something more than the objectivity of postpositivism and the subjectivity of naturalism. She is also critical of the tendency for researchers from both paradigms to leave the field of inquiry without implementing their findings. The contemporary emphasis upon evidence-based practice highlights the historical problem of transferring research findings into practice and the prevailing commitment to increasing clinical effectiveness through transforming practice in an evidence-based manner. Evidence-based practice is defined by Sackett et al (1996) as: “the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients” (p.71) Muir Gray (1997) adds to this definition when he suggests that practitioners share the best available evidence with patients and, through a process of consultation, decide which option is most suited to the patient. This interpretation is notable for combining the tenets of evidence-based practice with the beliefs of patient-centred care. Evidence may be found in critically appraised or collated research, such as the systematic reviews produced through the work of the Cochrane collaboration, in primary research publications, or in national and local guidelines. If evidence is lacking, primary research may be undertaken to begin generating evidence or consensus opinion may be sought (le May, 1999). The purpose of evidence-based practice is to enhance clinical effectiveness, which the NHSE (1996) defines as: “the extent to which clinical interventions when deployed... for a particular patient or population do what they are intended to do, that is, to maintain and improve health and secure the greatest possible health gain from the available resources” (p.341). Clinical effectiveness thus aims to enhance the quality and consistency of care by basing practice on evidence and ensuring that it is cost-effective. The implications are that practitioners should not only adopt a healthy scepticism towards un-researched traditional practices, but also be thoughtful of resource allocation. With their emphasis upon control, neither positivism nor postpositivism are likely to encourage research designs that focus upon both knowledge generation and practice transformation. Equally, the emphasis upon description in naturalistic studies generally means that the setting is not altered during the research phase. On the whole, positivists and naturalists disseminate their findings for others to implement. Carter (1998) argues that if practice is to be evidence based, research designs need to be adopted which enable practice transformation. One example of such a design is action research, which deliberately sets out to change practice and practitioners during the course of the study. The following section begins to consider the utility of action research in changing practice while generating knowledge and commences with a discussion of the philosophical background. 3.3 Action Research: A Design for Practice Based Disciplines It was as a result of his concerns with the objective and context removed positivist paradigm that Lewin (1946) first began to develop his ideas regarding action research. To be meaningful, Lewin (1946) believed that research needed to be conducted from the individual’s perspective and within their social setting. Furthermore, he argued that if research findings were to be viewed as useful in day-to-day practice, the research itself needed to address issues that were of direct concern to practitioners. However, Lewin’s ideas were themselves influenced by the dominant positivist paradigm and reflect a tendency to apply experimental methods to social situations and action (Hart and Bond, 1995b). Indeed, he viewed action research as a “change experiment” (Lewin, 1946). As a result, his approach has at times been criticised as a form of manipulative social engineering although most authors do comment upon Lewin’s aim to develop a democratic research design through his employment of critical reflection (Carr and Kemmis, 1986; Hart and Bond, 1995b). Lewin died before he completed his work, however, other authors have since elaborated upon his ideas (Adelman, 1993). In their seminal paper, Susman and Evered (1978) suggest that action research: “can base its legitimacy as science in philosophical traditions that are different from those which legitimate positivist science” (p.582). This implies that those who use action research will need to be able to argue that against the “gold standard” of positivism, action research has legitimacy. It is interesting to reflect upon why a research approach that has the potential to directly influence practice, should be viewed as less worthy than studies that may take years to filtrate through to clinical settings due to delays in publishing, practitioners’ perceptions of irrelevancy and problems with applying results in the real world (Webb, 1989; Newton, 1995). It is also interesting to question why nurses, who may be regarded as an oppressed group, should deliberately choose to adopt approaches, like action research, which require defending to ethics committees, medical staff and the research world at large (Playle, 1995; Fulton, 1997). The potential of action research to effect change in practice, and to enable nurses to understand the power bases within the work environment, may go some way to explaining nursing’s interest in this research design (Webb, 1990b; Hart and Bond, 1995a). 3.3.1 In Defence Of Action Research Philosophically, the defence for action research arises through an approach termed “new paradigm research”, which several authors contend is a new and emerging view of knowledge development (Reason and Rowan, 1981; Meyer, 1995a). New paradigm research is defined by Reason and Rowan (1981) as a: “collaborative, experiential, reflexive and action-orientated process” (p. xx). The approach differs from orthodox research by perceiving participants as co-researchers rather than subjects, thus valuing their inherent ability to participate as partners through-out the research process (Heron, 1981). Sometimes referred to as “co-operative enquiry” (Heron, 1996), new paradigm studies are described as being: “with and for people, rather than on people” (Reason, 1988, p.1). The paradigm is typified by its emphasis upon collaboration and close linkage of research and action. The focus is upon both practical issues and those which are of concern to practitioners. Researchers adopting this perspective seek to bring together social action and personal development through a process of education, critical reflection and empowerment (Reason and Rowan, 1981). It is within the area of new paradigm research that many authors would place action research and other essentially collaborative designs, such as illuminative evaluation, participatory and heuristic research (Reason and Rowan, 1981; Sanford, 1981; Meyer, 1995a). According to some writers, however, collaborative research has some way to go in its theoretical and philosophical development before it attains the status of a new paradigm (Basford and Slein, 1995). Indeed, certain authors view action research as currently belonging within the naturalistic approach due to its tendency to adopt essentially qualitative methods (DePoy and Gitlin, 1994). This perspective suggests that thinking around collaborative research may be constrained by a belief in the inherent rightness of the current knowledge and understanding of the naturalistic paradigm and collaborative research approaches (Kuhn, 1962; Dunbar, 1995, Carter, 1998). Others argue that modern day action research is linked to a new paradigm built upon humanistic philosophy and critical theory (Reason, 1988; Meyer, 1993). The next section considers critical theory in some depth. 3.4 Critical Theory: What Is It and What Does It Contribute To Action Research? Critical theory is concerned with the meanings that individuals assign to events, within the contexts in which these events occur. It thus denies the objectivity of positivism and postpositivism, believing that they only address surface issues and are unable to generate knowledge to explain the meaning of human action (Harvey, 1990; Titchen, 1993). Critical theorists also reject the reductionist view of these paradigms, where the focus of investigation is on parts rather than the whole. In its place, they believe in the concept of totality, which refers to the inter-relatedness of issues and a view that social phenomena should be investigated within historical, social, political and economic contexts (Harvey, 1990; Guba and Lincoln, 1998). These contexts are seen as both surrounding and influencing the participants and the problem/s being studied (Freire, 1972; Habermas, 1974). Critical theorists are particularly concerned with approaches which are able to transform the practice context through freeing people from their present constraints, liberating them to act and enabling them to resolve practice problems. They argue that positivism, postpositivism and naturalism are unable to alter the status quo and they thus reject these paradigms (Carr and Kemmis, 1986; Carter, 1998). The central features of critical theory are viewed as encompassing: emancipation reflexivity praxis ideology and theory development. Each of these is discussed in turn. 3.4.1 Emancipation Emancipation relates to the notion of freeing oneself from the constraints and control that arise from unequal power relationships. Based on the idea that social structures and actions are maintained through institutional, political and economic power, emancipatory activity suggests that social reality can be reconstructed. The assumption is that transference of ownership from dominant to less powerful groups is both possible and desirable. More recently, critical theorists have also proposed that emancipatory action can help to free individuals from dependency on habit and tradition (Carr and Kemmis, 1986; Peters and Robinson, 1984; Carson, 1990; Harvey, 1990). In addressing these issues, Habermas (1974) asserts that rational individuals can be enabled to act positively, politically and freely in society, if they engage in critical reflexivity. Thus, in terms of the proposed study reflection upon the constraints, power balances and traditions in postoperative pain management would be required if practitioners are to be freed to alter their practice through a critical theory stance. 3.4.2 Reflexivity Engaging in critical reflexivity enables participants to become aware of the conditions which “oppress” them, such as where and in whom power is invested, or where their beliefs and actions may have become distorted. The presumption is that all human beings have the potential to engage in reflective activity and, through this, to participate in identifying problems and generating emancipatory knowledge (Habermas, 1972; Susman and Evered, 1978). Reflexivity involves the promotion of self-reflection through critical discourse, or discussions. Discourse is regarded as a dialogue, that is, a two way process of communication involving the exchange of ideas and opinions (Freire, 1972). The process aims to be critical, based upon mutual respect and to raise participants’ conscious awareness of the ideas and opinions held both within and outside the group. Like naturalists, critical theorists value subjectivity and believe that the influence of values and beliefs need to be acknowledged. However, contrary to naturalists who offer interpretations based upon their own understanding of participants’ perspectives, critical theorists encourage participants to expose their values and beliefs through critical discourse. The discussion is directed at exploring different perspectives, exposing contradictions, imbalances and frustrations that exist within the situation, considering possible ways of altering beliefs and conditions and, through liberating participants from the habits and routines which paralyse their practice, providing the impetus for action aimed at improving practice (Habermas, 1974; Carson, 1990; Kim, 1993; Bellman, 1996). The purpose of critical discourse is, thus, to transform practice through a dialectical process involving systematic reasoning around opposing ideas (Craib, 1992; Carter, 1998). Reflexivity is also said to help individuals embark upon a process of self-study that aims to provide insight into the position of self within the specific situation and to enable continual personal growth and learning (Torbert, 1981; Schon, 1991). The ensuing revelations form the basis for individual action and knowledge generation (Reason and Rowan, 1981; Harvey, 1990; Schon, 1991). Thus, each participant is engaged with the researcher in a form of co-operative inquiry. To facilitate this process, researchers need to develop relationships with participants so that all parties achieve an understanding of the other’s ideas, together with the conflicts, tensions and power bases which face them within the setting (Winter, 1989; Craib, 1992). One of the issues with critical theory is that participants may regard researchers who adopt reflexive approaches as controlling, rather than emancipatory. To reduce this potential problem, Titchen (1993b) advocates that the process of critical discourse should be carefully introduced. A second danger is the possibility of creating vulnerability amongst participants who are being asked to challenge their underlying assumptions, values and beliefs. Consequently, the personal and interpersonal skills of the researcher need to be regarded as equally important to the methodological stance (Meyer, 1993). The work carried out by the author into the components of facilitation were viewed as being helpful in informing the approach to be adopted in the proposed study (Burrows, 1997a). Finally, the status of the researcher in relation to the participants is significant in determining whether individuals are truly empowered to act as co-researchers within the dialectical process. The issues of participant vulnerability, researcher/participant relationships and researcher status are returned to later in this chapter and in chapters four and seven. The following section addresses some of the elements surrounding the transformation of practice as a consequence of reflexivity. 3.4.3 Praxis In critical theory, practice transformation is encompassed within the concept of praxis. Originally referred to by Aristotle, praxis relates to the: “art of acting upon the conditions one faces in order to change them” (Susman and Evered, 1978, p.594). Thus, while reflexivity engages participants in critical discourse, praxis encourages action based upon the reflections and a view that theory should inform practice. Knowledge is therefore not just about finding out about the world, the position taken by positivist, postpositivist and naturalistic scientists, but also about changing it. The focus of investigation should thus be upon a critical analysis of practical problems that have relevance for practitioners (Carr and Kemmis, 1986; Harvey, 1990). Critical theories are practical in the sense that they stimulate participants to alter their lives through the development of critical self-knowledge and an understanding of their social situations. By acting to reform the situation, personal change and development may also take place. In terms of praxis, critical theory is a form of social action, where the aim of research is transformative and research and action are “inextricably intertwined in practice” (Torbert, 1981, p.145). Research based on praxis aims to uncover the main contradictions in a situation and to reconstruct participants’ ideology to enable emancipation and informed action (Susman and Evered, 1978; Reason and Rowan, 1981; Carr and Kemmis, 1986; Fay, 1987; Harvey, 1990). 3.4.4 Ideology Reconstructing participants’ ideology assumes that previous ideologies are, in some way, “false”. In this context, the term “false” does not suggest that the ideology is wrong, but rather that it is shown to no longer be necessary for the functioning of society and thus no longer acceptable to those who hold it. The ideologies held by society are organised around the values, beliefs and norms of a culture and its institutions. These institutions are said to both facilitate economic production and repress desires (Craib, 1992). According to critical theorists, repression leads to distortions in communication and interaction, which in turn affects behaviour. Individuals must become aware of their ideology and rid themselves of illusions if they are to be freed from social repression (Geuss, 1981). As ideologies reflect individuals’ subjective and personal opinions, values and beliefs, they cannot be directly measured. Investigators have to rely upon exposing them through critical reflection. The supposition is that raising consciousness about false ideologies will lead to frustration, which in turn will promote a discussion on alternative perspectives. Exposure to new ways of thinking will lead to enlightenment, action to reform ideology and the participant’s situation, and change that is universally acceptable. Ideology is therefore regarded as both informing, and arising from, social action (Geuss, 1981; Torbert, 1981; Carr and Kemmis, 1986; Fay, 1992; Craib, 1992). Geuss (1981) provides an in-depth examination of what constitutes a false ideology in critical theory. In summary, he suggests that delusions occur in one of three senses: Firstly, the generic sense is concerned with the origin of, and motivation for, ideologies. Delusions only arise in this sense if individuals are unable to acknowledge their motivations. It could be conjectured that relying on analgesics, to the exclusion of other approaches, is at odds with the current view of pain as a multidimensional phenomenon and the philosophy of individualised nursing care which seeks to address patients biopsychosocial needs. Maintaining the dominance of analgesic use may suit practitioners who unconsciously wish to retain power over patients. Secondly, in the epistemic sense, beliefs are supported by empirical evidence or self-fulfilling prophecies. These become delusions if they are shown to be untrue. In pain management, systematic reviews indicate that certain analgesics may be more limited than previously thought in dealing effectively with postoperative pain (McQuay and Moore, 1998). Yet, they remain the sole relief measure offered by most practitioners on the basis of empirical findings from single RCTs and individuals’ perceptions that they are effective for postoperative pain relief. Finally, functional ideologies are beliefs that justify social practice. It is within this area that the domination of one group, or set of ideas, over another, is placed. Historically, the medical model, which was underpinned by positivism, has tended to provide the ideological guide to postoperative pharmacological pain management. Although analgesics are generally an appropriate means of relieving postoperative pain, patients frequently comment that they “do not like taking painkillers/injections/tablets”. As was discussed in chapter two, pharmacological management may disempower patients if practitioners, without reference to the patients’ needs, wishes and pain management goals, retain control. While effective analgesic management is to be welcomed, a less unidimensional approach to the multidimensional problem of pain may enhance therapeutic management. Carter (1998) suggests that the person experiencing pain is now regarded as having an important role to play in both pain management and pain research. Pain is no longer viewed as being purely physical, there is an emphasis upon the individual and an acknowledgement that management needs to be approached holistically. This relatively recent ideological stance encourages questions about patient-centred pain management interventions. In the proposed study, the question posed relates to whether identifying patients pain relieving strategies and incorporating them into the plan of care, reduces postoperative pain, facilitates patient empowerment and promotes a more critical and multidimensional approach to postoperative pain management, or whether this constitutes a delusion and thus false belief? 3.4.5 Knowledge Development Developing knowledge from the critical theory perspective requires a consideration of the inter-connectedness of at least two of the three senses (generic, epistemic and functional) discussed in the previous section. In addition, an account of the roles of local history and the current context is essential for the generation of knowledge that aims to assist members of an organisation to solve their practical problems (Susman and Evered, 1978). As the product of a critiquing process, embarked upon by collaborating individuals to expose contradictions in social action, critical theory generates emancipatory knowledge (Carr and Kemmis, 1986). Emancipatory knowledge is viewed by Habermas (1972) as one of three approaches to human knowing. The other two are technical and practical knowledge. Each is briefly summarised below in order to locate emancipatory knowledge within context. Technical knowledge refers to the human interest in acquiring knowledge to facilitate technical control over natural objects. The outcome is instrumental, in as much as it offers causal, or scientific, explanations. Research related to this approach is empirical in origin, related to the analytical or natural sciences and underpinned by positivism and postpositivism. Practical knowledge is generated to enhance understanding through language. Knowledge is interpretative and informs and guides the judgements made during communication and dialogue. The research basis is that related to hermeneutic, or interpretative sciences, and is based upon the naturalistic paradigm. Emancipatory knowledge considers human interest in autonomy and freedom and is motivated by people’s interests and needs. Knowledge is developed through reflection and applied, through a consideration of power, to communication and social action. Research is informed by the critical sciences and underpinned by critical theory. According to Titchen (1993b), emancipatory knowledge provides a practical explanation for the way in which the structures, relations and conditions that restrict progress, can be transformed to enable practice development. No one way of knowing can suit a phenomenon as complex as pain. Indeed, each paradigm has the potential to add to our understanding of pain and its management. While technical knowledge may offer causal explanations regarding the pathophysiology of pain, disease associations and pharmacological management, practical knowledge may assist understanding of the way pain is perceived, communicated and expressed. Sociocultural influences may also be understood through this route. Conversely, emancipatory knowledge may enhance reflection upon political agendas, partnerships and power balances and the social actions of pain management partnerships. The proposed study is focused within the latter paradigm, without denying the relevance of the previous two. 3.4.6 Critical Theory and Action Research The philosophical underpinnings and central characteristics of critical theory may be closely linked to the reflective, emancipatory and collaborative elements of action research. Indeed Susman and Evered (1978) suggest that: “the immediacy of critical social issues form(s) an essential ingredient of action research” (p.587). Thus within some forms of action research, the features highlighted above are apparent. For example, Winter (1993) considers that action research is clearly underpinned by critical reflection, while Waterman (1995) addresses the emancipatory element of action research. She believes that action research is about developing people’s knowledge and skills to free them from the constraints of routine, lack of power and lack of knowledge. In this way she suggests that action research seeks to develop knowledge as a means to social improvement. Further discussion on the place of critical theory in action research is offered through the remaining sections of this chapter. The section above has attempted to locate action research within the current view of the philosophical foundations of research, by highlighting the central tenets of positivism, postpositivism, naturalism and new paradigm approaches. The nature and key features of critical theory have been described, together with an initial consideration of their relevance for action research. The chapter now moves on to look at the characteristics of action research. 3.5 Action Research: Characteristics and Definitions The characteristics identified by Holter and Schwartz-Barcott (1993), as being central to an understanding of action research, are used as a framework for this section. Explanations about each characteristic are interspersed with the interpretations and comments of other authors to highlight some of the differences and similarities in the thinking of those writing about action research. The four characteristics are: identifying practical problems specific to a particular situation creating change in practice collaboration and developing theory (Holter and Schwartz-Barcott, 1993). Each is addressed in turn below. 3.5.1 Identifying Situation Specific, Practical Problems Susman and Evered (1978) suggest that as research methods and designs have become increasingly sophisticated, they have also become less helpful in solving the practical problems faced by practitioners in their day-to-day work. Instead, research has focused upon generating knowledge which practitioners are expected to interpret, and then apply to their own situations. It is this issue which has not only created a perception amongst practitioners that research findings and theory lack relevance for day-to-day practice, but has also been the catalyst for the evidence based practice movement (Greenwood, 1984; Peters and Robinson, 1984; Sackett et al, 1996; Carter, 1998). As previously mentioned, it was partly as a consequence of his concerns regarding the distance of research from practice, that Lewin (1946) first developed action research. In his paper on the use of action research to investigate intergroup relationships, Lewin (1946) stressed the importance of focusing upon practical problems to ensure research utility. He also discussed the need for investigators and participants to consider problems within their specific situational context, thus acknowledging the importance of the people involved, their particular skills and the specific constraints that those individuals have to deal with. Researchers adopting action research designs therefore need to be empathetic to the context and the people within it, to appreciate problems from the participants’ perspective, to gain insight into the attitudes and goals of participants and to view context specific constraints from within (Meyer et al, 2000). Finally they need to be committed to focusing upon the practical or social concerns of those involved. In addressing the issues of practical, situation specific problems, Lewin (1946) suggested that social scientists should adopt action research, which he described as a series of planning, execution and fact-finding stages. He describes the planning stage as starting with a general idea which is examined in the light of the present situation. From this, an overall plan emerges and decisions are made regarding the initial action required. The action is then “executed” and is followed by a stage of fact-finding to evaluate the action, learn from the action, and plan the next step. Consequently, the researcher enters a second cycle of planning, executing and fact-finding. As participants move through this continuous spiral, the original plan may be modified in light of critical reflection upon the issues. In more recent literature, several authors have offered their own interpretations of the cyclical nature of action research. For example, Susman and Evered (1978) described a five stage process which involves diagnosing the problem, action planning, action taking, evaluating, and reporting the findings through the generation of critical theory. Conversely, Cohen and Manion (1985) propose a flexible eight stage process, comprising: problem identification; discussion and negotiation with interested parties and production of a draft proposal; literature review; refinement of the problem and research questions; selection of research procedures; selection of continuous evaluation procedures; implementation of research and action; and finally, interpretation of data, overall project evaluation, identification of recommendations and dissemination. In reality, although acknowledgement is made of an action stage, Cohen and Manion’s (1985) interpretation appears to mimic a generalised systematic research process rather than being specific to action research. As such, it seems to ignore the importance of reflection that many perceive as an essential component in action research (Carr and Kemmis, 1986; Winter, 1989; Elliott, 1991). In contrast, the four-stage process of planning, acting, observing and reflecting, described by Carr and Kemmis (1986), overtly addresses the need for critical reflection within action research. In nursing, Meyer (1995b) adopted a flexible three-dimensional spiral of action-reflection cycles (based upon the model described by McNiff, 1988) to summarise her study retrospectively. Her rationale for choosing this approach, rather than a more prescriptive model, related to the fact that she wished to adopt a facilitative role which would enable issues to be addressed as they emerged and yet allow her to communicate her study to others. She describes six stages of negotiation, assessment, planning, action, evaluation and withdrawal, each of which incorporated its own range of data collection and analysis. East and Robinson (1994) contend that approaches, which are consistent with the nursing process, may hold greater appeal for nurse researchers than the more prescriptive, predominantly educational models historically available. However, the nursing process can itself be prescriptive, and while holistic in nature does not incorporate critical reflection as a central characteristic. Indeed, some would argue that it has technical and positivist tendencies (Fish, 1991; Playle, 1995) which would thus put it in conflict with a critical theory approach to action research. The variety of stages suggested by the authors above can be confusing. However, as Meyer (1995b) points out the uniqueness of each study, arising from the collaborative nature of action research, demands flexibility in relation to the process chosen. A medley of schemes from which to choose may therefore be helpful to those planning and undertaking action research studies. McNiff (1988) warns, however, that it may be counter productive to adopt any one particular strategy. She suggests that those described in the literature may be constraining in the real world of practice and research and suggests that practitioners and researchers formulate their own model as they proceed with the study. While McNiff’s (1988) argument may be of use to experienced action researchers, Benner’s (1984) point about novices requiring frameworks and guidelines should be borne in mind. How for example can the brand new action researcher negotiate access, time and resources to carry out a study, if they have no idea or guidelines on how to approach the first stage? Furthermore, time and resources are precious commodities. Does it not therefore behove the action researcher to value participants sufficiently to at least have some idea of the stages to be developed in collaboration with them? A final point, is that the researcher may feel they require an idea of where they are going for their own peace of mind and confidence and in order to seek informed consent from potential participants. This latter issue is considered further in section 3.8. It could be argued that structures and frameworks are philosophically more akin to positivistic, rather than emancipatory, approaches. However, a number of authors, who have adopted critical theory or reflexive perspectives, advocate the use of frameworks to guide novice and more experienced participants (for example: Boud et al, 1985; Burnard, 1991; Schon, 1991; Johns, 1994). As highlighted above, action researchers have themselves developed models and guidelines and have entered into an open dialogue regarding methodology. As with any study, the totality of the design and the context in which the study is to take place need to be considered. Decisions regarding the use of models can then be made. In contexts where participants are novice researchers it is the author’s contention that a certain amount of structure is needed. This is not to advocate that action researchers plan their studies in detail, as that would counteract the collaborative, flexible nature of action research and ignore the reality that the stages tend to blur together in the real world of practice. However, it may be useful to identify a particular strategy to help communicate ideas to participants so that everyone can work together on the detail. This is especially important if participants are to be viewed as co-researchers and be involved in developing practice through critical discourse. In view of these arguments, it was considered useful to suggest to the participants in the proposed study, that a design incorporating Carr and Kemmis’ (1986) four stage process of planning, acting, observing and reflecting, should be adopted for the first cycle of the research. The details of this design are further explained in chapter four. 3.5.2 Creating Change In Practice A key element of action research is that there is a recognisable change in practice linked to the identified problem. In Lewin’s (1946) terminology, the action researcher carries out a change-experiment, the ultimate aims of which are to improve practice via developmental, structural, environmental or organisational change, and, through this, to contribute to the generation of practice theory. Action researchers believe that they have a moral responsibility to transfer research into action that is beneficial for those affected (O’Hanlon, 1994). The research outcomes are thus viewed as being bigger than the researcher and important for practice. The difficulty arises in both promoting and retaining the developments and changes. As Meyer et al (2000) highlight, problems can occur as a result of participants reluctance to change, lack of time, motivation and resources, lack of support and leadership, poor multidisciplinary communication and teamwork, and the often routine nature of nursing practice. Furthermore, the proposed change may threaten participants’ values, beliefs and current working styles, thus leaving them feeling criticised and vulnerable (Meyer, 1995a). As such, action researchers need to be motivated individuals, committed to practice development, skilled in change management and sensitive to participants’ feelings and the potential ethical dilemmas of combining research and action. The ethics of action research are discussed further in section 3.8. Lewin (1946) warned that changes achieved through action research might lack permanency. The degree of permanency attained may be linked to the particular approach adopted. For instance, action research studies based upon emancipatory designs are thought to effect change more permanently than those based upon technical or organisational approaches (Holter and Schwartz-Barcott, 1993; Hart and Bond, 1995b). Section 3.6 discusses the different approaches in more depth, however, suffice it to say at this stage that the more emancipatory the design, the more likely it is that change will be invested in the participants. If this is the case, participants should have the skills and qualities to influence practice following job changes. However, the original environment of change may “lose” the practice development unless it has also become internalised within the organisation. It could be further argued that if organisational transformation does takes place, practitioners of the future may come to regard the development as oppressive and constraining if they do not understand and agree with the underlying motivations and rationale. From a critical theory perspective, action can be seen as the development of self-knowledge and rejection of false ideologies. Change occurs as a result of empowering participants to become consciously aware of their beliefs and opinions, the conflicts and tensions surrounding their social context, and the power bases within which they operate. In emancipatory action research empowerment enables participants to alter their ideologies through critical discourse, to see where they have the authority to act, and to prepare, encourage and enable them to do so. As Susman and Evered (1978) point out, one of the goals of action research is: “to develop the self-help competencies of people facing problems” (p.588), Carr and Kemmis (1986) argue that people are more likely to act if their actions are underpinned by knowledge developed either through education or critical discourse. Thus the credibility of action research to change practice may be intertwined with the degree to which participants are able to express their values and beliefs through a process of empowerment. The next section examines the way in which participants may be involved in action research studies. 3.5.3 Collaboration The third characteristic of action research, identified by Holter and Schwartz-Barcott (1993), is collaboration. This may be viewed as working with others on, for example, a joint project (Collins, 1999). In action research the participants in the collaborative endeavour are the researcher(s) and participants, with the latter being seen as individuals who know: “the field or workplace from the inside” (Holter and Schwartz-Barcott, 1993, p.300). Participants are viewed as the experts in the field, with the researcher bringing knowledge of relevant theory, research design and processes and change management. The collaborative process is dependent upon mutual sharing and respect for all participants’ relevant knowledge and expertise. Each learns from the other and collaborates in a learning process, as well as in the research and practice transformation. In this way, action research is participatory, as group members are involved either directly, or indirectly, in initiating the research questions, design and methods, in analysing and interpreting the data and in changing practice. The involvement of participants as collaborators in action research was first addressed by Collier (1945), who commented upon the responsibility of participants in the action elements of the study, that is: in identifying the problem, discussing solutions and implementing change. Philosophically, action researchers believe in the innate ability of human beings to collaborate in these areas as well as to be involved in the generation of practice knowledge (Susman and Evered, 1978). While collaboration is viewed as an essential element of action research, the degree of collaboration will vary according to local need, the needs of the research and the particular action research approach adopted. Chein et al (1948) identified four levels of collaboration: In diagnostic action research, the researcher collects data to diagnose the problem and feeds the results back to the participants; in empirical action research, the researcher evaluates the actions implemented and feeds the results back. Conversely, in participant action research, the researcher works with the participants to identify the problem and plan the action. Finally, experimental action research involves the researcher and participants working together on all aspects by setting up an experiment for taking action and evaluating the outcomes. The first level appears to have closer links with audit than with action research. However as Closs and Cheater (1996) point out, while audit may draw on theory it does not aim to generate or test theory. The second level appears to reflect a postpositivist paradigm, while the fourth level has clear links with Lewin’s (1946) account of action research as a change experiment. The descriptions given by Chein et al (1948) illustrate early interpretations of action research and are perhaps less flexible in their view of collaboration than modern day action research. Titchen’s (1994) paper on collaboration provides a possible framework for those adopting action research within nursing. She describes three different levels of collaboration: collaborative informants, collaborative groups and co-action researchers. The co-action researchers are involved in the research and action as equal associates. However, depending upon the expertise and roles of participants at this level, each may take a lead in specific and appropriate aspects of the study. At the collaborative group level, participants are involved in some of the research activity, for example, in discussions on which data collection instruments to use, collecting data and possibly analysing the data. Members also participate in planning and co-ordinating the action and in reflecting upon the findings. In contrast, collaborative informants view themselves as ”subjects” of the research, providing data but not actually engaging in the research activity. However they should participate fully in planning and implementing the action if change and practice development are to occur. Titchen’s (1994) framework is useful in considering the structure of action research studies within the complex and often hierarchical settings of nursing practice. By allowing for different levels of participation, resources can be used in an effective way without excessive disruption to care delivery. This is particularly helpful where the study is across a number of settings and may indeed be more realistic than involving all potential participants in the research process. Furthermore, the model enables an “outside” researcher to work as a co-researcher with an “insider” change agent. Titchen and Binnie (1993a) themselves worked in this way and suggested that the model enables innovation and change combined with theory generation and testing. It should be noted that the “researcher” and “practitioner” do not have to be different people. Depending upon the size of the study, problem to be investigated and the knowledge base of participants, the researcher may be a practitioner. Whoever the researcher is, however, the guiding principal is that they collaborate with practitioners in identifying the problem, developing an action plan, implementing the action and critically evaluating the outcomes. A potential difficulty in adopting Titchen’s (1994) framework is that change may be undermined if some of the actors in the setting are excluded from a critical discourse aimed at problem identification and consciousness raising. If the ideologies of the group as a whole are not addressed, cohesive planning for action may be unworkable. As a result, the potential to develop an emancipatory knowledge for action may be limited. Previous experience with Titchen’s framework has led the author to view this model as invaluable in the practical management of action research studies in acute care settings (Burrows and Whybourne, 1996). It was therefore proposed that the model should be adopted for this study with the caveat that thought needed to be given on how collaborative informants could be involved to enable a critical theory approach to be adopted. This issue is addressed in further detail in chapter four. 3.5.4 Developing Theory The final characteristic, identified within Holter and Schwartz-Barcott’s (1993) framework, is the development of theory. Developing existing theory or generating new theory is an inherent goal of research. In action research the theory generated is grounded in action, hence Torbert (1981) refers to action theories which are developed and refined by identifying potential courses of action in response to the perceived problem. Theory generation is also grounded in practice, thus Cohen and Manion’s (1985) suggest that action research increases “functional knowledge” (p.209). In this way, action research may be used to test the limits to which existing knowledge may be applied to specific practice situations, thus refuting or supporting the original theory. As Holter and Schwartz-Barcott (1993) state: “If the ultimate purpose in developing nursing knowledge is to improve nursing practice, then knowledge that is validated and revised through practical application is extremely important for knowledge in nursing” (p.303). Given its situation specific nature, action research does not aim to develop generalisable knowledge, that is, knowledge that can be automatically applied in more than one setting. Rather, the aim is to produce knowledge for a specific situation and purpose (Cohen and Manion, 1985). Devotees of action research would argue that this increases the utility of the knowledge generated, as the research findings are applied to the practice situation instead of gathering dust on the investigator’s shelf. However, critics would counter this argument by pointing to the limited ability of the developed theory to apply to any other context than the one in which the changes were made. The issue of transferability, to which this discussion refers, is addressed in more detail in section 3.7. In developing theory, Holter and Schwartz-Barcott (1993) suggest that action research is not constrained to the use of any particular data collection method. Indeed a variety of methods are mentioned in the literature, including observation and interviews, reflective diaries and discussions, critical discourse, detailed field notes, questionnaires, rating scales and quasi-experiments (for example: Cohen and Manion, 1985; Webb, 1989, 1990; Waterman, 1995; Meyer, 1995a). The variety of methods used, appear to span those traditionally associated with one or other specific paradigm. It could be argued that this is because collaborative inquiry has yet to attain the status of a true paradigm, able to lay claim to its own methodological stance. Conversely, it may reflect the practice-focused perspective of new paradigm research, whereby the nature of the practice problem directs the choice of method to be adopted. Equally, the spiral character of action research studies may well demand the use of more than one method during the course of the study. Methods employed in action research studies need to inform both the action elements and the research questions. They therefore need to uncover the meanings that practitioners ascribe to their actions and beliefs (Lewin, 1948; Susman and Evered, 1978). Thus, for example, questionnaires containing open questions would be preferred to those relying upon closed questions. The methods adopted in the proposed study are outlined in chapter four together with a discussion of their appropriateness in action research. The following section critically analyses the utility of Holter and Schwartz-Barcott’s (1993) framework in relation to definitions of action research. 3.5.5 Action Research Defined Having considered the four characteristics identified by Holter and Schwartz-Barcott (1993), namely, identifying practical problems specific to a particular situation; creating change in practice; collaboration; and developing theory, it is useful to assess the extent to which they reflect existing definitions of action research. Rapoport (1970) defines action research as follows: “Action research aims to contribute both to the practical concerns of people in an immediate problematic situation and to the goals of social science by joint collaboration within a mutually acceptable ethical framework” (p.499). In this definition, Rapoport (1970) highlights the interest of action researchers in addressing practical problems occurring within specific situations, proposes that a contribution is made to creating change, emphasises the importance of collaboration, and through the statement “the goals of social science”, proposes that theory is developed. However, as Kendall and Sturt (1996) point out, this statement assumes that the goals of social science are both fixed and understood. Given the flexible nature of action research, it is clear that the research goals may in fact alter as the study progresses. The focus upon mutuality reflects the emphasis within co-operative inquiry that research should be beneficial for all involved and that praxis should be based upon universally acceptable ideas. On the whole, Holter and Schwartz-Barcott’s (1993) framework is supported in Rapoport’s (1970) definition. Conversely, the definition by Halsey (1972) is more limited. He states that action research can be viewed as a: “small-scale intervention in the functioning of the real world and a close examination of the effects of such intervention”. Here the focus is very much upon the action elements combined with an evaluation of the changes made. Although this definition aids understanding of the totality of action research, it is rather restricted in its description of the processes and characteristics involved. There is no reference to collaboration, which is clearly viewed as a defining feature of action research (Reason and Rowan, 1981). Furthermore, the choice of the words “small-scale”, may lead sceptics to perceive action research as unimportant and limited. In their definition, Carr and Kemmis (1986) highlight the reflective element of modern day action research, together with the characteristics of changing practice in specific settings. They state that action research is: “self reflective enquiry undertaken by participants in social situations in order to improve the rationality and justices of their own practices, their understandings of those practices and the situations in which those practices are carried out” (p.1067). Thus the focus of action research is stated here as being practical problems and social improvement. The involvement of participants in collaborative inquiry and in theory generation is also apparent. This definition not only incorporates the characteristics of action research discussed earlier, but reflects a critical theory philosophy involving reflexivity and justice in the social world. The definitions offered above each underpin, in their own way, the characteristics identified by Holter and Schwartz-Barcott (1993). However, they also introduce concepts, such as reflective inquiry, which are not addressed within the framework. These concepts are incorporated into Holter and Schwartz-Barcott’s (1993) discussion of the different approaches towards action research which are outlined in the following section. 3.6 Action Research Approaches Holter and Schwartz-Barcott (1993) describe three approaches to undertaking action research, namely the technical collaborative, mutual collaborative and enhancement approaches. To some extent these reflect Habermas’ (1972) views on the nature of human knowledge, as technical, practical and emancipatory, which were discussed earlier. The approaches are now considered in turn, together with examples taken from the nursing literature. 3.6.1 The Technical Collaborative Approach In the technical collaborative model the researcher aims to validate and refine existing theory by testing its application to specific practical settings. In this way, the approach is scientific and based upon technical action. The problem is defined in advance by the researcher who seeks out practitioners interested in the study and who are prepared to work with the researcher to implement a previously specified change. Collaboration is thus also viewed as technical. Change is likely to occur quickly and efficiently, but may be short lived as participants tend not to internalise the altered actions. The danger is therefore that once the researcher leaves the setting, organisational constraints and previous practices may return. Jones (1996) describes an action research study that set out to test whether the role of emergency nurse practitioners (ENPs) could be extended to assessing children in a paediatric casualty department. The need for this role development was linked to the reduction in junior doctor hours, together with an increasing number of patients. The relevant theories appear to be those related to skill attainment, role delineation and role conflict. Jones (1996) offers a detailed account of the setting in terms of the client group, but does not address the nature of the staff involved or her relationship to them. She does however refer to herself as a Sister at the hospital and cites literature that suggests that she was aware of possible compromises arising from unequal power relationships between her and the participants. The methods employed were predominantly quantitative, although a field diary was maintained to enable interpretation of the learning that occurred during the change period. The findings provided information on the types of children who could be successfully assessed and treated by ENPs and suggested that staff had the skills required to undertake this role. Jones (1996) discussed and reflected upon the findings in collaboration with the staff. Although not explicitly stated, it appears that Jones (1996) identified the problem and relevant theory in advance of seeking the consent of nursing and medical staff to embark upon a specified change process. While changes in practice occurred, she comments upon the constraints of an inappropriate skill mix to sustain the change on a 24-hour basis. Nevertheless, she also highlights how, through reflective discussions, other issues began to be raised. It appears then that although this study started from a technical collaborative stance, it may have moved towards a mutual collaborative approach. 3.6.2 The Mutual Collaborative Approach The mutual collaborative model relies upon the researcher and practitioners working together to identify the problem and possible solutions. Participants thus develop a shared understanding of the issues and can be said to have generated “local” descriptive theory on the topic. Because of the depth of collaboration, change is said to be lasting. However it is also invested in those involved and thus tends to dissipate as participants leave. An example of this approach can be seen in the work of Webb (1989) who commenced her research by working alongside nurses on the study ward for one shift per week over a three-month period. During this time she felt that she became an accepted member of the team and gained insight, through participant observation, interviews and questionnaires, into the context, relationships and problems that existed within the setting. The main areas of concern for the participants related to interpersonal relationships and team building. A mutual agreement was reached to address these issues before moving on to introduce practice developments connected to organisational change. Webb’s (1989) work clearly highlights the mutual respect that developed between the researcher and participants. Through the initial data collection, staff were able to develop a shared appreciation of the problems at hand and, in collaboration with Webb (1993), to identify possible solutions. However, it could be argued that the approach is more akin to developing practical, rather than emancipatory, knowledge for action. 3.6.3 The Enhancement Approach Researchers adopting the enhancement approach aim to engage practitioners in a process of critical reflection. The purpose of this is to enable participants to recognise and acknowledge the value systems, norms and conflicts that underpin the organisational culture and individuals work practices. The outcome is to develop a shared understanding, or “collective consciousness” (Holter and Schwartz-Barcott, 1993, p.302), of the underlying causes of a problem and constraints to change. As a result, changes that do occur tend to be internalised into the organisations work practices and are thus longer lasting. Holter and Schwartz-Barcott (1993) suggest that in enhancement approaches: “the emerging patterns of new practice and new theoretical insights stem from the newly formed culture of practice” (p.302). Consequently, this approach is regarded as more able to generate emancipatory knowledge. In a study on the use of the Roper, Logan and Tierney (1990) model of nursing, Bellman (1996) adopted an enhancement approach to enable nurses to reflect upon their use of the model and consider implementing practice change. Using theoretical concepts from both critical and change theories, an initial data collection phase aimed to identify nurses’ perceptions of the model. The findings, which revealed both a need and wish to develop psychological factors, were presented to staff and the action research process with its involvement of nurses as co-researchers explained. All the staff consented to participate. Regular taped meetings were held to facilitate group reflection, action planning and evaluation. Three concurrent changes relating to self-medication, patient controlled analgesia and operation specific information leaflets were identified, implemented and evaluated. In her paper, Bellman (1996) details the changes made and the manner in which staff took the lead in developing practice. Examples are given of the critical dialogue that took place during meetings. It is evident that the value systems, norms and conflicts that formed the groups’ ideology were brought to the level of collective awareness. Reflexivity is also apparent. However, in relation to emancipation, the participants continued to perceive themselves as lacking the power and authority to develop collaborative activities with the medical staff. Consequently, they tended to avoid involving the doctors - a practice with which Bellman (1996) admits she at times colluded. Of some concern, is Bellman’s (1996) self-criticism for not taking on responsibility for liaison with the medical staff herself. Given her use of critical theory, it would appear inappropriate for her to take such action. Rather she should perhaps have found an opportunity to address both her own and co-researchers actions and feelings in more depth so as to promote empowerment and emancipation. A useful aspect of Bellman’s (1996) paper is that she highlights areas in which both participants and patients became empowered. The examples offered above appear to suggest that the characteristics of action research can be variously interpreted according to the approach adopted. Furthermore, a search of the literature reveals that different authors conceive different typologies for action research. The following section aims to examine some of these typologies and to consider their relevance for the proposed study. 3.6.4 Alternative Typologies The three ways of knowing, identified by Habermas (1972), are apparent in Holter and Schwartz-Barcott’s (1993) interpretation of action research approaches. Thus the technical collaborative approach is based upon the natural sciences and linked to aspects of postpositivism, while the mutual collaborative approach adopts hermeneutics and embraces a naturalistic perspective to generating practical knowledge. Conversely, the enhancement approach embraces emancipatory philosophy and is related to critical theory. Thus it is only in the latter approach that one might expect to see the features of emancipation, reflexivity, praxis and ideology overtly conjoined. Carr and Kemmis (1986) use the labels technical, practical and emancipatory to describe their typology of action research designs. Unlike Holter and Schwartz-Barcott (1993), however, Carr and Kemmis (1986) assert that all action research is underpinned by critical theory and therefore involves reflexive activity. Their technical stance encompasses a reflective analysis on practice, the practical approach includes self-reflection upon the ideas and rationale for action and, finally, emancipatory action research incorporates critical discourse. Holter and Schwartz-Barcott (1993) suggest that their approaches are mutually exclusive. Although Carr and Kemmis (1986) are not quite so constraining, it is not until one considers the typology of Hart and Bond (1995b) that there is an acknowledgement that approaches may overlap. Hart and Bond’s (1995b) presentation is of four ideal types which they make clear are unlikely to be discreetly recognised in practice. Indeed, they suggest that any single action research study may move from one approach to another through the course of the project. Incorporating experimental, organisational, professionalising and empowering models, Hart and Bond’s (1995a; 1995b) typology is set against a continuum ranging from a consensus view of society to a conflict view. Each approach is further differentiated from the others in terms of its educative base, problem-focus and nature of improvement and involvement. Thus for example, the experimental approach is placed at the consensus end of the continuum on the premise that research aims to produce a generalisable consensual definition of improvement. Features encompassed within this perspective include re-education, administrative control and social change, the identification of causal links and the emergence of problems from the interaction of social theory and social practice. Conversely, the empowering approach is set at the conflict pole of the continuum on the basis that understanding is multi-faceted and research produces pluralist definitions of improvement. The emphasis is upon consciousness raising, shifts in the balance of power, negotiated definitions of problems and competing definitions of success. Holter and Schwartz-Barcott (1993) believe that action researchers must, from the outset, be clear about which approach they are adopting. This belief is based upon an opinion that as each approach is philosophically divergent, the researcher will need to consider widely differing issues. Carr and Kemmis (1986) also have some empathy for this view although they are not quite so restrictive. Conversely, Hart and Bond (1995b) propose that the purpose of identifying an approach is to provide an overall sense of direction for a study. The perspective here tends to be one of pragmatism in which the approach adopted is used to guide the researcher in considering different interests and potential problems. It could be argued that the dynamic nature of social settings, combined with the flexibility of action research, means that it is not always possible to state in advance which approach will be used, or whether a single approach will in fact suffice. Hart and Bond’s (1995b) argument that the approach is likely to alter through the course of the study is persuasive in this context. Hart and Bond’s (1995a; 1995b) typology appears dynamic and of practical help in raising the issues to be addressed by action researchers. However, it is limited by its neglect in overtly identifying the philosophical and theoretical underpinnings to each approach. In an analysis of the writings of eleven authors, who have attempted to refine and develop action research as a methodology, Peters and Robinson (1984) concluded that it is those who emphasise the critical and emancipatory features who present the strongest case for action research as a science. This point is well made if one believes that technical and practical action research contain features of positivism and naturalism, while emancipatory action research, through a focus on participation and critical discourse, belongs to a separate paradigm based on collaborative inquiry. The literature on action research in educational settings highlights how Lewin’s (1946) rather technical approach has largely been replaced by one more aligned to new paradigm research and critical theory. In the nursing literature, a similar movement is becoming apparent. It may be that a clear understanding of the philosophical and theoretical underpinnings is more helpful to those undertaking action research than an emphasis upon typology. This may be particularly apposite if Hart and Bond’s (1995b) point about the likelihood of overlap within the approaches and movement between them is accepted. The proposed study aims to engage patients in patient-centred postoperative pain management, to empower nurses to deliver effective pain management and to promote practice transformation. Collaborative inquiry and a critical emancipatory approach towards action research philosophically support these aims. However, the research idea was the author’s and access was negotiated. Although known to the Ward Sister and the Nurse Specialist for the Acute Pain Service, the author was, at the time, very much an outsider. Shortly after access negotiations were concluded, the position of Link Lecturer for the ward became vacant. In fulfilling this role it was possible to advance the development of a mutual relationship more quickly than might otherwise have been the case. As a result, the author was able to engage participants in collaborative critical discussions relatively quickly, to challenge their ideologies regarding postoperative pain management and to encourage them to participate actively in developing the research design. The intention to adopt a critical theory stance was thus possible. As will become apparent from chapter four onwards, both technical and emancipatory action and theory were generated. The idea of regarding typologies flexibly therefore appears important. Whichever approach, philosophy, or theory is adopted, action researchers need to consider the rigour of their study if the findings are to have any credibility. The following section addresses this issue, returning to some of the themes that have been touched upon earlier in the chapter. 3.7 Action Research: The Issue of Rigour “That (action research) should be lacking in scientific rigour ... is not surprising since the very factors that make it distinctively what it is ... are the antithesis of true experimental research.” (Cohen and Manion, 1985, p.216). Science may be defined as the systematic and rigorous collection of data to generate knowledge about human behaviour, experiences and environments (Depoy and Gitlin, 1994). Yet, Cohen and Manion’s (1985) statement suggests that there is only one way to judge scientific rigour, that is from an experimental viewpoint which focuses upon issues of reliability and validity. Given the philosophical basis of action research, Susman and Evered (1978) emphasise the inappropriateness of following such an argument, suggesting instead that the scientific merits of action research should be judged from alternative perspectives. The papers by Guba and Lincoln (1981) and Sandelowski (1986, 1990) are informative in this area. Like Susman and Evered (1978) they believe that it is inappropriate to judge all research against the criteria used by positivist and postpositivist researchers. Rather, they suggest that reliability and validity should be addressed by considering scientific rigour through an examination of truth-value, applicability, consistency, and neutrality (Guba and Lincoln, 1981). Each is discussed in turn below. 3.7.1 Truth-Value: Internal Validity versus Credibility In postpositivist research, with its emphasis on control, truth-value relates to internal validity. The latter refers to whether the study measures what it was intended to measure. Conversely, in naturalistic research truth-value considers whether findings faithfully interpret participants’ experiences and comments. This has been termed credibility (Guba and Lincoln, 1981; Sandelowski, 1986). Credibility may be addressed by asking participants to comment upon analysed data to establish whether they recognise their own and other participants’ views and experiences. However, Sandelowski (1990) argues in a later paper, that such “member checks” may not ensure credibility as participants views can change over time and may be altered as an effect of taking part in the research. In critical emancipatory designs, one might expect views to alter as a result of research participation. It seems that member checks may be a useful strategy for action researchers who must however remain aware that as participants’ views alter through the reflexive process, the level of credibility achieved should be regarded with caution. Indeed, Kincheloe and McLaren (1998) warn that it is extremely difficult to measure truth-value in dynamic research designs, such as those based upon critical theory. One possible strategy would be for action researchers to maintain their own reflective field notes highlighting any transformation of views as they occur (Koch, 1996). 3.7.2 Applicability: External Validity Versus Fittingness Applicability is associated with external validity, which in positivist and postpositivist designs is concerned with subject selection and the generalisability of findings. In action research the aim is to produce knowledge that is specific to the problem and setting. Samples are therefore generally small, being confined, for instance, to one particular ward, health care centre, or group of practitioners. As the theory generated by action research needs to be carefully re-examined and amended in each new research setting (Susman and Evered, 1978) findings cannot be viewed as representative in the positivist or postpositivist senses (Sandelowski, 1986). However, participants are seen as representing the groups to which they belong and the experiences of those groups. In place of generalisability, Guba and Lincoln (1981) therefore recommend considering “fittingness” as a means of assessing applicability. This relates to whether the findings “fit” contexts other than that in which the study was carried out. Sandelowski (1986) and Appleton (1995) believe that fittingness may be ascribed when practitioners in other settings view the findings as being relevant to their own experiences. In Greenwood’s (1984) view, assessing whether findings fit ones own personal reality is a far more useful criterion against which to judge research outcomes than the objective stance taken by positivists and postpositivists. However, given the need to critically examine the applicability of findings in each new setting, action research could be criticised in relation to its resource and cost implications. 3.7.3 Consistency: Reliability Versus Auditability The third area considered by Guba and Lincoln (1981) is consistency. In positivist terms, this refers to reliability and is based on the premise that precise replication of data collection procedures is possible and desirable (Sandelowski, 1986). Conversely, action researchers are interested in the human experience within its unique setting and surrounded by unique conflicts and tensions. They are thus generally not concerned with exact replication. To address consistency, Guba and Lincoln (1981) suggest researchers provide sufficient information to enable others to follow the thinking and decisions made. In-depth descriptions of the setting, context and participants are required, together with a critical discussion of the researcher’s involvement and influence on the study and the way in which the researcher and participants made decisions. Maintaining field notes and reflective accounts enables action researchers to undergo and encourage personal self-inquiry (Reason and Rowan, 1981). In publishing these accounts others can then follow the “decision trail” and judge the consistency of the study. This process is termed auditability. Consistency may also be demonstrated if a second researcher, reviewing the data and decision trail, arrives at the same or similar conclusions (Sandelowski, 1986). Titchen (1995) suggests that action researchers use “participant validation” to assist in this process. In this approach the participants involved in the study help the researcher with data analysis and interpretation, checking both the procedures followed and the interpretations made. As Titchen (1995) states: “we had to convince each other that our ideas were justified by, and did not distort, the participants’ accounts” (p.44). The way in which this was achieved was through critical discourse. Both the researchers and participants reflected upon the data agreeing and disagreeing until a reliable interpretation was reached. Reason and Rowan (1981) support such a validation process in their examination of rigour in new paradigm research. It also reflects the collaborative nature of both new paradigm and action research. 3.7.4 Neutrality: Bias Versus Rigour and Reflection The final area to be considered in relation to scientific rigour is neutrality. This is defined by Sandelowski, (1986) as being related to a lack of bias in both the research process and outcomes. Neutrality is connected to objectivity, which is attained in quantitative research when reliability and validity are established (Sandelowski, 1986). Lather (1986) comments upon the need to address neutrality in action research as otherwise, the: “rampant subjectivity inherent in the more phenomenologicaly based paradigms will prove to be the nemesis of new paradigm research” (Lather, 1986, p68). One of the difficulties with this statement is that neutrality may be neither possible, nor sensible, in action research where subjectivity and researcher involvement is valued. Carr and Kemmis (1986) argue that no researcher can truly detach themselves from their research and that to attempt to do so reduces understanding of the issues under investigation. They suggest that the action researcher’s familiarity with both the setting and participants enhance insight and understanding. This is very different from the requirement to remain objective and removed in positivist research. It also differs somewhat to the naturalistic position where the researcher acknowledges their influence on the situation but does not necessarily do so from a critical stance. Action researchers on the other hand are expected to try and critically explain how their presence and opinions altered the setting, values and beliefs of the participants (Susman and Evered, 1978) and how they in turn were influenced by the situation (Reason and Rowan, 1981). The issues of truth-value, applicability, consistency and neutrality are re-visited in chapter four in relation to the specific research methods adopted in the study. The following section briefly considers the ethical tensions involved in action research. 3.8 Action Research: The Issue of Ethics All research studies are governed by the Declaration of Helsinki that was originally adopted by the World Medical Association in 1964 and has subsequently been amended and updated (1975, 1983, 1989, 1996). Further advice can be found in professional guidelines which have been produced in support of the Declaration (Foster, 1997). However, Lathlean (1996) argues that the ethical issues arising in relation to specific research studies are, in part, determined by the methodological approach. It is therefore pertinent to briefly consider some of the issues that appear to have particular potential to create tension within action research studies. The concepts addressed in the following sections are participant vulnerability, voluntary consent, informed consent, anonymity and confidentiality. 3.8.1 Participant Vulnerability Participant vulnerability has been referred to in previous sections of this chapter, but is worthy of being re-emphasised and summarised here. It is suggested that emancipatory action research may cause participants to feel vulnerable because their working practices are subjected to challenge and change. Furthermore, Lathlean (1996) warns that the reality of health care practice, with its complex multi-professional and hierarchical structures, may prevent those who develop new understandings of practice from exercising power in their attempts to transform practice. As a consequence, this may lead to feelings of frustration and powerlessness that have the potential to cause emotional harm. Action researchers, adopting critical theory approaches, therefore need to determine in advance how they will support participants in reflecting upon their existing ideologies without increasing vulnerability. 3.8.2 Voluntary Consent The over-arching principle of research ethics is the requirement to obtain voluntary informed consent from all participants. Despite the emphasis upon collaboration, several action research studies have noted that participants do not always wish to maintain involvement over lengthy periods (Lathlean, 1996). The right to withdraw from a study is a standard criterion of voluntary consent, however, in action research an attempt to withdraw by one participant may leave others in the team applying pressure to retain that individual’s involvement (Meyer, 1993). The implications of non-involvement must therefore be discussed as part of the consenting process. An equally serious issue is the problem recognised by Titchen and Binnie (1993a) whereby the consent of original participants in the setting results in new members of staff being forced to participate in an on-going study. Even if applicants are informed of the existence and nature of the study at interview, it is unlikely that the impact of involvement will be sufficiently analysed by the individual at a time when they are focused upon employment opportunities. Students who are allocated their clinical placements by faculty have even less opportunity to decline participation. Strategies need to be developed to enable an on-going consent procedure for new staff, together with plans on how to move forward if certain team members decline to participate. It could be argued that the duty to care places an obligation on practitioners to be involved in activities that aim to explore ways of improving practice. Nurses also have a professional duty to provide evidence-based care and are accountable for the quality of care through both the Code of Conduct (UKCC, 1992) and Government legislation related to clinical governance (DH, 1997, 1998). The emphasis is upon creating an environment that allows excellence to flourish through the development of quality improvement programmes, evidence-based practice systems, life-long learning and partnerships in care management. These tenets could be viewed as consistent with those of action research, which aims to improve practice through collaborative research. However, tensions may arise between the over-arching principle of voluntary consent to participation in research and pressure to meet clinical governance requirements. Action researchers therefore need to be sensitive to these issues when seeking consent, otherwise ethical principles may be breached. Furthermore, if participants do not develop a sense of ownership regarding both the topic under investigation and the process of change, practice developments are unlikely to occur (Meyer et al, 2000). 3.8.3 Informed Consent The cyclical and collaborative nature of action research produces a further problem in seeking informed consent as the design tends to evolve through reflection upon previously collected and analysed data. The researcher is therefore unable to reveal the exact nature of involvement to participants in advance. Meyer (1993) points out that it is not simply the structure of the study that is relevant here, but also the difficulties that may be faced as the study progresses. As part of informed consent is weighing up the benefits of involvement against the risks, action researchers are in a difficult position. In section 3.5.1, it was argued that the design for the first cycle should be drafted in advance of seeking consent. If action researchers adopt a critical stance in an attempt to identify the related risks and benefits, potential participants can be provided with as much information as is available at this early stage. They also need to know, however, that the researcher cannot determine future risks and benefits. A further problem with informed consent is that potential participants may assume that the researcher knows what they are doing, particularly if they are perceived as having greater power than the practitioners (Jones, 1996). As a consequence, consent may be given without sufficient thought. To limit the impact of both of the problems addressed in this section, it is suggested that action researchers seek renewed consent for each stage of the study as it progresses. This was the position adopted in the proposed study. 3.8.4 Anonymity and Confidentiality Anonymity and confidentiality are important ethical considerations in all research projects. In positivist and postpositivist studies, with their emphasis upon objectivity and generalisability, participants are viewed as representative examples of the population and, as such, are of limited interest as individuals. Anonymity is therefore easier to facilitate (Lathlean, 1996). Conversely, in action research with its focus upon local practice problems and the generation of situation specific functional knowledge, the relevance of the context and participant characteristics are such that it may be difficult to conceal the setting. Data obtained through qualitative methods, such as field notes and interviews, may also reveal identifying details. Participants therefore need to be appraised of these issues in advance and agreement reached on whether anonymity should be aimed for. The right of participants to confidentiality must also be addressed and strategies identified for using data that may reveal personal or sensitive information. This is important for both the internal feedback of data to participants and external dissemination. Decisions need to be made regarding what constitutes data if field notes and/or reflective journals are being used. Otherwise, information that is confidential may inadvertently become part of the research data. The ways in which participant vulnerability, voluntary informed consent, anonymity and confidentiality are addressed in the proposed study are discussed further in chapter four. 3.9 Action Research: Design Issues Adopting a collaborative action research design to study the potential for engaging patients in their postoperative pain management is consistent with the current emphasis upon developing partnerships with patients through practice and research. It is hoped that by embracing a critical theory approach, nurses will be empowered to challenge their traditional ideologies and move from unidimensional, directional forms of pain management, to one that is collaborative, patient-centred and able to acknowledge the multidimensional nature of pain. By considering the generic, epistemic and functional elements of ideology, together with the development of new praxis, it is hoped that a new action theory will be generated related to the incorporation of patients’ pain relieving strategies into the plan of care. Finally, the effects of the collaborative process on staff and practice transformation will be critically evaluated with a view to elucidating social theory and emancipatory knowledge regarding the journey from traditional postoperative pain management, to one that engages patients and nursing staff in a partnership. Further details on the study design are presented in chapter four. 3.10 Conclusion The traditional paradigms, which underpin research, may be constrained in their ability to generate knowledge for patient-centred approaches to care. This chapter proposes that a design which combines action research and critical theory, and which lies within a new paradigm of collaborative inquiry, has the potential to develop new praxis in postoperative pain management. The following chapter details the methods adopted within the study. CHAPTER FOUR: METHODS 4.1 Introduction Chapter three suggested that a design combining action research and critical theory may have the potential to free nurses from their existing ideologies and empower them to engage patients in their own pain management. This chapter begins by identifying the research questions, followed by discussion on the three action research cycles that formed the locus of the design. A detailed description of the methods is offered. Issues related to the characteristics of action research, scientific rigour and ethics are interwoven through each section. 4.2 The Research Objectives and Questions In a somewhat old, but useful, text on the generation of research objectives, Campbell et al (1982) advocate that questions should be practical rather than purely theoretical, should promote activity and interaction, be applicable to the real world and bring together several ideas at the same time. In the proposed study the ideas of critical theory, patient-centred care and nurse empowerment are conjoined in an investigation into patients’ self-generated pain coping strategies. The main aim of the study was to identify whether patients possess and utilise self-generated strategies for acute pain and, if so, to ascertain the effects of engaging patients in their pain management by actively supporting their strategies postoperatively. To empower nurses to deliver effective pain management and promote evidence-based practice, a second and concurrent aim was to engage nursing staff in the process. Research based upon praxis aims to uncover the main contradictions in a situation and then reconstruct participants’ ideology to enable emancipation and informed action. The process in which staff were to be engaged was therefore a critically reflexive one. Bearing in mind Campbell et al’s (1982) proposals and the study aims, the research questions were as follows: What pain relieving strategies are used by patients in acute pain? What are the effects on postoperative pain of patients using their own pain relieving strategies? How can these strategies be integrated into the nursing management of postoperative pain? What are the affects of an action research and critical theory approach on the nurses involved and on the nursing management of postoperative pain? Each question was developed and refined, in collaboration with the participants, as the study progressed. They were incorporated into the cyclical research design as explained in the next section. 4.3 Developing The Action Research Cycles Three main action research cycles were conceived, each containing a variety of methods and involving both patients and staff. In formulating the outline design, it was deemed inappropriate to adopt one of the action research approaches described by Holter and Schwartz-Barcott (1993) as they appeared too mutually exclusive. Equally, the Hart and Bond (1995b) typology, although more flexible and capable of providing insight into the overall direction of the project, is complex and difficult to explain to practitioners who are new to action research. Careful consideration was given to the need for participants to comprehend action research if they were to be truly involved. As mentioned in chapter three, Benner’s (1984) theory that novices need to understand the concrete elements of an idea before they can move on to dealing with more abstract issues, was borne in mind. It was decided to take up Meyer’s (1995b) suggestion of a flexible spiral of action-reflection cycles. However, Meyer’s (1995b) six stage process for moving through the cycles was replaced with the less complex four stage process advocated by Carr and Kemmis (1986). Carr and Kemmis’ (1986) process involves planning, acting, observing and reflecting. The planning stage focuses upon formulating a general idea for action. Participants usually have some sort of objective but the process required to achieve it is unclear. The acting stage involves a careful examination of the idea and a critical consideration of how the objective might be met. As a consequence, a cohesive strategy for action is identified. The observing stage incorporates implementation of the strategy combined with data collection. Finally, the reflecting stage involves analysis and critical discussion of the findings, refinement of ideas and a general desire to move forward into a new planning stage. The process is regarded as a spiral of research and action through two or more cycles. Meyer (1995b) highlights that action research projects are dynamic and that researchers need to be sufficiently flexible to allow the design to emerge as the study progresses. The critical theory stance also supports this premise as social action grows out of critical discourse and the examination of ideologies. In developing a general outline to help staff understand the design, four cycles were initially proposed. As the researcher and participants began to work together, the original second and third cycles were amalgamated. The design is shown in diagram one overleaf. “Telling the situation as it is”, was viewed as providing an opportunity to examine current practice from the perspective of both patients and staff. The aim was to gain a descriptive understanding of patients’ pain coping strategies and nurses’ current practice and ideologies regarding pain management. The cycle addressed the first research question and provided a baseline for the last question. Collaborative reflection on the results led the study into a second cycle which focused on the second and fourth research questions and allowed ideas for integrating patients’ strategies into postoperative pain management to be examined. The emphasis was upon developing praxis through “testing the proposed changes”. Diagram 1: The Three Action Research Cycles.  Exposure of nurses ideologies and patients’ strategies through focus groups and a surveyCritical reflexivity and emancipatory action: workshops, a RCT, questionnaires and field notesPractice transformation and theory generation: researcher withdrawal, field notes and audit The final cycle, “creating change in practice - implementation and implications” allowed consideration of theory and emancipatory action both in terms of the focus of the study and within the wider pain management context. Additional insight has been gained into the third and fourth questions during the early stages of this cycle. Each cycle is discussed in the following sections of the chapter, together with the methods and approaches employed to undertake the research and action. The way in which participants influenced the design is also highlighted. 4.4 Cycle One: Telling The Situation As It Is The first cycle focused upon identifying patients’ acute pain relieving strategies and seeking insight into nurses ideologies and practices (Geuss, 1981). The cycle was complex, as essentially two concurrent processes of planning, acting, observing and reflecting were entered into. Table 1 summarises the activity at each of the four stages. Table 1: The First Cycle: Telling the Situation As It Is. Stage Activity  Planning 1. Negotiating access with Trust executives and directorate manager 2. Identifying interested ward team, discussion of design, consent to participate 3. Identification of collaborative group & initial discussion on research ideas Acting1. Collaborative group agreement of objectives for cycle one 2. Collaborative development of focus group method & questions, survey design & schedule 3. Submission to Local Research Ethics Committee Observing1. Exposure of collaborative informants existing ideologies and practices through focus group discussions and a survey 2. Recording of field notes Reflecting1. Initial analysis of data 2. Feedback of results and critical discussion with collaborative group and ward team at staff workshops 3. Refinement of results with collaborative group and co-action researchers.  Each element of the cycle is detailed in the following sections. 4.4.1 Negotiating Access The planning stage initially involved negotiating access. The Executive Nurse Director was approached and agreement reached to present the research ideas at a Surgical Directorate Ward Sisters’ meeting. Participants were informed of my background, existing links with the Trust and interest in investigating the potential for supporting patients in using self-generated pain coping strategies. The idea of action research was introduced and a brief summary provided of the possible implications for ward staff of becoming involved in the study. This included recognition of the time and commitment required, as well as the opportunities for staff to learn about research and develop their pain management practice. Three of the Sisters immediately declined to be involved. A fourth, although initially interested, subsequently declined, as she was new and participating in medical research. The final Sister, who I had known in a previous role, expressed interest and offered to discuss the ideas with her staff. The Acute Pain Sister and I were then invited to discuss the implications of participation with the nurses. A copy of the proposal was requested and once the Sister was satisfied that all nursing staff, both qualified and unqualified, were interested in taking part, permission was granted to commence the study. 4.4.2 The Setting The Ward has 28 beds designated predominantly for orthopaedic surgery. At the beginning of the study, in late 1995, there were also some rheumatology beds. During the summer of 1997, major changes occurred within the Surgical Directorate and the Ward was re-designated orthopaedic/trauma. The Ward contains four bays of six beds each and four single rooms. Attempts are made to separate men and women, however, this is not always successful. Initially I found the layout difficult as none of the beds can be observed from the nurses’ station or Sister’s office. The system of care is team nursing. In 1995, there were approximately six qualified nurses and three health care assistants (HCAs) for each of the two day duty teams and one night duty team. The philosophy, which is posted on the Ward where patients and visitors can read it, emphasises the provision of: “high standards of individualised holistic care”. Of note, the last paragraph reads: “As health carers our goal is to provide quality care through research-based practice by continually updating our knowledge and skills.” Five years ago it was rare to see research, or research based practice, acknowledged within ward philosophies at the local Trust hospital. It was with some excitement that the nurses commitment to research was noted, however it was not altogether surprising. The Ward has a reputation among students for being a popular placement with many learning opportunities, a focus upon conjoining theory and practice and, despite the exceptionally busy nature of the setting, for providing excellent support. The Sister had previously worked as a Senior Staff Nurse on an orthopaedic/trauma ward in another Trust where I was the Link Lecturer. This ward had also had a reputation for evidence-based education and practice development. In an early nursing publication on action research, Webb (1989) emphasised the importance of action researchers being viewed as clinically credible by the nursing participants. My previous relationship with the Ward Sister was pertinent. She was able to confirm to the team that I was not “simply” a researcher, but had also spent time in practice as a Link Lecturer. An unanticipated problem was that there was a consequent expectation that my knowledge of orthopaedics was better and more up-to-date than it was. However, this was subsequently turned to advantage when I asked to spend three shifts working on the ward to gain insight into the organisation of care, nursing practice and, in particular, the approach to pain management. During these shifts I emphasised that while I was bringing knowledge and practical understanding of pain management and action research to the setting, the Ward nurses were experts in orthopaedic nursing who knew the setting and care requirements from “the inside” (Holter and Schwartz-Barcott, 1993). Respect for individuals’ knowledge and the need for mutual sharing was emphasised from the start. The three observational shifts were undertaken in January 1996. Field notes were completed during and immediately following the shifts using a journalistic style to record initial impressions and a strategy based on the work of Spradley (1980) (see section 4.5.4.1). The observations were shared with the nurses at a lunchtime handover. Although they are part of the results of the study, a summary is presented here to provide insight into the setting. On each shift I worked with a nurse who was allocated the care of six or seven patients. The handover from night staff was brief, focusing purely upon how well the patient had slept and any specific needs that had arisen over night. The one exception was that patient’ status in relation to pain and analgesic management was mentioned frequently. As it was possible that my presence had prompted this, I asked several staff whether the detail was normal practice. The quizzical looks I received, and the emphatic “yes of course” responses, were sufficient to reassure me that the overt recognition of pain and its management, was real. Each of the qualified nurses commenced their shift with a drug round, which enabled them to ensure that their patients received analgesics prior to the morning activity. It also allowed the nurses to say good morning to patients and assess their needs thus enabling prioritisation of care. Care involved meeting hygiene needs, mobilising chair to bed and talking to patients. The attention to detail was considerable. Even on the busiest of the shifts time was made to offer bowls for feet soaks, talk to patients and educate them as required. Pain management was generally, but not solely, reliant upon analgesics. Staff had a positive attitude towards pain control and used education, positioning, reassurance, touch, ice, warm baths and distraction, in addition to analgesics. However, all the strategies were nurse identified and initiated. None of the patients were asked what they do to relieve pain. Consequently, power appeared to rest with the nurses. It was interesting to note that only the analgesics were recorded in the patients care plan, suggesting that Briggs and Dean’s (1998) assertion that lack of documentation reflects a lack of practice may be flawed. An additional conjecture was that the nurses only saw analgesic management as being valued by others and therefore worthy of being noted. Several comments were recorded about tensions that existed on the Ward relating to nurses attempts to have analgesic doses and medications altered. There was a sense that nurses perceived medical staff as using covert mechanisms to control prescribing, which left nurses frustrated and patients in pain. These issues are further addressed in section 4.4.5. My perception that communication was hindered by the ward layout was also flawed. Although the teams operated quite separately, there was considerable discussion between them and everyone appeared reasonably well informed about all patients on the ward. Further evidence of effective communication was noted in relation to the study. As the field notes highlight, I was: “very impressed with the fact that everyone either knew my name or at least that I was the “pain” lady!” Effort was made to introduce me to members of the multi-disciplinary team as they visited the ward. Many, including the Senior Registrar, knew about the study and spent time discussing it with me. This curiosity in a nursing study was viewed as being somewhat at odds with the tensions described by the nurses above. To some extent, the level of interest and awareness was also over-whelming, as I came to realise what I had committed to and to comprehend the importance of developing a collaborative framework to ensure that participation in the study added to, rather than detracted from, practice on the Ward. The collaborative framework is outlined below together with a description of my developing role on the Ward. 4.4.3 The Collaborative Framework Earlier experiences with action research led me to believe that it was not feasible to involve all members of an acute ward nursing team to the same degree in an action research study (Burrows and Whyborne, 1996). As mentioned in chapter three, Titchen’s (1994) collaborative framework was proposed. The framework has a number of apparent benefits: Firstly, by allowing different levels of involvement the Sister and Ward team felt they could cope with the demands of the study. Secondly, those who had a particular interest in either research or pain management would have the opportunity to work closely with the project, while those who were less interested would not feel bound to participate at all stages and levels. This was an important consideration as it was one of the ways in which participants could consent voluntarily. Thirdly, the model allowed for the involvement of myself as an “outsider” researcher working in collaboration with “insider” change agents, thus legitimising both the action and research (Titchen and Binnie, 1993b). Titchen and Binnie (1993b) define the “outsider” remit as being one where the individual: “holds legitimate authority to be in the situation, but does not hold legitimate authority to change things within it” (p.19). Although I was welcomed by the team and involved in ward activities, I was clearly an “outsider” at the commencement of the project. This status altered after six months when the Ward’s Link Lecturer resigned. I was asked to take over the role by both the University and Ward. Following discussion with the Sister regarding the advantages of greater collaboration and the disadvantage of my out of date orthopaedic knowledge, it was agreed that I should take on the role. As the Link Lecturer I had legitimate reasons for increasing my visits to the area. This enabled me to get to know the staff well, gain further understanding of practice on the Ward and the tensions and conflicts that exist in any team setting. I also had the authority to change aspects of the learning environment. However, despite being an accepted member of the team, I remained an “outsider” researcher according to Titchen and Binnie’s (1993b) definition. The collaborative framework was detailed in chapter three. The way in which the framework was applied is outlined below. 4.4.3.1 Co-action Researchers At the co-action researcher level I shared responsibility for the study with the Acute Pain Sister. We had first met when I was a member of her interview panel. On her appointment, I worked closely with her, involved her in the University pain management sessions and, at the request of the Trust, acted as her clinical supervisor. During this time we became effective working colleagues, as well as friends. The Acute Pain Sister knew of my research plans and offered to be actively involved. Titchen and Binnie (1993a, 1993b) advocate that participants operating at the co-action researcher level meet on a regular and frequent basis. Although we did meet regularly we felt that we wanted to critically consider this advice in relation to the study. Our concern was that if we held on-going, separate meetings, the collaborative emphasis would be at the co-action researcher level to the detriment of other levels. We decided to meet monthly during the early stages, to share and record our philosophies and understanding regarding pain and its management, discuss how we saw our roles and note our hopes and fears. After this, we aimed to meet quarterly, unless the situation demanded more frequent discussions. Field notes and minutes were maintained and were frequently referred to. The structure and analysis of these records is addressed in section 4.5.4 and the findings are considered in chapter six. Of note, certain members of the collaborative group participated at the co-action researcher level during some of the patient focused elements of the study. This flexibility in the framework is commented upon by Titchen (1994) and has been positively recognised, by myself, in a previous study (Burrows and Whyborne, 1996). 4.4.3.2 Collaborative Group The collaborative group initially consisted of myself, the Acute Pain Sister, two Senior Staff Nurses from one of the day teams, a Junior Staff Nurse from the other day team, and the Clinical Nurse Specialist (CNS) Rheumatology who had expressed interest in participating. At an early meeting, a senior member of the night duty team was identified and subsequently approached to act as a liaison between the group and night staff. With a background, which included a period of prolonged night duty, I was particularly keen to actively involve the night staff. Not only would we need them on board if practice was to be transformed, but all too frequently night staff are simply forgotten and thus excluded from essentially daytime activities. The core of the collaborative group remained fairly stable during the study. One of the Senior Staff Nurses changed wards but maintained her involvement. The Night Nurse transferred to days and became an active group member while continuing to liaise with the night staff. The Junior Staff Nurse left the ward and was replaced by another from the same team. However, she withdrew after six months due to other commitments. Two new Staff Nurses expressed an interest in increasing their involvement and subsequently became active members of the group. Sadly, none of the HCAs took up the standing invitation to join the group although they were committed to the project. The Acute Pain Sister and the CNS had maternity leave during the study. The CNS did not resume her active status on return although she continued to receive minutes and offer comments. The Acute Pain Sister maintained contact throughout her leave. From November 1995 to the summer of 1999, the group met on some 20 occasions with the regularity and purpose of the meetings varying according to the status of the study. Minutes were recorded and copies circulated to group members, the Ward Sister and a file maintained on the Ward for all staff to access. Certain meetings were audiotaped, as they constituted focused discussions on the experiences of the collaborative group members. The structure and analysis of these recordings are detailed in section 4.5.4. At the first meeting, it was evident that some members were present because of their interest in pain, while others were equally keen to learn more about research. We discussed the research ideas in further detail, the proposed structure for the collaborative framework and the proposed action research cycles. The idea of undertaking a survey to identify the pain relieving strategies used by patients in acute pain was tabled. I used my previous work on facilitation to guide communication within the meeting and to introduce the idea of critical reflection. The analysis of facilitation undertaken by Burrows (1997a) was informed by the work of several authors, notably Heron (1977), Rogers (1983), Brookfield (1986) and Townsend (1990). Although it was carried out before I was aware of the tenets of critical theory, Brookfield’s (1986) work, in particular, is based upon an understanding of critical theory. In retrospect, it is evident that the conclusions of the analysis were relevant to a critical theory approach. In summary, facilitation was defined as: “a goal-orientated dynamic process, in which participants work together in an atmosphere of genuine mutual respect, in order to learn through critical reflection” (Burrows, 1997a, p.401). I had discussed with my co-researcher the need to encourage critical discourse from the first collaborative group meeting. We believed that one way to promote this was to emphasise the expert partnerships involved in the study. Each participant’s theoretical and practical knowledge was valued for its worth and uniqueness. Group members were actively encouraged to share their thoughts, opinions and ideologies and each contribution was considered. A deliberate effort was made on the part of the co-researchers to emphasise their own novice status regarding what might, or might not, work with patients and staff in practice. As the first meeting progressed, it became clear that the groundwork had been laid for participants to feel sufficiently comfortable to challenge each other and be challenged. Group members successfully contested several of my own assumptions and thoughts in relation to the research ideas, design and current practices. This initial critical climate was maintained throughout the study and is commented upon further in chapter six. The role of the collaborative group was to participate fully in the planning and acting stages of each cycle in both a practical and critically reflexive manner. Involvement in the observation stage was negotiated on a group basis with regards to collecting data from staff and on an individual basis for patient data collection. The differentiation was important to promote confidentiality around staff data and to acknowledge individual preference, time and informed consent for involvement in patient data collection. Active participation in data analysis, which formed the first step in each of the reflecting stages, was difficult, as much of the initial analysis was computer assisted. Efforts were made to explain the analysis processes and, in turn, the group questioned and challenged the findings helping to uncover contradictions, reconstruct ideology and develop ideas for subsequent informed action. The group members were fully participative in the feedback processes and in critically discussing ideas for praxis transformation. It was the Ward based members who, together with the Acute Pain Sister, took the lead in planning and organising change. Members of the collaborative group were encouraged to maintain reflective journals. Reflections were shared and discussed at meetings, however, the records were never viewed by the co-action researchers. The reasons for this and its consequent implications are addressed in chapter six. 4.4.3.3 Collaborative Informants Titchen (1994) suggests that collaborative informants participate fully in planning and implementing action, but view themselves as subjects in terms of the research. Although it is helpful to have a level of involvement that requires the minimum time commitment possible, there is some conflict in constraining participants to operate only at this level. Firstly, to do so reduces the potential educational and emancipatory benefits for both the individual and team. Secondly, there is an ethical issue related to the rights of individuals who have given informed consent to participate, expecting the potential benefits that participation involves. I was collaborating with a motivated and interested team who understood that they could not all be involved at the collaborative group and co-action researcher level, but who nevertheless expected to invest time in the study and, quite rightly, wanted something in return. In contrast to Titchen’s (1994) model, we decided to involve interested collaborative informants in planning data collection and critically reflecting on results. This opportunity had to be balanced with the caveat that participation was voluntary, determined on an individual basis, with no pressure applied by the co-action researchers, collaborative group, or Ward Sister. Consent to participate was only sought for each new activity following receipt by individuals of letters and information sheets. The specific way in which this was operationalised is outlined in the relevant methods sections. It was hoped that by enhancing involvement, consciousness would be raised and collaborative informants would be able to challenge their ideologies, reflect critically upon practice and enter into a process of informed action led by the collaborative group. The additional involvement would enable those who wished to move into the collaborative group to do so with some confidence in their knowledge and understanding of the totality of the study. To some extent the approach also allowed individuals to exercise their right to withdraw from aspects of the study, such as data collection from staff. Unless the team as a whole decided to withdraw, it would not be possible for individuals to withdraw totally as changes in practice would have to be implemented across the ward. Sturt (1997) provides a useful discussion on weighing the balance between ethical principles of beneficence, which not only involves preventing or removing harm but also promoting good, and nonmaleficence, which refers to not inflicting harm (Beauchamp and Childress, 1994). By raising consciousness of false ideologies and empowering practitioners to act, Sturt (1997) argues that harm is removed and good promoted. The risk of exposing conflict and destabilising existing structures, which have the potential to inflict harm, is therefore said to be ethically out-weighed by the good arising from the development of critical self-awareness, praxis and the generation of new functional theories. However, this appears to be a somewhat paternalistic stance which ignores the risk of participant vulnerability within emancipatory action research. The above section has discussed the collaborative framework in detail. The chapter now moves on to consider the acting and observing stages of cycle one. 4.4.4 From Planning to Acting and Observing Once the collaborative framework had been established and staff were aware of the potential for moving between levels, the collaborative group were able to consider how they might address the agreed research questions. A decision was taken to record field notes to inform the different stages and assist with data organisation and analysis. Because the notes were maintained throughout the study the section discussing this method has been placed towards the end of the chapter to reduce repetition and enhance clarity. To obtain insight into the Ward nurses’ ideologies and perspectives on current practice and to promote consciousness raising, a series of focus groups was planned. The findings would help inform the question “what are the affects of an action research and critical theory approach on the nurses involved and on the nursing management of postoperative pain? Finally, in this first cycle, a survey was designed to establish whether patients in acute pain utilise self-generated strategies and to gain insight into their ideologies regarding the use of strategies both at home and in hospital. During the acting stage, data collection instruments were designed and ethical approval sought as was highlighted in table 1. Data were subsequently collected in the observing stage. Both the focus groups and survey are discussed below. 4.4.5 The Focus Groups To examine the effects of action research on participants in the study, it was thought to be necessary to establish their ideologies regarding pain management early on. This would then provide the basis for future critical discussions. Focus groups involve a research facilitator guiding six to ten participants in a discussion on a specific topic (Krueger, 1994). According to Schroeder and Neil (1992), the main aim of focus groups is to: “elicit informants’ unique, and often unanticipated, perspectives” (p.266). Thus knowledge is generated through gaining insight into the way participants view their world. Effective facilitation of focus groups should prompt a rich exchange of ideas in an open and honest climate where participants feel able to challenge and be challenged (Burrows and Kendall, 1997). The decision to use focus groups was based upon their potential to uncover the ideologies that informed pain management, to promote critical discourse, to raise consciousness regarding false ideologies and power balances and to be collaborative in nature. It was also hoped that participation would enable nurses to begin reflecting upon their individual and team approaches to pain management. Finally, the discussions would provide an opportunity to ask whether participants had ever considered identifying and supporting patients self-generated pain coping strategies and to determine their attitudes towards engaging patients in their own pain management. The focus groups thus offered an opportunity for both data collection and action. Both co-researchers read around focus group methodology and my previous experiences of using this method were used to inform discussions at the collaborative group meetings. Details on the approaches and processes may be found in Burrows and Kendall (1997) and Burrows (1998) (see appendix 1 and 2). However, some of the issues are summarised here to identify the way in which focus groups were undertaken in the study. 4.4.5.1 The Process Reiskin’s (1992) four-stage process of planning, conducting, analysing and reporting was used as a framework for the process. The most important initial step in planning was to determine the number of groups that would be sustainable by the Ward. Twenty-eight nurses and the CNS Rheumatology were eligible to participate. If more than eighteen consented to do so, it would be possible to run three groups. Krueger (1994) suggests that theoretical saturation, which is the point at which no new information is obtained, is unlikely to be achieved with less than three groups. To enable voluntary informed consent an information sheet was devised by the collaborative group and forwarded to each nurse (see appendix 3). Emphasis was placed upon the fact that nurses could chose whether or not to participate and the point was made that the discussions would be audiotaped. Krueger (1994) advocates that potential participants are informed that each individual’s views and experiences will be valued. A paragraph highlighting this point and laying the ground for promoting critical discourse was included in the information sheet. Burrows and Kendall (1997) note that informing participants of the discussion questions in advance may detract from interaction within the focus group as individuals arrive with formed responses. However, Lankshear (1993) advises that the questions are included on the information sheet to enable participants to consider their views prior to exposure to the group opinion. The Ward members of the collaborative group supported the latter view and suggested that critical discourse could be enhanced if participants had time to consider their opinions before attending. Their advice was taken and the questions were included on the sheet. DeWolf’s (1985) belief that focus group discussions initially elicit general responses, before participants relax and move forward to offer specific and in-depth responses, was borne in mind in deciding upon the sequence of the questions as follows: What sort of things do you feel you do well on the ward in relation to pain management? What areas do you find difficult? What non-drug interventions have you thought about using, or actually used when managing patient’s pain? Have you ever thought about asking patients what sort of things they do to relieve pain and incorporated their strategies into the plan of care? It could be argued that although the above sequencing enables movement from the general to the specific, it is likely to encourage an early exploration of power balances and tensions in the setting, before the ideologies surrounding the multi-dimensional nature of pain are considered. As a consequence, participant vulnerability was a risk. This was recognised by the co-researchers who consequently spent some time discussing their facilitation approach. Both co-researchers acted as facilitators and, as outlined by Burrows (1998), aimed to achieve a balance between neutral investigator and collaborator within the discussions. Skills of facilitation were employed (Burrows, 1997a) and active listening, probing, the use of silence and summarising were used to encourage discourse. I modelled the style and approach at the first session to enable the less experienced Acute Pain Sister to adopt a more active facilitation role in the second and third sessions. Each group was planned to last for one and a half hours with the first 20 minutes being spent on welcoming participants, offering refreshments to help to relax them, re-iterating the guidelines and questions for discussion, and checking consent for audio-taping. The questions were addressed during the subsequent hour. The final ten minutes were allocated to summarising the session and, with the audiotape turned off, asking participants how they had felt about their involvement. Seventeen nurses expressed an interest in being involved. This constituted 72% (n = 13) of the qualified staff and 33% (n = 4) of HCAs. A decision was made to go ahead with three groups. Care was taken to balance each group with a mix of day and night staff, qualified and unqualified nurses. Six participants had to send their apologies on the day, one failed to give a reason and her right to withdraw without doing so was respected. Four nurses requested an alternative date, indicating their interest in participating in both the focus groups and study. Unfortunately due to other role commitments and annual leave, the co-action researchers were unable to offer a fourth date. The participants accepted these reasons sympathetically. 4.4.5.2 Analysis Krueger (1994) suggests that focus group data are analysed in 3 stages. Firstly, viewing the raw data and ordering it into categories. Secondly, descriptive statements are given for each category followed by illustrative quotations. The choice of quotations should be determined by the purpose of the study and reflect the range and diversity of opinions and the typical and common. Kitzinger (1994) adds that quotations should demonstrate the flow of the discussion, highlighting complementary and argumentative interactions, by being presented in a sequential order and conversationally. This provides qualitative insight into the process and enables the researcher and others to assess whether changes of opinion occur as the focus group progresses (Kitzinger, 1994). Interpretation of the data is the final stage and incorporates analysis in light of the researcher’s contextual knowledge. While Krueger’s (1994) framework is clear, the rather simplistic overview did not provide the detail needed to analyse the focus group data with any perceived degree of rigour. Knodel’s (1993) discussion of coding however was helpful. He suggests that focus group data can be considered in two ways. Either data can be analysed in a superficial manner for practical recommendations, or it can be considered from a scientific perspective such that the phenomenon under investigation is explained and understood. The latter perspective involves a structured approach to the mechanical stage of coding the raw data. Codes consist of short words that summarise the content and are placed in the transcript margin next to the specific segment of discussion. A separate code is identified for each focus group question and sub-topic. Where a sub-topic occurs in a part of the text of a different question, it should be coded with both the sub-topic code and topic code. Once coding is complete, a table or “overview grid” is compiled. This provides a descriptive summary of the focus group content, the relationship between the topic variables and the extent and direction of consensus. The grid should enable similarities and differences between groups to be quickly assessed. 4.4.5.3 The Process Adopted The three focus group tapes were transcribed in full and I compiled a brief set of notes on focus group analysis for my co-researcher. We decided to use a combination of the processes described by Krueger (1994) and Knodel (1993). The transcripts were read, followed by a second read through of the first focus group transcript. Initial hunches regarding possible themes and sub-themes were recorded independently. The hunches were shared and a critical discussion entered into. The transcripts of focus groups two and three were then read again in light of our understanding. Further hunches were noted, shared and critically considered. With the exception of one area, we agreed all the codes without any differences of opinion. The exception related to whether a particular topic should be coded “communication between nurses” or “teamwork”. Given the possibility that these were in fact two separate topics, a “wait and see” approach was adopted enabling us to look for both during the analysis process. Codes were finalised and I took responsibility for marking the transcripts. Segments that were unclear were discussed and decisions made regarding the code to be applied. Once the segments from all three focus groups had been coded they were organised under their topic and sub-topic heading in sequential order and amalgamated into an “overview grid”. A further document was created giving a brief description of each topic and sub-topic. Illustrative quotations highlighting the range of statements, together with those that were typical and common were included. The two documents were appraised and discussed in an effort to interpret meanings. 4.4.5.4 Scientific Rigour In accordance with the qualitative nature of focus groups, rigour was assessed in relation to credibility, fittingness, consistency and neutrality (Guba and Lincoln, 1981) which were outlined in chapter three. Credibility was checked when sharing the results with the participants (see section 4.4.7). Fittingness can only be fully assessed once the results are published, however, the co-researchers recognised that much of what was said, particularly in relation to power balances and the constraints and tensions surrounding pain management, fitted their experiences elsewhere. Consistency was addressed in two ways, firstly by maintaining a decision trail, similar to that described in Burrows (1998), and secondly, by working with the co-researcher and collaborative group to enable participant validation (Titchen, 1995). Neutrality was considered by ensuring that the decision trail included reflections on the involvement and influence of the researchers. The focus group findings are presented in chapter six. 4.4.6 The Survey The focus groups provided insight into current practices from the perspective of the nursing staff. Similarly, the survey was designed to provide information on the patient’s perspective prior to practice developments being identified and implemented. Because action research focuses upon practice problems within a particular context, it does not specifically aim to develop knowledge which can be generalised from one setting to another. Thus the majority of action research studies tend to use qualitative methods which enable ideology and the meanings of social action to be investigated. Similarly, critical theory favours methods that prompt reflexivity and critical discourse. However, Holter and Schwartz-Barcott (1993) assert that action research is not bound by the use of particular methods and that researchers should therefore be sufficiently flexible to use appropriate methods for the question. Examples are available in the literature of studies employing methods such as quasi-experiments and surveys (Armitage et al, 1991; Jones, 1996). The initial research question was “what pain relieving strategies are used by patients in acute pain?”. Questionnaires or interviews appeared to be appropriate methods to employ. The adoption of critical theory and action research would suggest the use of unstructured interviews that enable respondents to talk freely around the topic area. The resulting in-depth data would be analysed qualitatively and provide the researcher with insight into the respondents unique perspectives and ideologies (Fielding, 1994). Although the idea of employing unstructured interviews was attractive, the emphasis upon changing and developing practice in action research raised questions as to whether this approach would provide the information sought. The collaborative group sensed that staff would require a certain amount of education to support patients’ strategies postoperatively. Indeed, previous work by Burrows (1990, 1997b) had indicated this to be the case. Given that time is a finite resource in any acute ward nursing team, we felt that we needed to know which strategies were most frequently used by patients in acute pain so that education could be prioritised. Equally, the range of strategies employed by patients was of interest so that staff were not disempowered by being unaware of the scope of strategies used. Finally, if we were to suggest that patients used their strategies postoperatively, we wanted to know that they were perceived as helpful in relieving pain, otherwise we could disempower patients. The literature on self-generated strategies, reviewed in chapter one, was unable to provide the answers that we sought. Studies were either too small to indicate the range of strategies that might be employed, lacked detail regarding the strategies used, or were North American in origin. As sociocultural background can influence pain behaviour (Madjar, 1985; Greenwald, 1991; McGuire, 1992) we felt that a survey of a local sample was needed to provide insight into the types of strategies patients admitted to the Ward might use. This perspective was also consistent with the situation specific nature of action research. 4.4.6.1 The Method As the Ward’s main focus was orthopaedic and trauma surgery, we decided to undertake a survey of patients attending orthopaedic and surgical outpatient clinics. These were an accessible sample who were representative of the types of patients admitted both to the research ward and the Surgical Directorate generally. Atkinson (1996) highlights that surveys can be analytical or descriptive, with the latter aiming to describe characteristics about a population and the former to investigate associations between different variables within the population. The survey for this study was descriptive as the aim was to establish the pain coping strategies used by patients and determine respondents views on their utility and potential for postoperative pain relief. Although surveys generally use questionnaires, they may also employ interviews (Fielding, 1994) and both approaches were adopted. The questionnaire was designed to provide information on the characteristics of the sample and existence, range and type of self-generated pain coping strategies. Specifically, respondents were asked: Are there any things you do (thoughts and actions) when you are in pain to help relieve your pain or discomfort? Are there any things your family and friends do for you when you are in pain to help relieve your pain or discomfort? What sort of things do you think the nurses and doctors might do to help relieve your pain? In determining the wording, the principles suggested by Barker (1996) were adhered to. The questions needed to be open and encourage multiple responses. During the pilot study, the questions commenced with “what sort of things do you/your family/the nurses and doctors do...” Although this produced the information sought, respondents suggested that they found the wording clumsy, hence the change to that shown above. The subsequent wording is less open, as it can be answered with a “yes” or “no” response, however, the information sought was obtained in practice. The questions also needed to indicate that both cognitive (thoughts) and behavioural (action) strategies were of interest and should be recorded. Lay, rather than professional language was used to enhance comprehension. Readability evaluations, using the Flesch Formula, indicated that the first and last question fell within the “easy” to understand category and the middle question within the “fairly easy” category (Hughes and Foster, 1997). The interviews were designed to elicit additional information from a sub-sample. Interviewees were asked to provide an example of a time when they had experienced physical pain and describe the pain relieving strategies employed, what prompted them to choose that particular technique, the way in which the strategy helped and how successful it was in relieving pain. They were then asked to repeat the process when recalling experiences of every-day aches and pains. Finally, interviewees were asked for their opinions about using self-generated strategies in postoperative pain management, thus providing some insight for the collaborative group regarding the implementation of change in practice. The questions posed during the interview were a mix of fact seeking questions and questions designed to explore ideology. An initial pilot study was undertaken to ensure that respondents understood the questions and that our administration strategies were workable within the outpatient setting. Following some minor changes, the main data collection period was commenced. 4.4.6.2 The Sample To undertake a survey which would be meaningful for local practice, we not only wanted to involve members of the local population who were likely to be admitted for surgery, but also ensure a sufficiently large sample to enable us to draw sound conclusions. Statistical advice was sought regarding the sample size. Following discussions about the proposed sampling method and variables known, or thought, to influence acute pain perception, the statistician advised that data would be required from 200 individuals in the questionnaire arm and that 10% of the sample (n = 20) should be interviewed. All new adult patients attending surgical and orthopaedic outpatient clinics and those attending pre-admission surgical assessment clinics were approached until the required sample figure was attained. All relevant patients, attending the clinics during the four-month data collection period, were given the same chance of involvement. it was hoped that this would enable the sample characteristics to reflect the local population in terms of gender, ethnic origin, religion and social class, and of patient admissions to the Surgical Directorate with respect to age, marital status, medical history and pain experiences. Combined with the sample size, this would not only provide us with sufficient data for the study, but should also enable generalisation to similar populations (Blacktop, 1996). 4.4.6.3 Ethical Issues Access and permission to undertake the survey were sought from the Outpatients Services Manager. The protocol was submitted to the Local Research Ethics Committee and approved with one or two minor amendments. Letters were written to the relevant consultants requesting access to patients in their clinics. The Manager introduced the research team to the clinic nurses, who in turn introduced us to the medical staff. Both parties showed considerable interest in the study and several nurses expressed interest in helping with data collection. Permission was sought from the Manager and full explanations provided regarding the implications of involvement. The way in which outpatient nurses collaborated is described in section 4.4.6.4. Written informed consent was sought from all patients (see appendix 4). National guidelines advocate that a minimum of 24 hours is available from first approach to seeking informed consent. This allows time for patients to ask questions and enables a “cooling off” period (Hughes and Foster, 1997). This was not possible within the clinic setting. Patients had to be approached on arrival and the Local Research Ethics Committee was informed that consent would be sought within 15 minutes. The questionnaire was attached to the information sheet so that individual’s could look at the questions before making up their minds. It was stressed that they should not begin completing the questionnaire until the researcher had returned and the consent form was signed. Allowing patients to view the questionnaire in advance may have increased non-response error (see section 4.4.6.4). However, with the lack of time available between initial approach and consent being sought the chosen method was deemed to offer the best balance between voluntary informed consent and scientific rigour. Patients were informed that, although there were no risks foreseen with the study, they would be asked to think about pain and the ways they dealt with it. The ethics committee were advised that the risk of distress was regarded as minimal and that if problems did arise, referral to the Acute Pain Service had been agreed. Warning of risks may increase non-response rates, however, all potential research subjects have a right to be appraised of possible discomfort and risks (Grubb, 1997). Separate consent was sought for participation in the interviews so that patients who only wished to complete a questionnaire were not prevented from participating. Consent was re-checked immediately prior to the interviews. The questionnaire and interview schedule may be found in appendix five. 4.4.6.4 Scientific Rigour According to Atkinson (1996), three types of errors can occur in survey designs: sampling error, non-response error and response error. Each is considered in turn followed by discussion of inter-rater reliability. Sampling error relates to the problem of inadvertently excluding individuals who represent the population, or selecting greater numbers of individuals with a specific characteristic that is not evenly distributed within the population. The result, in statistical analysis, is that the sample statistic does not reflect the true population parameter. Consequently, findings cannot be generalised to the population at large. Sampling error is of greater importance in analytical surveys where statistical analysis aims to consider associations between variables rather than the frequency of an occurrence (Atkinson, 1996). In this study, sampling error was considered from the viewpoint that different clinics might have different patterns of attendance in terms of population characteristics. All relevant clinics were, therefore, included. Non-response error refers to error that may be introduced if information is not obtained from all those included in a sample. While prisoners and individuals of 17 years or younger were listed in the exclusion criteria, patients who did not attend (DNAs) the clinic for their appointments, those who were not approached and those who declined participation could all add to non-response error. It is possible that DNAs were in control of their pain and less motivated to attend clinic, while individuals who declined may have done so because they believed that they did not use pain coping strategies. Non-response for these reasons could skew the findings to produce either under or over-estimations of the frequency, type and range of self-generated strategies. Atkinson (1996) suggests that clear explanations regarding the nature and importance of the study, combined with well-constructed data collection instruments, enhances motivation to participate and reduces non-response error. The relatively large sample size should also reduce the impact of non-response error. The response rates are detailed in chapter five. Response error incorporates random error which arises as a result of making mistakes when recording data, and systematic error which occurs if there is an error in the way in which measurements are made. Bias within the way questions are worded should be addressed through careful instrument development and piloting (Atkinson, 1996), while frequency checks can be used on data to uncover recording errors. The occurrence of random and systematic error may be increased if more than one researcher is involved in gathering data. Regardless of who is doing the measurement, the method should be sufficiently accurate and reliable to ensure that any influence or bias on the part of the researcher is reduced to a minimum (Presly, 1996). This relies not only upon using carefully prepared data collection instruments, but also upon inter-rater reliability where efforts are made to enhance the consistency of approach by different researchers. In using Titchen’s (1994) framework, there was potential for collaborative group members to be involved with the co-action researchers in collecting data. Although I had previous experience of data collection, the group, with the exception of the CNS Rheumatology, were new to research. All members were offered the opportunity of participating in the survey. One of the Senior Staff Nurses, who worked part time, expressed an interest in being involved. Following discussion with the Ward Sister, it was agreed that this nurse would act as a co-action researcher with some time being allocated from the Ward and the remainder from the nurse’s own time. The CNS Rheumatology also offered to assist with questionnaire administration. A paper on inter-rater reliability was produced for consideration and discussion. The two co-action researchers and the Staff Nurse met to agree the approach to be used. Decisions included what to wear, how to greet people, explain the information sheet, consent form and questionnaire and how to start and close the interviews. Scripts were written to enhance consistency and copies were given to the CNS Rheumatology to critique. Reflective records on how each interview progressed, non-verbal communication and any intervening factors related to ourselves or the patient were to be kept. Once the decisions were finalised we practised on each other, with one person acting as the patient, another as the researcher and the third as a critical observer. The sessions were taped and critically reviewed by one of the doctoral supervisors. Once we were achieving consistency, the pilot study was commenced. For the first half of the pilot we worked in pairs with one acting as the researcher and the other as a critical observer. All interviews were taped, listened to and copies sent to the supervisor for comment. Eliminating inter-rater bias can be difficult, however the procedure outlined above ensured that we acted as consistently as possible. As previously mentioned several of the Outpatient Nurses offered to assist with administering the questionnaire. Discussion took place within the collaborative group and with the supervisors regarding the advisability of involving these nurses. The tension of collaborative action research, versus scientific rigour, formed the basis of these discussions. Our conclusion was that if we trained the nurses to follow the questionnaire script, stressed the importance of voluntary informed consent, observed them in action, and worked with them, rigour and collaboration could be combined. Three of the original five who expressed interest, consented and collected questionnaire data from 22 patients. 4.4.6.5 Analysis The main analysis questions, identified by the collaborative group, were: What are the characteristics of the sample? What type of strategies do patients in acute pain use? What is the frequency of use for different strategies? What is the range of use? What themes can be identified for the strategies? Do patients find their strategies helpful in reducing acute pain? Is there any difference between the types of strategies used for “everyday” pain and those used for more “serious” acute pain? How do patients feel about using their self-generated strategies in hospital. Variable labels were identified for the questionnaire data, which were then entered into the Statistical Package for Social Science (SPSS) computer software programme (SPSS Inc., 1994). With the exception of the strategy labels, variables were easily identified, as the questionnaire was pre-coded. Nevertheless, the co-action researchers spent some time checking and agreeing the appropriateness of the labels. The strategy labels were more difficult to identify. Initially, each new word used by a patient was coded. However, this became unwieldy and unhelpful. For instance, there was no need to treat different anti-inflammatory drugs separately. The co-action researchers, with assistance from the Senior Staff Nurse, spent time agreeing codes and checking and re-checking the data. The questionnaire data were analysed using descriptive statistics including, frequency counts, means, medians and modes. Bar graphs, pie charts, histograms and tables were produced as relevant to enhance visualisation of the data. The strategies were categorised into draft themes. The content and labelling of the themes were discussed at a collaborative group meeting and are presented in chapter five. The interview data were transcribed and a simple overview grid produced. Columns contained responses to the interview questions and separate rows were designated for each of the 20 respondents. Numerical data were added to the SPSS file where appropriate. A summary report was produced which included the qualitative data. Much of the latter relied upon verbatim quotations, together with one or two critical incidents. The report was discussed at a collaborative group meeting and used to prompt ideas for practice transformation. The findings are detailed in chapter five. 4.4.7 Reflecting Upon The First Cycle The reflecting stage in the first cycle, not only included analysis and discussion of the focus group and survey results at collaborative group meetings, but also included providing feedback to others. Several survey respondents requested copies of the results and a two-page summary was produced and sent out. Copies and a covering letter were forwarded to the Local Research Ethics Committee, Executive Nurse Director, relevant Consultants and Outpatients Nurses. The results were also verbally presented to the Outpatients staff at the Manager’s request. The majority of the nurses attended and discussion was animated. The co-action researchers were asked to keep Outpatients informed of progress. The main group with whom the results needed to be shared were the collaborative informants. The best way to approach this was debated within the collaborative group and with the Ward Sister. The Executive Nurse Director and Ward Sister were keen to enable staff to participate in discussions away from the ward. A series of staff workshops were suggested. 4.4.7.1 Staff Workshops Four workshops were organised, with each lasting two hours. To facilitate reflection on the first cycle, the first hour was spent discussing and refining the focus group and survey results. The second hour involved planning for cycle two through a critical discussion on ideas for future research and action. As with the focus groups, nurses were invited to attend the workshops by letter. It was agreed that those attending in their own time could either claim time from the ward, or be recompensed at bank staff rates. No monetary claims were made and many staff did not re-claim their time, highlighting their overall commitment to the study. The information sheet inviting participants to the workshops is contained in appendix six. Twenty-eight staff were eligible to attend the workshops of whom two were on maternity leave and two were about to leave the ward. Of the remaining twenty four, all but one of the qualified nurses on day duty, together with one HCA attended a workshop. None of the night staff were able to attend. A comprehensive pack was sent to non-attendees containing the results and a summary of the decisions reached for cycle two. The night staff liaison and one of the Senior Staff Nurses in the collaborative group took responsibility for speaking to each of these individuals to respond to any questions they might have and check their understanding of, and consent to, the proposals. The co-action researchers took it in turns to facilitate the workshops, present the findings, prompt critical discussion and record notes. The latter captured the general directions of the discussion and with consent, individual’s specific comments. The notes were transcribed, added to the field note data (see section 4.5.4) and used to inform the decision making process. Some of the data are presented here to illustrate the rationale for the decisions regarding action and research within cycle two. The remaining data are incorporated into chapter six. The focus group and survey results were regarded with interest. Some of the difficulties linked to pain management on the Ward had been resolved through raising consciousness. In turn, this led to new praxis whereby the nurses felt empowered to take the initiative in improving communication with the medical staff. The range and type of strategies used by the survey respondents raised considerable curiosity. At the first workshop it was suggested that patients’ self-generated strategies should be identified on admission to the ward and actively supported postoperatively. Staff thought that this might enhance pain management and: “help the patient feel more in control, you know, if they’re doing their own thing rather than having us tell them what to do all the time”. Several participants indicated they would need advice on ways to support patients and proposed that the collaborative group listed strategies that could be supported on the ward, together with ideas on how the nurses could enable patients to use their strategies in the hospital environment. One example that was given was: “if the patient wanted ‘peace and quiet’ one possible way of at least simulating this would be to draw the bed curtains”. Another suggestion related to a patient who was currently on the ward and who: “normally gets up in the night to make tea and relieve her stiffness and distract herself from the pain. We could tell her to ring the bell to get help getting out of the bed, you know for safety reasons, and then she could go and get her tea from the trolley and that could help her pain”. These examples highlight how the nurses were grasping the components of patient-centred care, namely, focusing upon the individual, working with them to elicit and understand their values, beliefs and choices, identifying and meeting needs and viewing patients as partners with whom care management could be negotiated. Further details on the nurses altering ideologies and praxis are offered in chapter six. Prior to the workshops the collaborative group had tentatively discussed ideas for data collection in cycle two. Firstly we wanted to know the impact on patients of engaging them in their own pain management and secondly we wanted to consider how patients’ self-generated strategies could be incorporated into the plan of care. As an action research study it was evident that data needed to be collected from the perspectives of nurses and patients regarding the process. However, when faced with the idea of possibly interviewing patients to investigate their ideologies and perceptions about being engaged in their pain management, the collaborative group were adamant that they wanted to: “know that this thing, this approach, will work. It’s all very well us telling other nurses and the doctors that patients like it and it seems to work, we have to prove it works, can’t we do some sort of experiment?” This view was replicated during the workshops: “well, quite frankly I don’t want to spend time adding to my workload unless I know for a fact that it works. Is it going to help their pain, let them be less dependent on drugs and so on? We need to test this, do it properly”. “What’s the point of doing some sort of mumbo jumbo interview with patients? They will ask us ‘well does this really work...I’ve always wondered... and now your telling me it will even work after surgery, wow!’ It will make us look stupid if we can’t be sure and it would be unfair to the patients - not very empowering in my book. I think we have to do an experiment. Then we can do it properly, you know, know some answers for us and the patients”. Despite the emphasis in action research and critical theory on using methods, which explore perceptions and ideologies, through, for example, critical discourse, the nurses were keen to focus upon measurable issues. My attitude towards this and the tensions arising as a consequence of the design, are discussed in chapter six. Suffice it to say, at this stage, that participants’ views were acknowledged. The workshop discussions considered designs that would enable us to test the effect of engaging patients in their self-generated strategies against measurable postoperative outcomes. The views and opinions expressed at each workshop were used to inform critical discourse at the following workshop. Views were presented as offered and the co-researchers attempted to critically analyse the advantages and disadvantages of each with participants. Overall, there was a perception that an experiment was required to answer the research question “what are the effects on postoperative pain of patients using their own pain relieving strategies”. The detailed decisions regarding design and implementation of the trial are discussed in section 4.5.2 below. 4.4.8 Summary of Cycle One Cycle one described the current situation from the perspective of both nurses and patients in relation to ideologies regarding pain management and the use of self-generated pain coping strategies. The findings obtained from the focus groups and survey produced ideas to enable the group to move into cycle two. 4.5 Cycle Two: Testing The Proposed Changes The main focus of cycle two was to investigate the effects of patients using their own pain relieving strategies on postoperative pain. A concurrent aim was to consider how practice might be transformed to enable patients’ strategies to be identified and supported. In accordance with the overall design, the impact of the approach on participants’ ideologies, reflexivity and praxis was also considered. 4.5.1 From Planning to Acting, Observing and Reflecting Table Two outlines the stages in cycle two. Table 2: The Second Cycle: Testing the Proposed Changes. Stage Activity  Planning 1. Formulation of ideas for practice transformation, by collaborative group and at Staff Workshops 2. Acknowledgement of perceived need to test proposed changes Acting1. Critical discussion and development of research and action strategies at Cycle 1 Staff Workshops, collaborative group and co-action researcher meetings 2. Finalisation of RCT design and data collection instruments for submission to LREC 3. Collaborative planning for undertaking the RCT in practice 4. Co-researcher development of staff questionnaires Observing1. Testing of proposed practice transformation through RCT 2. Recording of field notes 3. Staff questionnaire 4. Researchers’ critical reflexive discussion Reflecting1. Initial analysis of data 2. Feedback of results and critical discussion with collaborative group and ward team 3. Refinement of results with collaborative group and co-action researchers 4. Further critical reflection on involvement in study.The staff workshops enabled a general idea for action to be formulated, namely, that we should test the proposed changes through a randomised controlled trial. The acting stage involved development of the research and action strategies, including the RCT, two staff questionnaires and items for the researchers’ critical discussion. As a consequence of the Local Research Ethics Committee bringing a meeting date forward by several weeks, the time set aside by the collaborative group to work on the RCT design and data collection instruments fell after the new closing date. I therefore had to take over the design but was able to meet or telephone the co-action researcher, individual members of the collaborative group and the Ward Sister. The focus of these brief discussions were practical queries, checking relevant information and, to some extent, critical discussion. To this end, I summarised the elements of the plan and requested critical comment. Several times, participants telephoned following the encounter with revised opinions and new challenges. On the whole, these reflected personal views, but there was evidence that collaborative group members were attempting to check out issues with collaborative informants, who were then able to offer their own comments. Where possible, all ideas were incorporated into the proposal. The co-action researcher and two members of the collaborative group checked the proposal and critically reviewed the data collection instruments prior to submission. It was acknowledged that considerable additional work would be required by the group to prepare us all for undertaking the trial. The acting and observing stages are now discussed in more detail. 4.5.2 The Randomised Controlled Trial Sackett et al (1996) suggest that systematic reviews of RCTs provide the most reliable sources of evidence for treatment efficacy. Where insufficient trials exist to undertake a review, a single, rigorous RCT is regarded as supplying the next best level of evidence on which to base practice. With their emphasis upon control and objectivity, RCTs are regarded as being firmly embedded within the positivist / postpositivist paradigms. Their use therefore seems an anathema to action researchers. Indeed, Allcock (1996) states that the: “experimental approach based on the traditions of positivism and the importance of empirical observation appears to be ideologically opposed to the basis of action research” (p.35). This perspective becomes even more pertinent when action research is combined with critical theory. Nevertheless, the literature contains some examples of positivist methods being used within action research designs. For example, Jones (1996) and Armitage et al (1991) used quasi-experiments to test the extension of nursing roles in emergency care and primary nursing, respectively. Barker (1992) used an RCT to test whether an empowering method of health visiting could make a difference to the developmental levels of young children. Forty health visitors acted as controls and 46 as intervention visitors. Intervention families were visited monthly for two years. Findings indicated that the intervention positively influenced child development and maternal self-esteem. Barker (1992) concluded that RCTs can be used to test the impact of empowering interventions within action research. In pain management, Sofaer (1984) adopted a combined action research and experimental design to test the impact of a ward-based educational programme on patient outcomes. Patients whose pain was managed by nurses, who had undergone the programme, experienced significantly less pain than those nursed by a control group who had not attended the course. Hart and Bond’s (1995b) typology, which was outlined in chapter three, was reviewed at this stage to explicitly consider the conflicts and tension of employing an RCT within a study which aimed to empower both patients and nurses in relation to pain management. The typology places experimental action research at the opposite end of the continuum to empowering approaches. The educative base of experimental action research focuses upon the researcher and is aimed at re-educating staff. Conversely, the empowering type is user and practitioner focused and seeks to raise consciousness to shift power balances. In this study, nurses were being encouraged to challenge their existing ideologies and the balance of power within pain management. Likewise, we wished to raise consciousness amongst patients about the use of self-generated strategies. Consciousness raising, rather than re-education, was thus deemed appropriate. Change interventions in the empowering approach rely upon exploring identified practice problems through understanding the context and processes of change. Action and critical discourse are emphasised. In experimental approaches, problems are tested through research, which may produce generalisable knowledge. As such, an RCT is a feasible method to employ within experimental action research. However, despite the survey employed in cycle one and the proposed RCT within cycle two, the overall design of the study leant towards an empowering, rather than experimental, approach. Rationale for the use of the RCT therefore needed to come from a different source. From the outset, the study emphasised collaboration and involvement. Cycle one lasted some 20 months and during this time participants were increasingly exposed to research, both through the study and the escalating emphasis upon evidence-based practice within health care. Resources were tight and staff did not wish to add to their workload unnecessarily. They therefore wanted to be assured that any new approach was evidence-based and consistent with an acute care perspective of working towards measurable outcomes. The rationale for using the RCT thus lay initially within the nurses’ view of the practice problem and research question, as something amenable to testing rather than “simply” describing. There was also a sense that an RCT would empower nurses in defending new praxis to medical staff. Comments during the workshops suggested that participants regarded the findings from an RCT as providing a more ethical basis for new praxis in comparison to qualitative findings. This was related to the view that they would be able “to prove“ to patients that “it works”. Finally, the point that action research is not constrained to any particular method, but can be flexible in drawing upon the most appropriate method for the research question, was acknowledged. Further discussion about the choice of method is presented in chapter six. The detail of the RCT is discussed in the following sections, commencing with identification of the research hypothesis. 4.5.2.1 The Research Hypothesis The research question - “what are the effects on postoperative pain of patients using their own pain relieving strategies?” - was transferred into a one-tailed hypothesis as follows: “Overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care, reduces postoperative pain intensity and distress, analgesic consumption, anxiety and length of stay, and enhances patient satisfaction with their postoperative pain management.” The outcomes to be measured were pain intensity and distress, analgesic consumption, anxiety, length of stay and patient satisfaction. The parameters of these measures are discussed in section 4.5.2.4. The independent variable, or intervention, was both the preoperative identification of patients self-generated strategies and the postoperative provision of active support to patients by the nursing staff. The manner through which the interventions were implemented is addressed in section 4.5.2.5. To secure a representative sample and reduce the effects of extraneous variables, all patients meeting the inclusion criteria (see section 4.5.2.3) were to be approached. Following consent and the collection of baseline data, subjects were to be randomly allocated to an experimental group who would receive the trial intervention, or to a control group who would receive the usual pain management offered on the Ward. Allocation was carried out using a random numbers table. Due to the nature of the intervention the nurses could not be blind to the treatment allocation. Because of the collaborative approach, whereby some nurses also acted as researchers, those collecting data were also aware of the treatment allocation. Patients were not told which group they were in, however it is feasible that they were able to identify their group. 4.5.2.2 Determining The Sample Size Power calculations were used to determine the sample size. Hillier and Gibbs (1996) emphasise that legitimate approaches to sample size calculations must be used if trials are to be ethical. Otherwise, time and resources are wasted and subjects may be exposed to harm without benefit. The purpose of carrying out calculations is to determine the smallest sample size that will detect statistically significant differences. However, Getliffe (1998) warns that statistical significance may be different from clinical significance. This point is also made by McQuay and Moore (1998) who advocate that health care researchers must take into account the practical potential of their findings. This issue is further considered in chapters five and seven. The primary comparison measure between the experimental and control groups was identified as the mean pain intensity scores of the two groups. As no previous RCTs have been undertaken on patients self-generated strategies in acute care settings, the literature on taught strategies was used as a basis for the calculations. Flaherty and Fitzpatrick’s (1978) study on the effectiveness of taught relaxation for postoperative pain used a 0 - 10 numerical scale to measure pain intensity. The experimental group pain intensity mean was 5.55 (SD 1.8) and that for the control group was 7.45 (SD 2.2). The difference in mean pain intensity between the two groups was thus 1.9. Taking the standard deviation (SD) as 2.5, and assuming 90% power and alpha = 0.05, 36 subjects were required for each group to detect a difference of at least 1.9 in the mean change in pain intensity. To allow for dropouts, it was deemed sensible to recruit 40 patients to each group. Thus a total sample of 80 patients were to be recruited and randomly allocated to either the control or experimental group. The power calculations are detailed in appendix seven. 4.5.2.3 Inclusion and Exclusion Criteria Patients scheduled for booked surgery on the Ward, who were aged 18 years of age and over and able and willing to give written, informed consent were to be offered the opportunity of inclusion in the study. Patients below the age of 18 years, emergency admissions, prisoners, those who were mentally incapacitated and anyone who was unable or unwilling to give written informed consent were to be excluded. 4.5.2.4 Data Collection Methods The data collection instruments used during the RCT are summarised in the table below, together with their primary purpose and time of completion. The instruments are included in appendix eight. The method, and rationale for choice, are discussed in the subsequent section. Table 3: Summary of the Data Collection Instruments Used During the RCT. Instrument Purpose Time of Completion  Preoperative Patient Questionnaire  To obtain baseline data on sample attributes, the outcome measure of anxiety and other potential intervening variables related to pain perception and strategy choice. Following consent, prior to randomisation. Preoperative Strategies Questionnaire  To identify the self-generated pain coping strategies of experimental group subjects. Following randomisation, prior to surgery. Self-report Pain Intensity and Distress Scales To obtain data on the outcome measures of pain intensity and pain distress. 12.00 - 14.00, day 1 - day of discharge. Analgesic and Antiemetic Consumption Record To obtain data on antiemetic use and the outcome measure of analgesic consumption. Daily, from day 0 post surgery to discharge. Postoperative Patient Questionnaire  To obtain data on the outcome measures of anxiety, length of stay and patient satisfaction.  The day prior to, or the morning of, discharge. Postoperative Strategies Questionnaire  To identify self-generated pain coping strategies used by patients in both groups during their hospital stay. The day prior to, or the morning of, discharge.  A Preoperative Patient Questionnaire was devised to record demographic data (such as age, gender, ethnicity and social class), a brief medical history, past pain experiences, current pain perceptions, and baseline data related to current pain intensity and distress ratings. The purpose of the questionnaire was to provide information on sample characteristics and the intervening variables potentially related to pain perception and strategy choice. This would enable the sample to be described and allow comparisons between the control and experimental groups to ensure that there were no obvious differences at baseline (Barker, 1992). The questionnaire also incorporated the “Short Form Spielberger State-Trait Anxiety Inventory” (STAI) which has recognised validity and reliability (Marteau and Bekker, 1992). Anxiety has been closely linked to pain perception (Craig, 1999) and was identified as one of the outcome measures. The inventory consists of six affective statements, each of which is scored by the respondent according to how they are feeling at the time. The scoring system is ranked as “not at all”; “somewhat”; “moderately”; or “very much”. By determining the preoperative scores, comparisons could be made between the two groups to establish that no significant differences existed prior to the experimental group being exposed to the intervention. The results would also enable comparisons against postoperative anxiety scores as an outcome measure. A Strategies Questionnaire was developed for administration to subjects in the experimental group preoperatively and to both groups prior to discharge. The questionnaire contained a list of strategies devised from the survey data, thus conferring internal validity. Respondents were expected to place a tick in column A, to identify all the pain relieving techniques that they had used in the past (or postoperatively); place a score in column B, to indicate how helpful they found each strategy in relieving pain; and on the preoperative questionnaire only, experimental subjects were asked to place a tick in column C against any strategy they wished to use in hospital following surgery. The relative simplicity of the tool was deliberate to enable patients to complete the questionnaire by themselves and allow experimental subjects to use it as a basis for discussion with the nurse regarding how they might use their strategies postoperatively. The purpose of the postoperative version was to ascertain experimental subjects’ strategy use following surgery and identify the degree to which control subjects utilised self-generated strategies during their hospital stay, if at all. Before the Strategies Questionnaire was developed, consideration was given to using either the Cognitive Control Strategy Inventory (CCSI), or Coping Strategies Questionnaire (CSQ). Both instruments were developed with patients in North America, with the former evolving from work with acute pain patients and the latter with chronic pain sufferers (see section 1.5 and 1.7). The CCSI relies upon the administrator matching taught strategies against high inventory scores. However, a study by Rokke and al’Absi’s (1992), which was described in section 1.7, concluded that offering choice appeared to be more effective. As the current study not only aimed to offer choice, but also to base this on patient generated preferences, the CCSI was deemed inappropriate. The CSQ has been used frequently in pain related research and has acknowledged validity and reliability. However, with the notable exception of the study by Kashikar-Zuck et al (1997) on pain coping during mammography (see section 1.7), the tool has predominantly been used in chronic pain research. Additional testing is required before the CSQ can be used with confidence for a study involving postoperative patients in the United Kingdom. The Burrows pain assessment chart was not considered for use in this study as the section on patients’ self generated strategies was limited and did not reflect the totality of strategies identified through the survey. A final issue was the collaborative nature of the study. There was a perception that, because the Strategies Questionnaire had been developed within the study, the nurses identified with it and felt a sense of ownership. Self-Report Pain Intensity and Distress Scales were adopted to obtain data on the identified outcome measures related to pain. Pain is a subjective phenomenon and, as reflected in McCaffery’s (1968) definition (see section 2.7), is best measured by patient self-report. In acute care, with its emphasis upon objective measurement, subjective reports are frequently regarded as lacking rigour (McQuay and Moore, 1998). However, McQuay and Moore (1998) point out that the sensitivity and consistency of pain self-reports have been found to be remarkably accurate. The most commonly used instruments designed to assess pain intensity are visual or verbal numerical analogue scales and categorical scales. The latter may be converted into numerical scales for analysis (e.g. none = 0, mild = 1, moderate = 3, severe = 4). They are useful tools for clinical practice, but lack sensitivity in research as respondents may be constrained by the categories. Visual analogues scales avoid the limitation of forcing scores into a particular response, as the individual is required to place an X on a 0 - 100 mm. line where 0 is no pain and 100 is unbearable pain (or similar wording). However, McGuire (1984) suggests that visual scales require considerable concentration to complete, which may be difficult postoperatively. Verbal numerical scales enable respondents to give a number between 0 (no pain) and 10 (unbearable pain). They avoid the degree of enforcement found in categorical scales while providing additional anchorage in comparison to visual analogues. Jensen et al (1986) found that verbal numerical scales present fewer problems for older patients than visual scales, which was deemed important in terms of the orthopaedic population involved in the current study. They are also widely acceptable, quick and easy to score. The self-report scales used in the study were of this type and designed to measure both pain intensity and pain distress. In earlier chapters, it was emphasised that pain is a multidimensional phenomenon incorporating physiological, sensory, affective, cognitive, behavioural and sociocultural dimensions (McGuire, 1992). The self-report scale adopted for this study measured just two dimensions, and for each of these, just one element. Thus pain intensity was measured in the sensory dimension and pain distress in the affective. Consideration was given to using the McGill Pain Questionnaire (MPQ) which, although unable to measure pain in all six dimensions, was designed to consider the three areas proposed by Melzack and Casey (1968) and Melzack (1973) (see section 1.4). However, McGuire (1984) records that the MPQ may take 20 - 30 minutes to administer and score. Although a short form is available, it was still felt to be too complex for self-report in a postoperative context and too time consuming to administer in a busy surgical setting. The relatively simple verbal numerical scale was preferred. The scales were to be completed preoperatively and then daily from the first postoperative day to the day of discharge. The intensity scale was anchored by “no pain” (0) and “unbearable pain” (10), while the distress scale was anchored with “not at all distressing” (0) and “extremely distressing” (10). It was anticipated that patients would keep their own sheets and fill in their scores between midday and 2 p.m. This is a relatively quiet period and by suggesting the scales were completed at lunchtime, the arrival of lunch would act as a reminder. Control for completing the scale lay with the patient, although a researcher would visit each afternoon to check data entries. The Analgesic and Antiemetic Consumption Record was designed to document the type of analgesic, the dose per 24 hours and the route of administration. Similar details were requested for antiemetics. The record was kept with the drug chart to enable nurses to complete it during drug-rounds. Conceding that this would not always be possible, the record would be checked and, where necessary up-dated by a researcher each afternoon. Analgesic consumption was regarded as an outcome measure on the assumption that the less pain experienced, the less analgesic would be required. Antiemetic consumption was included at the request of the Acute Pain Sister on the basis that opiates can cause postoperative nausea and vomiting (PONV). She felt that we needed to be able to assess whether any differences in analgesic consumption between the control and experimental group were linked to PONV. The Postoperative Patient Questionnaire was to be administered to all patient participants prior to discharge, either during late afternoon on their penultimate day or during the morning of their final day, depending upon the anticipated time of discharge. The questionnaire was designed to enable length of stay to be recorded, which according to the National Institute of Nursing Research (1994) is an infrequently measured outcome. Verbal numerical analogue scales were also devised to measure perceived control, with 0 anchored by “not at all in control” and 10 with “completely in control”. The item was included twice, once in relation to the extent the patient felt in control of their pain management and a second time regarding the extent to which they felt the staff were in control. Patient satisfaction with pain management was also to be measured on a verbal numerical scale with 0 anchored as “not at all satisfied” and 10 as “completely satisfied”. The benefit of using verbal numerical scales to consider perceived control and satisfaction was that respondents would already be used to completing these scales. The limitations of the approach are addressed in chapter five. 4.5.2.5 The Trial Protocol The proposal was submitted to the Local Research Ethics Committee and approved without amendment. The Ward consultants were contacted in writing and formal approval requested for permission to approach their patients. Each of the collaborative informants was sent a letter up-dating them on activity since the workshops and briefly summarising the study design. A uniform pocket sized booklet was devised by the collaborative group and sent to all participants. The original A4 copy is attached at appendix nine. The first section of the booklet detailed the trial design, nurses’ role, researchers’ role and the data collection instruments. The trial intervention was clearly stated and emphasis placed upon supporting experimental patients’ strategies, rather than teaching new strategies. The second section of the booklet detailed the way in which strategies might be supported in the hospital setting. Finally, the nurses were advised how to manage patients in the control group should the issue of self-generated strategy use arise. Posters summarising the trial protocol were placed in the Ward office and at the nurses station. All of the nurses were offered the opportunity to act as “co-action researchers” during the experiment. Although the offer was positively welcomed, none of the collaborative informants took it up initially. However, as the trial progressed, two D Grade nurses became increasingly involved. With the Ward Sister’s permission they joined the “co-action researchers” three months into the seven-month data collection period. Following completion of the trial, they continued as active members of the collaborative group. The Acute Pain Sister commenced maternity leave within a few weeks of the trial beginning but remained in touch by telephone and through meetings with myself. The Senior Staff Nurse who changed wards, moved before the RCT started. However, with her new Ward Sister’s consent she maintained some involvement, particularly at weekends. Until the three month point, when the two D Grade nurses joined the research team, there were three main co-action researchers: myself, the Senior Staff Nurse from the collaborative group who had participated in the survey and the Staff Nurse who had transferred from night duty. For brevity, the collaborative informants are referred to as “the nurses” and the co-action researchers, as “the researchers”. 4.5.2.6 The Procedure for Recruiting Patients Following admission, all patients meeting the inclusion criteria were handed the patient information sheet. During the late afternoon or early evening they were approached by one of the researchers to ascertain whether they were interested in participating. After any questions had been answered, those who remained interested were consented by the researcher. The patient information sheet and consent form are attached in appendix ten. The Preoperative Patient Questionnaire was completed and returned to the researcher in a sealed envelope. The patient was randomly allocated to the control or experimental group. The researcher explained the Pain Intensity and Distress Scales and handed the Strategies Questionnaire to experimental subjects for completion. The Analgesic and Antiemetic Consumption Record was placed in the nurses file at the bottom of the bed and the admitting nurse was informed of the patient’s consent and trial status. At approximately six to eight weeks into the study, it was decided that a coloured square of paper should be pinned to the notice board behind the patient’s bed to signify to the nurses and multidisciplinary team that the patient was in the trial. Two different colours were used so that all nurses were quickly aware of which patients should have their strategies actively supported. If patients or non-nurse members of the multidisciplinary team asked why the paper was different colours, they were to be told that “its just the paper we have available”. In the event, no one asked. The rationale for this addition is explained in section 4.5.3. During the evening of the admission day, the nurse checked that the patient had completed the Strategies Questionnaire and discussed with them ways of supporting any identified strategies postoperatively. Following surgery, the nurses offered this support. The researchers visited trial patients each afternoon, checked that the Pain Intensity and Distress Scales had been completed and, where necessary, up-dated the Analgesic and Antiemetic Consumption Record. The researcher also handed the Postoperative Patient Questionnaire to those being discharged, together with a sealed envelope for its return. On discharge, the nurse ensured that all relevant data collection instruments were returned to the patient’s study file. The researcher then erased the individual’s name from the research status list in the Ward office. I removed all completed files at my next visit. A small pilot study involving ten patients was undertaken prior to commencement of the trial. No alterations were made to the data collection instruments, although changes were made in the way we recorded analgesics delivered via Patient Controlled Analgesia Systems and epidurals. Originally, the length of the trial was anticipated as three and a half to four months. In the event, winter pressures leading to an increase in emergency admissions and cancellation of booked patients meant that data collection lasted for seven months. The difficulties caused by this are addressed in chapter six. The following section briefly outlines the analysis strategy for the RCT data. 4.5.2.7 The Analysis Strategy Following data input to SPSS (1994), frequency counts were undertaken to check the accuracy of data entry. The analysis strategy then incorporated three main stages. Each is listed below and then discussed in turn. Baseline comparison of the control and experimental groups to identify: sample characteristics whether intervening variables potentially related to pain perception and strategy choice varied between the two groups. Outcome measure analysis to test the hypothesis and determine whether there were any differences between control and experimental group outcomes. Additional analysis related to strategy use. The purpose of a randomised design is to ensure that variables are equally distributed between groups. It could, therefore, be argued that there is no need to establish whether any significant differences exist between control and experimental groups at baseline. However, several authors (Barker, 1992; Donnan, 1996; Hallett, 1997) believe that it is important to undertake checks so that the researcher can be satisfied that the randomisation process has worked. A decision was reached to proceed with baseline checks to facilitate insight into the characteristics of the study sample, to acknowledge the many factors which may influence pain perception (and potentially strategy choice) and to understand whether any differences existed between the groups in relation to the dependent variables considered to be potentially influential. Twenty variables were identified for analysis as shown in Table 4. Table 4: Variables Identified for Baseline Analysis. Age Time since diagnosis Pain at the momentGenderType of surgeryBaseline pain intensitySocial classPrevious surgeryBaseline pain distressMarital statusPrevious hospitalisationExpected pain on day 1Ethnic originPain in the pastIntensity of expected painConsultantMost painful past encounterSTAI baseline scoreDiagnosisScore of most painful past encounter  Univariate descriptive statistics were undertaken on each of the variables. The mode and variance were identified for 12 nominal and two ordinal variables and, where relevant, bar charts and pie charts were produced to enable visualisation of the data. Medians, means and standard deviations were established for six interval variables and histograms were produced. Bivariate analysis allows comparisons to be made between two variables. This strategy was employed to determine whether there were any significant differences between the control and experimental groups on entry to the study. Contingency tables were produced for nine of the nominal variables that had insufficient counts in certain cells to allow statistical testing. Visual analysis of this data was thus necessary. The remaining nominal and two ordinal variables were subjected to Chi-square tests. Two of the interval measures were normally distributed and subjected to parametric t-tests. Non-parametric Mann-Whitney U tests were undertaken on the remainder. Clustered bar charts and boxplots were produced as appropriate. The results of the baseline comparison are summarised in chapter five. In testing the hypothesis, eight main outcome variables, some with sub-variables, were identified for analysis. These are listed in Table 5. Table 5: Variables Identified for Hypothesis Testing. Pain intensity daily pain intensity scores mean postoperative pain intensity Pain distress daily pain distress scores mean postoperative pain distress Control of pain management patient in control nurse/doctor in control State Trait Anxiety  Analgesic consumption total morphine via PCA/epidural number of IM/oral opiates number of non-opiate doses number of NSAID doses Antiemetic consumption Length of stay Patient satisfaction  Frequency distributions and univariate descriptive statistics were conducted on each of the variables to gain insight into the characteristics of the variable and assess the distribution. Bivariate analysis was then carried out to establish whether or not there were any significant differences between the control and experimental groups. Ordinal data were subjected to Chi-square tests. For interval data with normal distributions, parametric t-tests were undertaken. Where the distribution was not normal, non-parametric Mann-Whitney U tests were used. The results are presented in chapter five. Strategies recorded by experimental subjects preoperatively were initially tabulated by frequency of use, degree to which they helped to relieve pain and the number of subjects who wished to use the strategy in hospital. A similar technique was used to describe the strategies used postoperatively by patients in both the control and experimental groups. Real values were altered to percentages to aid comparison between the groups. The frequency of use was calculated for each strategy and comparisons made within and between groups and against the survey data. Visual examination of the tables was used to inform the interpretation of results. The findings are included in chapter five. 4.5.2.8 Summary Having outlined and discussed the design and conduct of the randomised controlled trial, this chapter now moves on to consider the other research elements contained within Cycle Two. The following sections address the manner in which data was collected from the nurse participants during the RCT and from the nurses and researchers following the RCT. 4.5.3 The Staff Questionnaires and Researchers Critical Discussion The RCT provided data from the patients’ perspective but could not inform the process element of the study. To ascertain how patients’ strategies might be integrated into the nursing management of postoperative pain in the long term, and to assess the effects of the action research and critical theory design additional data was required from the nurses. Data were collected during and on completion of the RCT through field notes, questionnaires and a critical reflexive discussion. The questionnaires and discussion are described below and the field notes in section 4.5.4. 4.5.3.1 Data Collected During the RCT - The First Staff Questionnaire To reflect the participative and critical nature of the design, the researchers felt that it would be useful to engage nurses in critical discourse regarding the progress of the RCT in terms of both research and praxis. The discussion was also viewed as an opportunity to highlight any difficulties being experienced by the nurses. Two dates were organised eight weeks following commencement of the trial. A poster was placed on the office notice board informing the nurses of both the dates and main discussion questions. These were: 1. How do you feel the study is going? 2. What, if any, problems are you having? 3. How are you identifying experimental patients’ pain relieving strategies? 4. In what ways have you been able to support experimental patients’ strategies? In the event, neither discussion took place. On the first date the Ward was exceptionally busy and it was impossible to proceed. I abandoned my researcher role and assisted with care through supervising a junior student. On the second date, staff sickness meant that the Ward was short staffed. Although three nurses offered to participate, a researcher, who was to co-facilitate the discussion, could not be freed up and we both felt that removing nurses from the Ward would add stress and result in them having to work late. We decided to send out a questionnaire based upon the discussion questions. The questionnaire and covering letter may be found in appendix 11. In the covering letter, the researchers acknowledged that we were aware of the difficulties being experienced by undertaking a trial within the practice arena. The addition of placing coloured squares on the pin boards behind subjects’ beds was introduced at this stage in response to one of the problems being experienced. To enable honesty and encourage critical comment, we decided that the questionnaire would be anonymous. To encourage response from those who might not wish to complete the questionnaire and were not concerned with anonymity, we also suggested that nurses could give verbal responses directly to the Ward based researchers if they preferred. This dual strategy worked well; just three written responses were received, however, nearly all the nurses approached one or other researcher offering critical comments and suggestions. These were recorded and added to the data for analysis. Commentary upon the findings is incorporated in chapter six. 4.5.3.2 Data Collected on Completion of the RCT - The Second Staff Questionnaire Again to reflect the collaborative and critical reflexive nature of the design, the co-action researchers had originally planned brief one-to-one discussions with all consenting nurses following completion of the RCT. These would focus upon participants’ experiences during the RCT, their thoughts regarding the development of new praxis and their insights into personal development and alterations in ideology. For a number of reasons the idea had to be abandoned in favour of a further questionnaire. Firstly, the collaborative group felt that staff would prefer questionnaires. It was acknowledged that everyone was tired after the length of the RCT and that questionnaires would be perceived as less time consuming and therefore more likely to achieve a good response. Secondly, I had been asked to take on a part-time secondment at a different Trust, which would make access to the Ward nurses at times convenient to them, difficult. Thirdly, the Acute Pain Sister had just returned from maternity leave and thought it might be difficult for her to undertake half the interviews and be fully participative in the consequent qualitative data analysis. However, she offered her time to design, administer and analyse a questionnaire if this was thought to be a possible way forward. Following discussion it was acknowledged that this would be good development opportunity for the Acute Pain Sister. We agreed to proceed with a questionnaire. The questionnaire contained items on the nurses role in and perception of the RCT, their opinions regarding the helpfulness of both the researchers and booklet and three questions which aimed to gain some insight into the impact of the study regarding improvements in awareness, knowledge and practice in pain management. All questions required a closed response or a score. One further item asked respondents to record any other comments that they had. The nurses were sent a covering letter with an anonymous questionnaire (see appendix 12). Completed questionnaires were returned to an envelope in the Ward office and were analysed using descriptive statistics. No qualitative analysis was required as none of the respondents added comments in response to the open question. A poster of the results was placed in the Ward office and copies given to all collaborative group members. The findings are included in chapter six. 4.5.3.3 The Researchers’ Critical Discussion The researchers asked the Acute Pain Sister to facilitate a critical reflexive discussion on their participation in the RCT. In the event, reflection was broader than this and addressed participants involvement in the study as a whole. The discussion was open and unstructured and the facilitator used the skills she had developed in undertaking the focus groups to guide her in identifying the broad areas to address. The session was taped and analysed by the co-researchers following the strategy outlined in section 4.4.5.3. The findings are briefly mentioned in section 4.5.5 and are included in chapter six. The penultimate section of Cycle Two considers the approaches adopted to recording and analysing the field notes. 4.5.4 The Field Notes According to Porter (1996) field notes are: “observations written manually into a notebook by the researcher. They involve detailed descriptions of social situations and interactions that occur in the ‘field’ of research” (p.119). Silverman (1993) suggests that observing ordinary conversations and routines is important in developing an understanding of the field of research, while Sturt (1997) adds that observations enable the researcher to build a picture of everyday professional life. The field notes were also perceived as a way of recording the decision trail. Table six outlines the different contexts in which field notes were recorded during cycles one and two and specifies who undertook the note taking. Table 6: The Range of Field Notes Collected During the Study. Cycle Activity Field Notes Recorded by:  Telling the Situation As It Is 1. Observation of ward practices 2. Co-action researcher meetings 3. Collaborative group meetings 4. Staff focus groups 5. Patient survey log reflections 6. Staff workshops Author Co-action researchers Author Co-action researchers Researcher/clinic nurse daily Researchers daily and following each interview Co-action researchers Testing the Proposed Changes 1. General observations, co- action researcher and collaborative group meetings 2. Patient RCT log reflections 3. Researchers’ Critical Discussion 4. Feedback of Results Co-action researchers Researchers Co-action researcher (APS) Collaborative group Dey (1993) suggests that field notes can be recorded in either an unstructured manner as direct observations on what was spoken or what occurred, or against a pre-determined structure. Both types of notes were maintained in the study and are discussed below. 4.5.4.1 Structured Observations The observation of Ward practice, which commenced at the beginning of cycle one, was guided by Spradley’s (1980) framework. Table seven outlines the framework and the way it was applied in this study. Table 7: The Framework Used for Recording Observational Field Notes (Spradley, 1980). Framework Application to Study  1. Space: the physical place Layout of the Ward 2. Actor: the people involved Team nursing and the make-up of teams 3. Activity: set of related acts people do Interaction between the teams 4. Object: physical things that are present Equipment and resources for pain management 5. Act: single actions that people do Pain management actions  6. Event: related activities that people do General approach to care. 7. Time: the sequencing that occurs The way in which pain management was integrated in care 8. Goal: what people are trying to achieve Nurses ideologies of pain management 9. Feeling: the emotions expressed Nurses ideologies of multidisciplinary pain management.  Spradley’s (1980) categorisation aims to encourage thick description and was used to gain insight into the setting and current ideologies of the participants. Some of the findings were presented in section 4.4.2, the remainder are incorporated into chapter six. A second example of the use of structured field notes related to the records maintained during the co-action researcher meetings. These were guided by the literature on collaborative action research and, in particular, the work of Titchen and Binnie (1993a). Their publication on co-action researcher partnerships led us to consider the following areas: Our ideologies regarding health care, personal and professional development and pain management. Our understanding of the collaborative co-action researcher partnership, including individual strengths and weaknesses, perceptions of each other, our roles and the way in which the relationship developed. Our outsider versus insider status, including our roles, their effect and our relationship with the participants. Our interpretation of pain management on the Ward and within the broader context of the Trust. The costs and benefits to ourselves of participating and collaborating in the study. Finally, we added our own category related to continuing critical reflection on our interpretation and utilisation of action research and critical theory. The data were added to at each of our discussions and the interpretations are included in chapter six. The notes maintained during the collaborative group meetings also used a structured format. The meetings were balanced between “business” and critical discourse. The former were framed originally by the action research cycles and Carr and Kemmis’ (1986) stages and, latterly, by the research process and matters arising. The critical discourse elements were less structured and are dealt with in section 4.5.4.2. Finally, the researchers kept logs during the survey and RCT. These incorporated two elements: factual and reflexive. The latter is addressed in the section below. The factual element focused upon some of the issues identified in Spradley’s (1980) framework, such as the space, the actors, the activity and so on. Additional notes were also kept of the numbers of patients expected (either in clinic or for ward admission), the numbers who attended, reasons for non-approach and, finally, the number who consented to participate of those approached. This recording was prompted by the high DNA (did not attend) and cancellation rates which occurred during the survey and RCT respectively. 4.5.4.2 Unstructured Observation The term “unstructured” is used in this context to describe recordings which were either unstructured or were less bounded by structure than the notes gathered above. Three broad categories of data were identifiable: Data from the collaborative group meetings, staff workshops and researchers critical discussion formed one aspect related to the critical theory process and final research question. The data were highly interactive and critically reflexive in nature. The personal reflections carried out by those adopting researcher roles during the focus groups, survey and RCT formed a second category of data. The main purposes of these recordings were as a check on scientific rigour, as a reflexive account of personal development and action and, finally, as a critical discourse with self and others regarding changing ideologies. Data recorded during the feedback and refinement of results with staff and the collaborative group formed the last category. The emphasis here was upon reaction to the results and the identification of new praxis based upon raised consciousness, altering ideology and social action for practice transformation. Data collection was either by tape recordings or by jottings made in the field. Tapes were transcribed and added to the field note data. Jottings incorporated brief descriptions of observations, key words and phrases. These were transformed into field notes by using the jottings as a basis from which to expand the accounts. The process was usually completed on the same day, although occasionally it was the following day. Field notes were typed straight into the computer, with each new entry headed with the date, context and topic focus. As the study progressed, it was possible to begin creating files according to the context. Files were developed on co-action researcher and collaborative group interactions, collaborative informant data, the survey and the RCT. No attempt was made to construct files according to topics at this stage as doing so could have contaminated later analysis. However, hunches and initial interpretations were recorded in Italics. Crabtree and Miller’s (1992) reflection framework informed the strategy adopted for recording hunches. The appeal of this framework lay in its acknowledgement of critical reflexivity, which was consistent with the overall study design and its emphasis upon scientific rigour. Reflections were recorded on analysis, method, dilemmas and conflicts, points of clarification and the researchers’ frame of mind. Each of these is now briefly explained. Crabtree and Miller (1992) refer to reflection on analysis as: “the beginning of the dialogue you have with yourself” (p.59). Hunches are recorded and connections made. Confusion is acknowledged and the dialectical process valued. Reflections on method is similar to the process of recording decision trails, which was discussed in section 3.7.3. Thus the way in which choices were identified and the rationale for selection and action are recorded. Reflection on ethical dilemmas and conflicts acknowledges the reality of natural settings and the involvement of participants. From a critical theory perspective, bearing this element in mind overtly prompted reflection upon the barriers, tensions and conflicts surrounding pain management and the ethical issues arising as ideology was exposed and transformed and new praxis identified. The points of clarification category acted as a check on whether I understood what others, or I had observed. Sections were highlighted for further personal reflection, discussion with supervisors, or, at co-action researcher and collaborative group meetings. Finally, reflections on the researchers’ frames of mind were explicitly recorded during several stages of the study as highlighted above. They were also incorporated into other field notes. For instance, my reflections included commentary upon the way my personal and professional life influenced my ability to prompt critical discourse during collaborative group meetings. Thus, if I was tired and busy I was less likely to ask challenging questions. With the exception of the survey and RCT factual logs, each field note file was transferred into Atlas.ti. once it was complete. Atlas is a computer software package designed for analysing qualitative data (Scientific Software Development, 1998). The analysis strategy is outlined below. 4.5.4.3 Field Note Analysis Two approaches were proposed for analysing the field notes and other qualitative data. The first approach was to identify codes in advance. The second, and subsequent, approach was to re-read the data in an effort to determine themes grounded within it. The rationale for this dual approach was to enable critical theories to be developed in a focused manner, while at the same time ensuring that important issues were not missed. The identification of pre-codes was based on a conscious decision to seek evidence that illustrated the action research process and the critical theories generated through the nurse and patient data. The strategy was overtly focused upon answering the research question: “what are the affects of an action research and critical theory approach on the nurses involved and on the nursing management of postoperative pain?” Two areas of codes were generated, the first related to the characteristics of action research and the stages of planning, acting, observing and reflecting. The second focused upon pain management. The codes are shown in Table Eight. Table 8: The Codes Identified for Analysing the Field Notes Focus Code Illustrative Summary of Possible Sub-Codes  Action research: theory and practice Situation specific, practical problems Creating change in practice Collaboration Developing theory the context / the problems changes: practical and personal / barriers / support and leadership co-action researcher, collaborative group, collaborative informants experiences emancipation / reflexivity / ideology /praxis generic, epistemic and functional theory Pain managementAssessment Pain relieving strategies Multidisciplinary pharmacological / non-pharmacological taught vs. self-generatedCoding was undertaken and then each theme and sub-theme were defined and expanded. The data and resulting interpretations are presented in chapters six and seven. 4.5.5 Summary of Cycle Two: From Reflecting to Planning A considerable amount of data were produced during cycle two. In the reflecting stage, the collaborative group spent time discussing the best way of sharing results and prompting critical discourse with the collaborative informants. The findings from the first staff questionnaire were essentially fed straight back by the researchers, altering aspects of the trial protocol and enabling discussion of the issues and problems with the nurses. Some of the field note data also informed this process. The findings from the second questionnaire were easily presented in poster format and were attached to the office notice board for all to consider. The RCT analysis strategy and findings were discussed at co-action researcher and collaborative group meetings. Results were revealed as they were identified to promote reflexivity, enhance interpretation and enable critical consideration for praxis. Once data analysis was complete, detailed summary packs were given to each of the collaborative group members and discussed over several meetings. The Ward based collaborative group members offered to share and discuss the findings with the collaborative informants and to engage them in critical discourse regarding new praxis. I assisted in organising the process, working with the collaborative group and the off duty to ensure that all informants, whether part or full time, working nights or days, were involved in the discussions. The process took three weeks to complete and was carried out on a one-to-one or small group basis. By its conclusion, every member of the Ward team had been involved. The participants’ response to the results is included in chapter six. Their ideas for new praxis are incorporated with those of the collaborative group in section 4.6. Finally, the analysis and findings from the researchers’ critical discussions were shared and reflected upon at a collaborative group meeting. Some of the findings are presented here to illustrate the reflecting stage in cycle two and the basis for cycle three planning. The findings are fully incorporated into chapter six. On the whole, collaborative group members felt that participation had removed many of their fears regarding research. One stated that she now felt: “empowered to tackle something (further)”. Although another member felt that she needed a break, the general consensus was that progress should be maintained and practice transformed on the basis of the findings. Members also perceived participation in the collaborative group, and the opportunity to work as a team, as positive features that they did not want to loose. They felt increasingly able to challenge ideology and transform praxis, and were motivated to do so. The fluctuation of interest and motivation on the part of collaborative informants were noted and comment made that: “the interest was there when we were physically doing something”. Collaborative informants’ interest in the RCT results, their increasing critical reflexivity and their ideas for future praxis were positively regarded by the collaborative group. Overall, there was a feeling that forward movement should, and could, be maintained and that this did not need to rely upon my leadership. The reflections led into planning for cycle three. 4.6 Cycle Three: Creating Change In Practice - Implementation and Implications The Acute Pain Sister and collaborative group have taken responsibility for research and action in cycle three, reflecting the potential of action research to empower those involved. My role became critical support and gradual withdrawal. As a consequence, the following discussion is brief and focused upon the planning and acting stages. The full draft plan for this cycle is summarised in Table Nine. Table 9: The Third Cycle: Creating Change In Practice - Implementation and Implications. Stage Activity  Planning 1. Collaborative discussion on range of ideas for practice transformation Acting 1. Critical discussion on transformation process 2. Strategy for action agreed 3. Gradual withdrawal of author involvement Observing 1. Phased implementation of practice transformation 2. Collection of data to be carried out by nurses / Acute Pain Service 3. Recording of field notes Reflecting 1. Analysis of data to be carried out mainly by staff with support from audit department and author as consultant / link lecturer 2. Discussion of results within ward team and Acute Pain Service 3. Refinement of results.  A brief overview is now offered of the actions agreed to date. Following feedback of the RCT results to the Ward nurses and a collaborative group meeting held in the summer of 1999, several ideas were identified for further action and research. In essence these could be divided into actions related to practice transformation on the Ward, which were to be led by collaborative group members, and actions to be implemented through-out the Trust, which were to be led by the Acute Pain Sister. Both sets of actions incorporated further data collection through audit strategies. On the Ward, there was an additional intention to maintain critical reflexivity and consider other aspects for developing pain management that might lead into further research. The main issue for all participants was that the RCT findings were viewed as important and relevant for practice. To maintain momentum and promote evidence based practice, it was agreed that the strategies questionnaire should be implemented. Prior to implementation, the questionnaire was re-titled the Pain Strategies Questionnaire (PSQ) and altered to include a brief introduction. Patients were to be invited to complete two columns; one asking them to indicate all the strategies that they use when in pain and the second to highlight those they wished to use in hospital. The author took responsibility for the alterations and the Ward team for implementation. The PSQ may be found in appendix 13. The Questionnaire was also to be implemented on the ward to which one of the collaborative group members had moved. Including this ward meant we were responding to the criterion that our colleague had continued to be involved on the basis that her new ward would benefit. Equally, this second ward could act as a trial for wider implementation in the hospital. This was particularly important as, given the collaborative action research nature of the study, we were unaware of how other teams might react to the idea of identifying and supporting patients’ self-generated strategies. The Ward-based collaborative group members took responsibility for this action and were supported by the Acute Pain Sister. The Nurses Research and Strategy Guide was to be revised so that it was suitable for patients. Following agreement by the collaborative group, the revision was discussed at a Trust’s Pain Link Nurse meeting, which is chaired by the Acute Pain Sister. The revision was then incorporated into the Acute Pain Service’s Paediatric Pain Management Booklet, which was about to be piloted. Following the pilot, the intention is to develop a similar booklet for adults. In the meantime, nurses on the study Ward planned to continue using their own guides. Initially, it was thought that the PSQ and booklet would be handed to patients on admission. However, the orthopaedic clinic nurses informed the Acute Pain Sister that they would be interested in handing out the information when patients attended for pre-admission assessment. The advantage would be that booked patients would have a chance to read the booklet and complete the PSQ prior to admission. Emergency patients would still need to be provided with the pack by ward staff and it was suggested that this should occur within 24 hours of admission whenever possible. The Acute Pain Sister agreed to maintain liaison with the clinic nurses regarding the introduction of the pack. The collaborative group regarded internal dissemination as an important element in promoting evidence-based practice within the Trust and in easing practice transformation across the Trust. It was recognised that presentations would need to be arranged with a variety of nurses, doctors and other members of the multidisciplinary team. In addition a summary of the survey and RCT results was prepared for distribution. Although collaborative group members were invited to participate in both these activities, the main responsibility fell to the co-action researchers. The final agreement was that the collaborative group meetings should be maintained to provide a focus for praxis in pain management. The author suggested that she should stand down as chair and a practitioner was appointed. Further details regarding practice transformation are included in chapters six and seven. 4.7 Conclusion The variety of methods adopted within this study reflects the nature of practice problems in acute care settings. The ideology of these settings is one that values measurable outcomes, hence the survey and RCT. The twin processes of patient-centred care and nurse empowerment required different approaches to data collection. The next two chapters present the results and related discussion. Chapter five focuses upon the patients’ perspective and chapter six upon the nurses. CHAPTER FIVE: SELF-GENERATED PAIN COPING STRATEGIES: ANALYSIS OF THE PATIENT DATA 5.1 Introduction Chapter four described the three action research cycles and methods employed to collect data from patients and staff. In this chapter, the results of the patient survey and randomised controlled trial are presented, together with relevant excerpts from the field notes. To enhance clarity in the process of theory building, discussion of the results is included within the chapter, with wider issues being addressed in chapter seven. 5.2 The Survey Findings As previously mentioned, the survey was designed to address the question “what pain relieving strategies are used by patients in acute pain?” Specifically, data analysis aimed to identify: The characteristics of the sample The type of strategies used The frequency of use The range of use Strategy themes The perceived effectiveness of strategies Whether any differences exist between the types of strategies used for “everyday” pain and those used for more “serious” acute pain Patients feelings regarding the use of strategies postoperatively. 5.2.1 The Sample Table 10 illustrates the number of patients recruited during the four months of data collection. Table 10: Summary of Reduction From Population To Sample Category  Number New patients who met the inclusion criteria June - September 1996 401 Patients who attended for their appointment * 338 Patients who consented to participate # 220 Patients who completed the questionnaire 200  * The DNA (did not attend) rate for the first six weeks of the survey was approximately 11%, however, from mid-July to mid-September this figure increased to some 22% of the total numbers booked. This reflects national trends which show a rise in DNA rates during school holidays (Aylesbury Vale Community Health Council, 1995/6). # Of the 338 patients, who were approached, 35% declined to be involved. Of those who consented, two were later withdrawn by the researchers as one was confused and thus deemed unable to give informed consent and the other was under 18 years of age. A further four actively withdrew part way through completing the questionnaire and fourteen did not return their questionnaires. The main reason for this was a shorter waiting time than originally anticipated, resulting in individuals wishing to leave the hospital prior to completing the questionnaire. These patients, together with some 15 others, were given the opportunity to return their questionnaires in stamped addressed envelopes. In total, 200 patients completed the questionnaire. All patients were given the opportunity to participate in interviews until the required number had been undertaken. Twenty-five patients originally consented to take part, however, quiet, private rooms were only available on 20 of these occasions. The remaining five were thanked for their interest. 5.2.2 Questionnaire Sample Characteristics The characteristics of the questionnaire sample are detailed in Tables 11 and 12. Table 11 outlines the demographic profile of the sample. Table 11: The Demographic Profile of the Survey Sample. Characteristic  Statistic (n = 200)  CommentsAge: range mean standard deviation Gender: male female Marital Status: married / stable relationship single divorced widowed separated Social Class: I II III IV V not classified * Ethnic Origin white - UK other Caucasian Asian black Caribbean Chinese Religion Christian Moslem other/none Clinic: general surgery orthopaedic / rheumatology 18 - 89 45 years 1.58 57.5% 42.5% 72.0% 10.5% 8.5% 5.5% 3.5% 7.8% 16.2% 27.3% 5.0% 2.8% 40.9% 93.5% 3.5% 1.5% 1.0% 0.5% 86.0% 10.0% 4.0% 85.4% 14.6% Based on the Registrar General’s Social Class by Occupation Classification system. Not classified 21% retired; 13% housewives; 6.9% HM forces, students, unemployed.* Ethnic origin is representative of the local population census data. * The data collection instrument failed to ask those who had retired to record their previous occupation. Consequently, the social class data are incomplete. Nevertheless the percentage distribution reflects that of the local population. The clinical characteristics, which could potentially influence pain perception and strategy choice, are highlighted in Table 12. Table 12: The Clinical Profile of the Survey Sample. Characteristic  Result (n = 200) Diagnosis: varicose veins arthritis hernia prostatitis cholecystitis abdominal pain (unspecified) haemorrhoids urinary problems cancer other not yet known Length of time since diagnosis: 0 - 3 months (Acute) 4 months or longer (Chronic) Previous surgery: yes no Previous illness: yes no Previous most painful experience: surgery trauma abdominal pain arthritis back pain labour infection other Score of previous most painful experience range mean SD Expect pain after surgery: yes no unsure Intensity of pain expected on the day after surgery range mean SD 11.5% 10.0% 8.5% 8.0% 7.5% 5.5% 5.0% 5.0% 5.0% 20.0% 14.0% 42.5% 57.5% 70.7% 29.3% 58.6% 41.4% 16.0% 14.0% 13.0% 11.0% 11.0% 10.0% 7.0% 18.0% 0 - 10 7.00 2.40 61.5% 7.5% 31.0% 0 - 10 4.42 2.56 5.2.3 Pain Relieving Strategies Respondents were asked to record what they do to relieve pain, what their family and friends do that helps to relieve the individual’s pain and what sort of things they think nurses and doctors might do to help relieve their pain. The results are presented in the following three sections. 5.2.3.1 Patients’ Self-Generated Strategies Of the 200 subjects surveyed, 2.5% (n = 5) omitted to respond to the item asking them to identify the things they did to help relieve pain. A further 8.5% (n = 17) stated that they did nothing. The remaining 89% (n = 178) recorded a total of 368 self-generated strategies. The maximum number used by any one individual was eight strategies. The majority recorded between one and four. In total, 43 different strategies were recorded. These were categorised into 10 themes as shown in Table 13. Table 13: Classification of Self-Generated Pain Relieving Strategies. Theme  Examples of Strategies Within The Theme Analgesics Behavioural Resting alone Cognitive control Mobilising Positioning Seeking treatment Cognitive and expressive reassurance General help with activities Miscellaneous  All types of opiate and non-opiate analgesics, together with NSAIDs Distraction, relaxation, breathing exercises, imagery, music, massage, warmth, cold Lying down, trying to sleep, peace and quiet, being left alone Mind over matter, positive thinking, grin and bear it Walking, moving about, exercise Changing position, supporting the painful area, immobilisation Seeking medical attention to treat the cause or the pain, advice, information Talking about the pain, physical contact, confidence in someone else, verbal reassurance, expressive touch Help with hygiene needs, housework, shopping Humour, praying, tea, alcohol, cigarettes, being irritable, equipment aids, “complementary therapies”The frequency with which respondents reported self-generated strategies is outlined in the table below. The themes are listed in their ranked order. Table 14: Frequency of Patients Self-Generated Strategies by Theme and in Ranked Order. Strategy Theme  Number of Times the Theme was Recorded  Percentage of Total Number of Themes Recorded Behavioural Analgesics Resting alone Cognitive control Mobilising Positioning Seeking treatment Cognitive and expressive reassurance Miscellaneous  128 90 47 36 20 16 8 3 20  35% 25% 13% 10% 5% 4% 2% 1% 6%  Total  368 100% The most frequently occurring category was the behavioural theme, followed by analgesics. Resting alone and cognitive control were also used regularly. 5.2.3.2 Family and Friends Generated Strategies Just under 50% (n = 99) of respondents recorded strategies used by their family and friends to help relieve their pain. The majority mentioned one strategy, although a few recorded as many as three. A total of 143 strategies were identified overall. Table 15 shows the frequency and ranked order for each theme. Table 15: Frequency of Family and Friends Generated Strategies by Theme and in Ranked Order. Strategy Theme  Number of Times the Theme was Recorded  Percentage of Total Number of Themes Recorded Cognitive and expressive reassurance Behavioural General Help Resting alone Analgesics Seeking treatment Positioning Miscellaneous  34 29 28 20 14 5 4 9  24% 20% 20% 14% 10% 3% 3% 6%  Total  143 100% Cognitive and expressive reassurance were the most common strategies used by family and friends. Behavioural strategies and general help were important, while allowing the person to rest alone and providing them with analgesics were also mentioned. The theme “general help” appeared only in the family and friends category, while neither “mobilising” nor “cognitive control” were recorded in this section. 5.2.3.3 Patients’ Perceptions of Practitioner Generated Strategies In response to the question “what sort of things do you think the nurses and doctors might do to help relieve your pain” 78% (n = 156) of subjects recorded strategies. The majority listed just one strategy although a significant number registered two and a few recorded up to five. The total number of strategies identified as being potentially offered by practitioners was 233. Not surprisingly, analgesics were the most frequently mentioned strategy and 40% (n = 62) of those responding to this item listed them as the sole practitioner strategy. However, 60% (n = 94) anticipated that doctors and nurses would use additional non-pharmacological strategies to treat or relieve their pain, while 30% (n = 47) of respondents did not list analgesics as a practitioner generated strategy at all. The frequency with which practitioner generated strategies were recorded and their ranked order are shown in Table 16. Table 16: Frequency of Practitioner Generated Strategies by Theme and in Ranked Order. Strategy Theme  Number of Times the Theme was Recorded  Percentage of Total Number of Themes Recorded Analgesics Seeking treatment Cognitive and expressive reassurance Behavioural Positioning Resting alone Mobilising Cognitive control Miscellaneous  109 53 45 13 4 4 2 1 2  47.0% 23.0% 19.0% 5.6% 1.7% 1.7% 0.8% 0.4% 0.8%  Total  233 100% The table shows that patients feel there will be an emphasis upon analgesics, treatment and reassurance. 5.2.3.4 Summary The final table in the series on strategy use amalgamates the results of the patient, family/friends and practitioner generated strategies. Table 17: Overall Frequency of Strategy Use by Theme and in Ranked Order. Strategy Theme  Number of Times the Theme was Recorded  Percentage of Total Number of Themes Recorded Analgesics Behavioural Cognitive and expressive reassurance Resting alone Seeking treatment Cognitive control General help Positioning Mobilising Miscellaneous  213 170 82 71 66 37 28 24 22 31  28.6% 22.8% 11.0% 9.5% 8.9% 5.0% 3.8% 3.2% 3.0% 4.2%  Total  744 100% The results indicate that patients use a range of self-generated strategies for pain relief on an everyday basis, that they receive assistance from family and friends in their attempt to relieve pain and that they anticipate that health care practitioners will employ a variety of pain relieving strategies and treatments. Forty-nine separate strategies were identified and collapsed into ten themes that were generated from the data rather than the literature. The most frequently used self-generated strategies were linked to the behavioural theme, those most often employed by family and friends were from the cognitive and expressive reassurance theme, while analgesics were the most frequently recorded theme for practitioners. When the frequencies were combined, analgesics proved the most frequently mentioned strategy, closely followed by behavioural techniques and then cognitive and expressive reassurance. The following section considers the perceived effectiveness of self-generated strategies in relieving pain; whether there were any differences between strategies employed for everyday pain and those used for more serious acute pain and respondents’ opinions regarding the use of pain relieving strategies postoperatively. 5.2.4 The Efficacy and Use of Pain Relieving Strategies The data regarding perceived efficacy and use of strategies were obtained from the interviewees. To locate this data within the overall survey sample, the demographic and clinical characteristics of the sub-sample were tabulated and are attached in appendix 14. Eighteen (90%) interviewees recorded 32 self-generated strategies and 21 practitioner generated strategies on their questionnaires that were completed prior to the interview. Eleven (55%) of the sample listed 17 family and friends generated strategies. The ranges of strategies, by theme, are shown in Appendix 15. In comparing the range with those of the overall questionnaire sample, it was possible to determine that, with one exception, strategy use was broadly equivalent. The exception was the cognitive and expressive reassurance theme which interviewees appeared to record more frequently. The numbers are small and care must therefore be taken in drawing any conclusions from the frequency counts. However, they are included to highlight the direction of the data. Appendix 16 contains an interview transcript that provides a typical example of the interview schedule and responses. 5.2.4.1 Perceived Efficacy of Self-Generated Strategies Interviewees varied in their perceptions of how effective their strategies were. Table 18 contains some of the comments made regarding perceived efficacy, including both positive and negative comments. Table 18: Interviewees Comments Regarding the Efficacy of Strategies for Relieving Pain. Strategy  EffectivenessLocal anaesthetic NSAIDs Non-opioid oral analgesic It numbed the pain. It worked quite well after a while. They were no good whatsoever. It helps, but it’s not that successful. It immediately cleared. I worry about the side effects It takes the edge off and allows you to carry on with activities. It took the pain away and I was just left with discomfort BreathingIt takes your mind off the pain, very relaxing and quite successful. DistractionIt masks the pain and relaxes the mind. It absorbs the pain, but then it comes back. It’s not that successful. It takes your mind off it. Whether it relieves pain, I don’t know. It’s quite successful at diverting attention though. I immerse myself and don’t notice the pain because I’m concentrating. The pain gradually goes. It’s very successful RelaxationIt relaxes body and mind. It helps to disassociate the pain. It’s really good. Instrumental touchVery successful. WarmthIt’s comforting. The pain comes back soon, but it does help at the time. I thought a salt bath was healing. It made it draw a bit. It helped. I never felt a thing once I got in the bath. The pain disappears and I don’t need anadin. ImageryIt stimulates the brain and takes pain away fairly quickly. ColdNot much, but distracting. It took the throbbing away and cooled my ankle down. It was a great relief. Within a few more days there was no more pain. Cognitive & expressive reassuranceIt’s reassuring and it’s some relief. It makes me feel better, very successful. Resting alone: It’s relaxing and you can be alone with your thoughts. It takes your mind off it and it seems to work. Took the pain away completely. Cognitive control: It relieves some pain and I feel better for it. It’s about breaking through the pain barrier and proving physical capability. You just deal with it. It’s about being in control. Positioning: Not particularly helpful. Oh, a great deal, there was no pain when it was still. It helps. Mobilising - especially going for a walk It clears my head, but it doesn’t always work. Helps relieve it, generally feel better afterwards. It’s about not giving in.  Many of the strategies discussed by interviewees were either learnt during childhood or had been tried on the recommendation of friends or health care practitioners. Practitioners were particularly likely to suggest analgesics, warm baths, resting and medical attention, while family and friends tended to advise the use of ice packs, heat pads and some of the more unusual strategies in the miscellaneous theme. In adulthood, interviewees had also developed strategies such as cognitive control and mobilising. Positive self-talking and going for walks were frequent examples. These strategies were perceived as promoting distraction and allowing the individual to take their mind off the pain. As one interviewee commented: (I) “take a long walk. The fresh air helps clear my head and the aches and pains just disappear.” The extent to which a strategy was regarded as positively impacting upon pain was related to whether it was being used for everyday aches and pains or for more serious acute pain. The next section considers these differences. 5.2.4.2 Differences In Strategy Use According to Pain Intensity The examples given of everyday aches and pains included headaches, cuts, dental problems, muscular injuries and arthritis. All strategy themes were mentioned with the exception of positioning. Examples of serious pain included infections, trauma, postoperative, abdominal and labour pain. Again all strategy themes were referred to with the exception of mobilising. Differences were distinguished in relation to pain intensity and strategy use in terms of the number of strategies employed. The average number used for everyday pain was 3.3, compared to 4.5 for acute pain. It seemed that the more intense the pain experienced, the greater the number of strategies employed. Differences were also noted in relation to the type of strategies used as can be seen in Table 19. Table 19: Number of Times Strategy Theme Recorded for Everyday Versus Acute Pain. Theme  Everyday Pain  Acute Pain  Analgesics Cognitive and expressive reassurance Cognitive control Mobilising Behavioural Resting alone Seeking treatment Positioning Miscellaneous General help with activities  14 4 8 9 19 5 1 0 4 0  9 1 1 0 37 10 5 14 4 0  Those experiencing everyday aches and pains were more likely to use cognitive and expressive reassurance, cognitive control and mobilising. Conversely, those in acute pain tended to increase the number of behavioural and resting alone strategies, treatment orientated strategies and positioning. There was some evidence that the greater the fit between the strategy themes and the type of pain, the more effective the strategies were perceived to be. Of note, analgesics were mentioned more frequently for everyday pain. The explanation for this apparent anomaly lies in the degree to which over-the-counter medicines are used in a self-generated manner for everyday pain, while medical treatment is sought for acute pain and prescribing thus lies in the hands of the practitioner. This issue is addressed further in the following section. 5.2.4.3 Strategy Use in Hospital Interviewees suggested that they would like to see a range of strategies used in hospital, some being self-generated and others practitioner generated. The general help and miscellaneous themes were not included in the list. The number of times the remaining themes were mentioned is shown in Table 20. Table 20: Strategy Themes Identified, by Interviewees , For Use in Hospital. Theme  Self-Generated  Practitioner Generated  Analgesics Cognitive and expressive reassurance Cognitive control Mobilising Behavioural Resting alone Seeking treatment Positioning 1 0 1 5 20 2 2 0 11 7 0 2 5 2 6 2  As can be seen in the table, analgesics were identified as a method respondents expected and hoped practitioners would use in hospital postoperatively. Interestingly, although cognitive and expressive reassurance was not identified as frequently for acute as for everyday pain (see Table 19), there was an expectation that it would be provided by practitioners in the hospital setting. For example interviewees commented: “That’s part of the trust between patient and nurse that your nurse is there to know more about what is happening.” “They gave me some painkillers and I just took them because they said you need to do that sort of thing.” “Someone who is more experienced would say this would be good, would you like to try it?” “I would just like to be guided by the staff.” These quotations highlight the importance for some individuals of having confidence in practitioners. Continuing to use behavioural strategies was popular and, in particular, the opportunity to use warm baths. However, many interviewees perceived that this would not be allowed in hospital, stating for example: “I would be relieved to do that...I don’t think I’m going to be allowed to have a bath though”. These and other comments regarding what would and would not be allowed reflect Copp’s (1974) findings that patients believe health care practitioners might disapprove of self-generated strategies being used in hospital. During discussions on the use of strategies in hospital, an additional theme was generated from the interview data. Labelled “being empowered”, issues relating to this theme are discussed in the next section. 5.2.4.4 Being Empowered It was evident from the data that some interviewees wanted to be treated as individuals, given choice and control and through this, to be actively engaged in their pain management. Conversely, others appeared to prefer practitioners to adopt a nurturing role in which power and control remained in the hands of the practitioner. Phrases such as “take care”, “attend to you” and “stop me being concerned” were typical of this latter group. As one interviewee stated: “If professionals decide it’s the way to go, it’s the way to go.” A second added: “I don’t like pain very much. They said they’d keep it under control and I’m sure they will.” These statements have some cross-over with those elicited in relation to the cognitive and expressive reassurance theme. It was apparent that those who mentioned this theme frequently were also likely to prefer a more passive role in their pain management. Respondents seeking empowerment used phrases like “listen to exactly what you say”, “treat the individual as a person” and “control is important to me”. The following quotations offer further insight into the importance of being empowered for this group: “I dislike taking painkillers and think my pain threshold is high. So when in hospital if I ask for strong analgesics it is humiliating to be given paracetamol. I think I should be more in control of my body.” “Pain is a personal thing that others find difficult to understand.” “I want choice and to use my own strategies.” “I would like to be given a full explanation...to be told what sort of pain I’m likely to have and how long its going to last for. If I know what’s coming I can take time to think about how I’m going to deal with it. ...I would like some control.” The tone and body language of these individuals was frequently assertive and there was a sense that they perceived hospitalisation as synonymous with a loss of control. This extended to their expectation of pain following surgery, which was regarded as “frightening”. Typically, comments implied that there was a lack of trust in practitioners’ ability to deal effectively with the individual’s pain and that they would let it get out of control. For example, one interviewee stated: “The trouble with pain is once you let it get a grip, it’s very hard to get rid of.” Conversely, those who appeared to prefer a nurturing approach were relatively relaxed, anticipating that the nurses and doctors would know what they were doing and would do it well. In chapter two, the point was made that true empowerment must allow for individuals to exercise their right to engage in care management, or, if they prefer, to leave decision making in the hands of the practitioner. Those interviewees who appeared to express a preference for nurturing, could be regarded as advisors in the decision process, however, it is evident that others wished to be acknowledged as equal partners (Elliott and Turrell, 1996). The role of the nurse is to determine the degree to which patients wish to participate, as it is only through doing so that care can be patient-centred. This issue is addressed further in chapter seven. 5.2.5 Summary The survey enabled respondents’ beliefs and practices concerning pain relieving strategies to be exposed. Reflecting McCaffery’s (1968) definition, there was evidence that pain was viewed as a highly personal experience demanding a range of individualised strategies. The strategies adopted differed according to the pain intensity and respondents beliefs about pain and its relief. A wide range of strategies was used enabling a multidimensional approach to pain management. Strategies were elicited that reflected the physiological dimension through seeking treatment and the biochemical impact of analgesics. The sensory component was best typified by the apparent practice of respondents engaging in different strategies according to the intensity of pain. The affective and cognitive dimensions could be respectively linked to the strategy themes cognitive and expressive reassurance and cognitive control. The behavioural dimension could be linked to all of the themes, however, the behavioural, mobilising, positioning and resting alone themes are specifically mentioned here. Finally, the impact of the sociocultural dimension could be seen through interviewees’ comments regarding learnt strategies and the influence of family, friends, practitioners and life experience on strategy choice and refinement. As highly interactive themes, cognitive and expressive reassurance and general help with activities may also be linked with the sociocultural dimension. Several contradictions were noted in the data and discussion on these may be found later in the chapter. Of note is the evidence that individuals use a range of pain relieving strategies and yet believe that their strategies are either not transferable to, or would be frowned upon in, the hospital setting. This was an important issue in constructing the intervention for the RCT. The results from the RCT are presented in the next section. 5.3 Findings From The Randomised Controlled Trial The RCT set out to test the hypothesis: “Overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care, reduces postoperative pain intensity and distress, analgesic consumption, anxiety and length of stay, and enhances patient satisfaction with their postoperative pain management.” The design of the RCT and the three main analysis questions were previously identified in chapter four. The analysis questions are summarised below: Were the control and experimental group comparable at baseline? Did incorporating patients spontaneous, self-generated strategies into the plan of care alter : pain intensity pain distress analgesic consumption antiemetic consumption anxiety length of stay patient satisfaction control What was the difference, if any, in strategy use between the experimental and control groups The following two sections provide a description of the sample and the sample characteristics, together with discussion related to the first analysis question. 5.3.1 The Sample The RCT lasted for 31 weeks although data were only collected during 24 of these weeks. Two weeks were lost because no researchers were available. The remaining five, were at Christmas and Easter when there were no booked admissions. The data collection period was twice the length anticipated. On average, the Ward had 16 booked admissions per week, totalling some 375 patients over the period. However, actual admissions were considerably lower than this due to pressure on beds. Approximately 45% of patients were either cancelled or admitted to other wards. Table 21 provides a breakdown of the number who consented to participate in the trial from those admitted. Table 21: Summary of Reduction From Admission to Consent. Category  Number Percentage Patients admitted Patients who did not meet the inclusion criteria Patients who were not approached * Patients who declined to be involved Patients who consented to participate  206 33 33 43 97 100% 16% 16% 21% 47% * A few of the patients, who were not approached, were deemed by the nurses to be unsuitable on the basis of their clinical condition. The majority, however, were missed because researchers were either off sick or the Ward was too busy. In total, 97 patients consented to participate in the trial. However, usable data were only collected on 80 of these subjects. The reasons for withdrawal are detailed in Table 22. Table 22: Reasons for Withdrawal of Subjects from RCT. Category  Number Ratio by Group  Significant amount of missing data Operation cancelled Transferred to ITU/other wards on the day of surgery Withdrawn by researchers as confused/too ill to participate Withdrew on own request Died during surgery 7 3 3 2 1 1 4C : 3E * 2C : 1E 3C : 0E 1C : 1E 1C : 0E 0C : 1E Total 17 11C : 6E* C = control, E = experimental The data on these subjects were excluded from the analysis. 5.3.2 Sample Characteristics The demographic profile of the RCT sample is detailed in Table 23 below. Table 23: The Demographic Profile of the RCT Sample. Characteristic  Overall Sample (n = 80)  Control Group (n = 40) Experimental Group (n = 40) Age: range mean standard deviation Gender: male female Marital Status: married / stable relationship widowed single divorced / separated Social Class: I II III IV V other Ethnic Origin: white - UK other  18 - 89 59.40 16.27 43 37 56 10 9 6 6 22 22 7 1 22 78 2  21 - 89 60.25 17.35 24 16 32 2 4 2 4 9 16 2 0 9 38 2 18 - 86 58.55 15.28 19 21 24 8 5 3 2 13 6 5 1 14 40 0 With a mean age of 60, the sample was skewed towards an older population and was somewhat older than the survey mean of 45 years. The skew was not unanticipated given the orthopaedic setting. The data were, however, normally distributed around the mean. The experimental group had a slightly lower mean age than the control group, however, there was no significant difference revealed by a t-test (t = .47; p = 0.64; CI - 5.58, 8.98). Despite the experimental group containing more female subjects than the control group, a Chi-square test showed no significant difference between the groups according to gender (X² = 1.258; df = 1; p = 0.26). Given the relatively low numbers in several of the marital status categories, data were re-catagorised into married/stable relationship and “other”. A Chi-square test showed that no significant difference existed (X² = 2.286; df = 1; p = 0.15) between the control and experimental groups. Again with low numbers for each social class, significance tests were conducted after amalgamating classes I and II; III non-manual and manual; and IV and V. A Chi-square test indicated that a statistically significant proportion of the experimental group were located in the lower social classes (X² = 7.415; df = 2; p = 0.025). However, 28% (n = 22) of the overall sample and 33% (n = 13) of the experimental subjects did not record a social class which may explain the apparent differences. Descriptive statistics on ethnic origin were sufficient to demonstrate that the control and experimental groups were similar for this variable. The clinical characteristics of the sample are detailed in Table 24 overleaf. As can be seen from the table, the two groups were comparable in terms of diagnosis. Length of time since diagnosis could potentially influence the use of pain relieving strategies, with those experiencing pain for longer periods employing a wider range of strategies. Thus, a Chi-square test was undertaken to assess the significance of any difference between the groups in relation to the tabulated categories. The results were X² = 4.908; df = 4; p = 0.30 indicating that no significant difference existed between the groups in respect of length of time since diagnosis. Descriptive statistics on the type of surgery planned, previous stay in hospital and previous surgery were sufficient to demonstrate that the control and experimental group were similar on these variables. Table 24: The Clinical Profile of the RCT Sample. Characteristic  Overall Sample (n = 80)  Control Group (n = 40)  Experimental Group (n = 40) Diagnosis: arthritis (unspecified) arthritis (osteo, ankylosing spondilitis, RA) back pain Hallux Vulgus torn ligaments other Length of time since diagnosis: < 6 months 6 - 12 months 1+ - 5 years 5+ - < 10 years > 10 years not given Type of surgery: THR TKR other joint surgery arthroscopy ligament repair back surgery other Previous stay in hospital: yes no not stated Previous surgery: yes no not stated  35 20 12 4 3 6 14 17 23 7 13 6 31 18 9 2 4 12 4 71 8 1 66 13 1  18 10 7 2 0 3 4 8 15 2 7 4 17 7 6 0 1 7 2 37 3 0 34 6 0  17 10 5 2 3 3 10 9 8 5 6 2 14 11 3 2 3 5 2 34 5 1 32 7 1  Table 25, overleaf, shows the pain profile of the sample upon admission to hospital. Table 25: The Pain Profile of the RCT Sample. Characteristic  Overall Sample (n = 80)  Control Group (n = 40) Experimental Group (n = 40)Pain in the past: yes no not stated Type of most painful past experience: surgery arthritis trauma back pain abdominal pain mobilising other not given Score of most painful past experience: range mean SD Pain at the moment: yes no not stated Baseline pain intensity: range mean SD Baseline pain distress: range mean SD Expect pain day after surgery: yes no do not know Intensity of pain expected: range mean SD STAI baseline scores (Anxiety) range mean SD 78 1 1 17 13 11 10 6 6 13 4 0 - 10 8.04 2.20 65 14 1 0 - 10 3.85 2.85 0 - 10 3.33 2.85 73 5 2 0 - 10 5.44 2.47 6 - 24 17.98 3.60 40 0 0 7 6 6 5 4 5 6 1 3 - 10 8.22 2.00 34 6 0 0 - 10 4.19 2.93 0 - 10 3.47 2.87 35 4 1 0 - 10 5.50 2.78 6 - 24 18.03 4.47 38 1 1 10 7 5 5 2 1 7 4 0 - 10 7.89 2.40 31 8 1 0 - 9.5 3.51 2.77 0 - 10 3.18 2.85 38 1 1 0 - 10 5.37 2.11 13 - 24 17.92 2.70  Descriptive statistics were also sufficient to show that the control and experimental groups were similar for pain in the past, type of most painful past experience, pain at the moment and expected pain on the day after surgery. A non-parametric Mann-Whitney U test indicated that the groups did not differ significantly in relation to their most painful past experience scores (U = 723; p = 0.57). There were also no significant differences regarding current pain intensity (U = 684; p = 0.35) or pain distress (U = 728; p = 0.61) on entry to the study. A t-test carried out on the intensity of pain expected on the day after surgery showed no significant differences between the groups (t = 0.23; p = 0.82; CI - 1.03, 1.30). Similarly, a Mann-Whitney U test revealed no significant differences (U = 537; p = 0.51) in the two groups’ STAI scores. Eleven subjects failed to complete the short form STAI on entry to the study as they appeared to misunderstand the instructions and had scored just one item out of the six. The STAI formed part of the preoperative questionnaire. As baseline data, the questionnaire was completed prior to randomisation and remained sealed in an envelope until the subject exited from the study. The fact that the data were incomplete could therefore not be identified in advance. However, once the initial problem was recognised, instructions were added to the effect that all items on the STAI should be completed. This appeared to reduce the frequency of the problem. The distribution of incomplete data were eight from the control group and three from the experimental group. In summary, with the exception of social class, analysis of the baseline characteristics indicated that there were no significant differences between the control and experimental groups on entry to the study. The amount of missing data may explain the social class anomaly. Overall, it was concluded that the randomisation process was effective in terms of the sample characteristics and identified intervening variables potentially related to pain perception and strategy choice. The following section moves on to consider whether, or not, the hypothesis was supported in relation to any of the outcome measures. 5.3.3 Outcome Measures Eight main outcome variables were identified in testing the hypothesis. Each of these is now addressed in turn. A table summarising the main findings is presented in section 5.3.3.9. 5.3.3.1 Pain Intensity Pain intensity was analysed in two ways. Firstly, the overall mean for each group was calculated and a t-test undertaken to assess the significance of the results. These are summarised in Table 26. Table 26: Overall Pain Intensity.  Control Group  Experimental Group Statistic n  (  SD  n  (  SD  t-value p =  CI Result  40  3.95  2.01  40  2.74  1.71  2.91 0.005  .385, 2.050  Range Maximum score Median  0 - 8.5 8.5 3.65  0 - 7 7.0 2.42   0.025 0.009  Overall, subjects in the experimental group experienced significantly less pain than those in the control group. The experimental group also had lower maximum and median scores on the 0 - 10 scale. The second approach to analysis of the pain intensity data involved calculating mean scores for each group on a daily basis. This was done to gain insight into whether there was a pattern in the difference observed in the overall pain intensity, related to the postoperative stage. Again t-tests were undertaken on the results which are shown in Table 27. Table 27: Daily Pain Intensity.  Control Group  Experimental Group Statistic n  (  SD  n  (  SD  t-value p =  CI Day 1 40 4.86 2.91 39 3.72 2.63 1.83 0.070 -.098, 2.387 Day 2 38 4.71 2.58 40 3.75 2.65 1.62 0.109 -.219, 2.140 Day 3 39 4.04 2.46 38 3.00 2.44 1.86 0.067 -.074, 2.151 Day 4 37 3.86 2.65 37 2.92 2.46 1.59 0.116 -.240, 2.132 Day 5 34 4.09 2.89 35 2.34 2.06 2.88 0.006 .532, 2.958 Day 6 32 3.28 2.44 25 2.16 2.41 1.73 0.089 -.176, 2.419 Day 7 26 3.35 3.02 26 1.96 2.07 1.93 0.060 -.062, 2.831 Day 8 23 3.35 2.74 23 1.83 2.12 2.10 0.041 .062, 2.981 Day 9 20 3.50 2.86 18 1.67 1.85 2.37 0.024 .260, 3.406 Day 10 17 3.18 2.92 16 1.75 2.52 1.51 0.142 -.506, 3.359 Day 11 13 3.31 2.69 9 2.22 2.63 0.94 0.359 -1.339, 3.510 Day 12 12 2.92 2.97 9 1.78 2.54 0.95 0.356 -1.385, 3.663  When daily scores were calculated it was evident that experimental subjects had lower mean scores for each of the days analysed. The difference in the mean scores was significant on days five (p = 0.006), eight (p = 0.04) and nine (p = 0.02). Of note, days six (p = 0.08) and seven (p = 0.06) were approaching significance. By day 10 the majority of subjects had been discharged home. Most patients on the Ward were administered opiates via PCAs or epidural for three to four days following surgery. It could be postulated that the reason for the occurrence of significance, or near significance, between days five and nine was due to experimental patients using self-generated strategies pro-actively once their PCAs or epidural were discontinued. In conclusion, the hypothesis “overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care reduces postoperative pain intensity” was supported when looking at overall pain intensity scores. 5.3.3.2 Pain Distress Data relating to pain distress were analysed in a similar fashion to that on pain intensity. Table 28 shows the results of the overall pain distress analysis. Table 28: Overall Pain Distress.  Control Group  Experimental Group Statistic n  (  SD  n  (  SD  t-value p =  CI Result  40  3.08  2.29  40  1.91  1.42  2.74 0.008  .316, 2.015  Range Maximum score Median  0 - 9.5 9.5 2.55  0 - 5.5 5.4 1.67    Significantly less pain distress was experienced overall by subjects in the experimental group, compared to the control group. The experimental group also had lower maximum and median scores on the 0 - 10 scale. The results of the daily pain distress analysis are shown in Table 29, overleaf. Table 29: Daily Pain Distress.  Control Group  Experimental Group Statistic n  (  SD  n  (  SD  t-value p =  CI Day 1 38 3.96 3.26 39 2.79 2.97 1.64 0.106 -.254, 2.585 Day 2 38 3.82 3.12 40 2.85 2.69 1.48 0.143 -.339, 2.297 Day 3 39 3.10 2.97 37 2.05 2.17 1.76 0.082 -.138, 2.235 Day 4 37 3.00 2.65 37 2.59 2.60 0.66 0.508 -.810, 1.621 Day 5 34 3.23 2.94 36 2.08 2.33 1.81 0.076 -.122, 2.426 Day 6 32 2.50 2.77 25 1.92 2.66 0.80 0.426 -.871, 2.031 Day 7 26 2.27 2.78 26 1.35 2.06 1.36 0.180 -.442, 2.288 Day 8 23 2.48 2.83 23 1.04 1.80 2.05 0.047 2.036E-02, 2.849 Day 9 20 2.80 3.17 18 0.89 1.57 2.39 0.024 .274, 3.548 Day 10 17 2.82 2.92 16 0.75 1.06 2.74 0.012 .497, 3.650 Day 11 13 3.23 3.03 9 1.22 1.64 2.00 0.060 -9.01E-02, 4.107 Day 12 11  2.45 2.94 9 1.11 1.69 1.28 0.219 -.882, 3.569 Significant differences were observed between days eight and ten. In conclusion, the hypothesis “overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care reduces postoperative pain distress” was supported when looking at overall pain distress scores. 5.3.3.3 Analgesic Consumption Data about analgesic consumption were collected in four main categories: morphine via PCAs or epidural was calculated as total consumption, while consumption of IM/oral opiates, non-opiates and NSAIDs were recorded as number of doses. One subject from the control group, who underwent an arthrodesis, received no analgesics. Otherwise all subjects were administered analgesics. The table below summarises the number of subjects administered analgesics by category. Table 30: Analgesic Administration by Category and Group. Category Number of Control Subjects Receiving Analgesic Number of Experimental Subjects Receiving Analgesic  Total Morphine via PCAs/epidural IM/oral opiates Non-opiates NSAIDs  26 26 38 21 25 16 40 32  51 42 78 53 Total  111 113 224 The next table looks at the amount of morphine used in the two groups. Table 31: Amount of Morphine via PCAs/Epidural. Statistic  Overall Sample (n = 51)  Control Group (n = 26) Experimental Group (n = 25) Range Median Mean SD  3 - 491 mg 137 mg 159 mg 105.91 3 - 491 mg 123 mg 156 mg 122.38 30 - 343 mg 158 mg 163 mg 87.99  Although the dispersion was wider for the control group, the means were similar. No significant differences were found between the groups using a Mann-Whitney U test (U = 294.5, p = 0.57) or in assessing for categorical differences around the median split using Chi-square tests (X² = 1.01, df = 1, p = 0.30). Table 32 shows the number of IM/oral opiate doses received by subjects in the two groups. Table 32: Number of IM/Oral Opiate Doses. Statistic  Overall Sample (n = 42)  Control Group (n = 26) Experimental Group (n = 16) Range Median Mean SD  1 - 56 doses 9 doses 13.62 doses 13.89  1 - 56 doses 9 doses 14.62 doses 15.60  1 - 28 doses 9 doses 12 doses 10.84  No statistical differences were found between the two groups in relation to the mean number of IM/oral opiate doses received (U = 184.5, p = 0.54). The total number of doses for each group was markedly different. The control group received 380 doses and the experimental group 192 doses. Multiplying the experimental mean by the control group numbers (n = 26) results in a figure of 312. It would be interesting to cost the difference of 68 doses in relation to nursing time, pharmacy time and the cost of analgesics. When the data from both opiate categories were considered together, it was apparent that there were differences in the number of subjects from each group who were administered opiates via the PCAs/epidural and, or, IM/oral routes. The analysis is shown in the table below. Table 33: Number of Subjects Receiving Opiates. Route  Control Group (n = 36)  Experimental Group (n = 35)  Significance Morphine via PCAs/epidural PCAs/epidural alone IM/oral opiates IM/oral alone Both PCAs/epidural and IM/oral Neither opiate route  26 10 26 10 16 4  25 19 16 10 6 5  X² = 7.31, df = 1, p = 0.006 X² = 5.01, df = 1, p = 0.024 X² = 7.31, df = 1, p = 0.006 As can be seen in Table 33, significantly more experimental subjects received opiates via PCAs/epidural alone (p = 0.006), while significantly fewer received IM/oral opiates (p = 0.024) or opiates via both the PCAs/epidural and IM/oral routes combined (p = 0.006). The following table considers non-opiate consumption. Table 34: Non-Opiate Consumption. Statistic  Overall Sample (n = 78)  Control Group (n = 38)  Experimental Group (n = 40)  Range Mode Median Mean  1 - 60 doses 2 doses 12.5 doses 15.14 doses  1 - 60 doses 7 doses 13 doses 15.32 doses  1 - 51 doses 2 doses 11 doses 14.97 doses  Data on non-opiate consumption revealed no statistical (U = 707.0, p = 0.60) differences between the mean doses administered to each group. There were also no differences in the number of subjects from each group receiving non-opiates. The difference in the modes and medians might have a clinical significance if the findings were replicated in a larger sample. The category “non-opiates” referred to simple analgesics. Non-steroidal anti-inflammatory drugs (NSAIDs) were considered separately. Table 35 shows the results of the analysis on NSAIDs. Table 35: NSAID Consumption. Statistic  Overall Sample (n = 53)  Control Group (n = 21)  Experimental Group (n = 32)  Range Median Mean SD  1 - 41 doses 10 doses 11.64 doses 8.87  2 - 27 doses 11 doses 12 doses 7.80  1 - 41 doses 9.5 doses 11.41 doses 9.62 Again mean NSAID consumption was comparable between the two groups (U = 298.5, p = 0.49), as was consumption calculated around the median split (X² 0.7256, df = 1, p = > 0.25). However, considerably more experimental subjects used NSAIDs in comparison to control subjects. This difference was statistically significant (X² = 6.7645, df = 1, p = 0.009). In summary, no statistical differences were found between the two groups in relation to mean consumption within each of the four analgesic categories. When the number of subjects by group and analgesic category were considered, significantly fewer experimental subjects were found to have received IM/oral opiates alone or in combination with PCAs/epidural morphine. Conversely, significantly more experimental subjects received NSAIDs. The hypothesis “overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care, reduces analgesic consumption” was supported in part. 5.3.3.4 Antiemetic Consumption Antiemetics were not identified as an outcome measure within the hypothesis. However, the potential link between opiates and postoperative nausea and vomiting was recognised. It was therefore deemed important to analyse antiemetic consumption as a possible intervening variable in relation to analgesic consumption. The results are shown in Table 36. Table 36: Postoperative Antiemetic Consumption. Statistic  Overall Sample (n = 49)  Control Group (n = 27)  Experimental Group (n = 22)  Number of doses Range Median Mean SD 95 doses 1 - 6 doses 1 dose 1.99 doses 1.25 56 doses 1 - 6 doses 2 doses 2.07 doses 1.36 39 doses 1 - 5 doses 1 dose 1.77 doses 1.12A Mann-Whitney U test confirmed that the result was not statistically significant (U = 261.5, p = 0.43). However, fewer experimental patients received antiemetics and those who did so, tended to have fewer doses than subjects in the control group. This may reflect the finding that fewer experimental subjects received IM/oral opiates. There may be a clinical significance in terms of quality of life, time and cost in the observed difference. 5.3.3.5 Anxiety Despite a wide dispersion of STAI scores in the control group, no statistically significant differences were observed between the groups on entry to the study (U = 537, p = 0.51). Postoperatively, mean scores were increased in both groups indicating a lessening of overall anxiety. However, again there were no significant differences between the two groups postoperative scores (U = 522, p = 0.78). The relevant descriptive statistics are summarised in Table 37. Table 37: STAI Scores. Statistic  Baseline STAI (Overall) Control Group Baseline STAI  Experimental Group Baseline STAI Postoperative STAI (Overall) Control Group Postoperative STAI Experimental Group Postoperative STAI Range Mean SD  6 - 24 17.97 3.60 6 - 24 18.46 4 13 - 24 17.93 2.74  8 - 24 19.86 3.89 8 - 24 19.32 4.36 9 - 24 20.23 3.49 The STAI data were then considered from the perspective of comparing entry data against exit data using Wilcoxon Matched-Pairs Signed Ranks tests. The results are shown in Table 38 overleaf. Table 38: Comparison of Entry and Exit STAI Scores. Statistic  Entry vs. Exit Control Entry vs. Control Exit  Experimental Entry vs. Experimental Exit  Mean Rank Z statistic 2-tailed p  20.79 / 27.23 -2.5763 0.01 11.45 / 13.38 -0.6857 0.49 10.17 / 13.89 -2.7311 0.006 The results demonstrate that there was a significant difference between the pre and postoperative STAI scores. A further test was undertaken to identify whether or not the difference existed in both groups. As the table demonstrates, it was the experimental subjects who showed a statistically significant reduction in anxiety during their period of hospitalisation. No such reduction was observed in the control group. The hypothesis, “overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care reduces anxiety”, was supported as experimental subjects showed a significant reduction in anxiety during their hospital stay, whereas control subjects did not. It could be postulated that this difference arose as a result of experimental subjects being empowered to take control while participating in their postoperative pain management. However, given that the difference between the mean entry and exit scores is only 2.30 (on a scale with a possible range of 0 to 24), it is questionable how much clinical impact this statistically significant reduction in anxiety actually had. 5.3.3.6 Length of Stay No significant difference was observed between the control and experimental group’s mean length of stay (U = 701, p = 0.34) as shown in Table 39 overleaf. Table 39: Length of Stay. Statistic  Overall Sample Control Group Experimental Group Range Median Mean SD  3 - 23 11 10.95 4.58 3 - 23 11 11.38 4.49 3 - 22 11 10.53 4.69 However, a contingency table revealed that 80% of experimental subjects had been discharged by day 13 in comparison to 65% of control subjects. Two Chi-square tests were undertaken on length of stay divided into 13 days or less, versus 14 days or more. The first of these was undertaken with an outlier from the experimental group included revealing a non-significant result (X² = 2.2570, df = 1, p = 0.13). With the outlier excluded, the result was approaching significance (X² = 2.9418, df = 1, p = 0.086). Munro and Page (1993) suggest that although outliers frequently reflect errors of recording, they may also represent extreme values for an unusual subject, as was the case in this study. Where the results of significance testing are dissimilar, as a consequence of including or excluding the outlier, Munro and Page (1993) suggest that analysis which is resistant to outliers can be used. Median values are one example of outlier resistant statistics. The median values suggest that the non-significant results of the first Chi-square test are more reliable. Although it is possible that a larger sample might have obtained a significant result on this outcome measure, the hypothesis “overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care reduces length of stay”, could not be supported. 5.3.3.7 Patient Satisfaction No significant difference was observed between the control and experimental groups mean patient satisfaction scores (U = 781, p = 0.85) as shown in Table 40 overleaf. Table 40: Patient Satisfaction. Statistic  Overall Sample  Control Group Experimental Group Range Mean SD  3 - 10 8.68 1.80 3 - 10 8.63 1.89 4 - 10 8.74 1.71  The hypothesis “overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care enhances patient satisfaction with their postoperative pain management” was therefore not supported. The data collection instrument used to measure patient satisfaction was a 0 - 10 verbal numerical scale anchored by “not at all satisfied” and “completely satisfied”. Fifty per cent of respondents in both groups scored nine or above suggesting that the scale lacked sensitivity in assessing patient satisfaction with pain management. 5.3.3.8 Patients’ Perceptions of Control Over Pain Management Data about control of pain management were collected on two 0 - 10 verbal numerical scales anchored by “not at all in control” and “completely in control”. One scale referred to subjects’ perceptions of self-control and the other to their perceptions of the degree to which health care practitioners were in control. Patient control of pain management showed a bi-modal distribution with 24 subjects scoring 5 - 6 and 23 subjects 9 - 10. The overall mean was 6.39, SD = 2.59. Table 46 summarises the distribution by group. Table 41: Patient in Control of Pain Management. Statistic  Overall Sample  Control Group Experimental Group Range Modal category 5 - 6 Modal category 9 - 10 Mean SD 0 - 10 n = 24 n = 23 8.68 1.80 2 - 10 n = 14 n = 9 8.63 1.89 0 - 10 n = 10 n = 14 8.74 1.71As anticipated from the descriptive statistics, a Mann-Whitney U test showed no significant difference between the two groups mean scores (U = 680, p = 0.24). Similarly a Chi-Square test, based upon the bi-modal categories, showed no significant difference between the two groups (X² = 1.7412, df = 1, p = < 0.25 > 0.1). Table 42 summarises the distribution for practitioner control of pain management. Table 42: Nurse/Doctor in Control of Pain Management. Statistic  Overall Sample  Control Group Experimental Group Range Modal category 8 Modal category 10 Mean SD  2 - 10 n = 15 n = 25 7.75 2.22 3 - 10 n = 9 n = 12 7.79 2.20 2 - 10 n = 6 n = 13 7.71 2.28  Again, a bi-modal distribution was observed with 25 subjects scoring 10 on the 0 - 10 scale and 15 scoring eight. A Mann-Whitney U test showed no significant difference between the two groups mean scores (U = 793, p = 0.95) in relation to nurse/doctor control of pain management. The hypothesis “overt identification and systematic incorporation of patients’ self-generated strategies into the plan of care enhances patients’ perceptions of control” was therefore not supported on this measure. However, on hindsight the manner in which perceptions of control were ascertained was flawed. A more useful approach, assuming the continued use of a 0 - 10 verbal numerical scale, may have been to anchor one end with “nurse/doctor in control” and the other with “patient in control”. This might allow the balance of control (and thereby the power base) to be assessed, which was the original intention. 5.3.3.9 Summary In summary, the hypothesis was supported in relation to anxiety, pain intensity and pain distress and partially supported in terms of analgesic consumption. No significant differences were observed with reference to length of stay, patient satisfaction or perceptions of control of pain management. Table 43 summarises the main findings from the RCT. Table 43: Summary of the RCT Findings. Outcome Measure  Finding Pain intensity and pain distress Analgesic Consumption STAI Length of stay Patient satisfaction Control of pain management  experimental subjects’ mean pain intensity (p = 0.005) and pain distress (p = 0.008) were significantly less than control subjects differences were particularly noted on days five to nine post surgery no statistically significant differences were observed in mean analgesic consumption between the control and experimental groups in each of the four analgesic categories experimental subjects received fewer IM/oral opiates which may have clinical significance in terms of time and cost significantly fewer experimental patients received IM/oral opiates alone or in combination with PCAs/epidural morphine (p = 0.006) significantly more experimental subjects received NSAIDs (p = 0.009) no statistically significant differences existed between the control and experimental groups mean STAI scores experimental subjects had a significant reduction in anxiety during their stay in hospital (p = 0.006) no statistically significant difference was observed regarding mean length of stay a larger sample might have revealed a difference no statistically significant results were observed the data collection instrument may have lacked sensitivity no statistically significant results were observed the data collection instrument was flawed lacking validity The following section moves on to consider strategy use during the RCT. 5.3.4 Strategy Use The 40 experimental subjects recorded a total of 331 self-generated pain relieving strategies on admission to hospital and listed a total of 254 strategies that they would like to use postoperatively. The breakdown by theme is shown in Table 44. Table 44: Experimental Subjects’ Strategy Use On Admission to Hospital and Preferences for Postopertive Use. Strategy Theme  Number of Times the Theme was Recorded on Admission  % of Total Number of Times the Theme was Recorded For Use in Hospital  % of Total Behavioural Resting alone Analgesics Positioning Cognitive and expressive reassurance Seeking treatment Cognitive control Mobilising General help Miscellaneous  121 35 33 30 29 25 24 21 9 4  37.0% 11.0% 10.0% 9.0% 8.5% 7.5% 7.0% 6.0% 3.0% 1.0%  94 22 31 27 24 21 9 19 5 2  37.0% 8.5% 12.0% 10.5% 9.5% 8.5% 3.5% 7.5% 2.0% 1.0%  Total  331 100%  254 100%  As can be seen in the table, the percentages by which subjects wished to continue using existing self-generated strategies postoperatively were similar for many themes. Of note, resting alone fell from being the second most frequently recorded theme at admission, to fifth place for use in hospital. Cognitive control was also less popular, while analgesics and positioning were listed more frequently as strategies subjects would like to use in hospital. Experimental subjects recorded using 265 self-generated and 136 practitioner initiated strategies following surgery. Subjects in the control group used a similar number, recording 240 self-generated and 150 practitioner initiated strategies postoperatively. Table 45 shows the number of strategies used by individuals according to group and whether the strategy was perceived to be self or practitioner initiated. Table 45: Number of Strategies Used by Individuals Postoperatively. Number  Number of Self-Generated Strategies Used by Subjects Postoperatively (Range 1 - 14)  Number of Nurse/ Doctor Initiated Strategies Recorded by Subjects Postoperatively (Range 1 - 12)  Experimental Group (n = 39)  Control Group (n = 38) Experimental Group (n = 37) Control Group (n = 34) 1 - 5 6 - 9 > 9  10 21 8  14 16 8  28 8 1  25 5 4  Table 46 shows the breakdown by theme. Table 46: Subjects Postoperative Strategy Use, by Group and Theme  Experimental Group  Control Group  Self-Generated  Practitioner-Initiated  Self-Generated  Practitioner-Initiated  Theme  Nos. of Times the Theme was Recorded  % of Total Nos. of Times the Theme was Recorded  % of Total Nos. of Times the Theme was Recorded  % of Total Nos. of Times the Theme was Recorded  % of Total Behavioural Resting alone Mobilising Positioning Cognitive & expressive reassurance Analgesics Seeking treatment Cognitive control General help Miscellaneous  100 30 25 24 21 20 18 15 8 4  38.0% 11.0% 9.5% 9.0% 8.0% 7.5% 7.0% 5.5% 3.0% 1.5%  23 7 19 12 15 35 16 0 7 2  17.0% 5.0% 14.0% 9.0% 11.0% 26.0% 12.0% 0.0% 5.0% 1.0%  93 38 16 22 18 15 15 18 9 10  37.0% 15.0% 6.5% 8.5% 7.0% 6.0% 6.0% 7.0% 3.0% 4.0%  31 6 23 16 13 32 14 2 9 4  21.0% 4.0% 15.0% 10.5% 8.5% 21.5% 9.5% 1.5% 6.0% 2.5% Total265100%136100%254100%150100%The table shows that the number of self-generated strategies recorded by experimental patients postoperatively was more or less equivalent to the number they anticipated using. The theme resting alone was the second most frequently used self-generated strategy by both groups, with the experimental group percentage at the same level as on admission and the control group somewhat higher. Mobilising was used more frequently by experimental subjects, and cognitive control less frequently. Otherwise the percentage distribution across the groups and for experimental subjects was remarkably consistent across time, from admission to discharge. Unsurprisingly analgesics were the most frequently recorded practitioner initiated strategy, although in the control group this figure was matched with strategies from the behavioural theme. Overall, frequencies for practitioner initiated strategies were more or less consistent across the groups A visual comparison of the RCT and survey data suggested that cognitive and expressive reassurance and positioning were more frequently recorded by the inpatient sample. Otherwise, strategy use was fairly consistent across the two samples. The postoperative strategy questionnaire asked subjects to score how helpful each recorded strategy was in relieving pain. The scoring system was based upon a five-point scale ranging from “not at all helpful” to “completely helpful”. The sum for each strategy was identified by group and the mean calculated. Table 47 presents the results. The table is complex in so far as each individual strategy is listed. Although calculating mean scores might have enhanced clarity by theme, this would have been clinically questionable. It was also against the philosophy of the study, which aimed to empower experimental subjects to use their individual self-generated strategies. Presenting the results thematically would hide detail such as the perception that breathing exercises were only regarded as fairly helpful, while the use of cold was viewed as very helpful. Both these strategies are from the behavioural theme and amalgamating the scores would have created a meaningless result. Table 47: Mean Score of Perceived Strategy Efficacy by Group and Strategy. Theme Strategy Efficacy of Self Generated Strategy (( scores 0 - 5) Efficacy of Nurse / Doctor Strategy (( scores 0 - 5) Experimental Group Control Group Experimental Group Control Group Analgesics painkillers 3.9 3.4 4.0 3.9 Behavioural distraction. relaxation breathing exercises imagery music massage warmth cold 3.1 3.0 2.4 2.9 3.1 2.8 3.4 3.6 2.8 3.2 3.3 3.7 3.2 3.3 3.3 3.0 3.3 3.4 3.0 4.0 4.0 3.0 4.0 4.0 2.0 3.4 2.8 __ __ 3.0 3.4 3.9 Resting alone resting alone lying down trying to sleep peace & quiet 3.1 3.6 3.6 3.7 3.5 3.4 3.5 3.3 5.0 2.5 2.5 4.0 __ 4.3 __ 4.0 Cognitive control grin & bear/ positive thinking 3.1 2.8 __ 3.5 Mobilising  mobilising walking moving about exercise 4.0 3.3 3.1 3.0 3.4 2.6 2.8 3.0 3.9 3.5 2.3 2.0 3.0 3.2 2.3 3.5 Positioning  positioning changing position support painful area 3.0 3.2 3.3 3.7 2.8 3.6 3.5 3.7 3.6 2.9 3.5 4.0 Seeking treatment treatment seek medical attention advice/information 4.2 3.6 3.5 2.8 3.2 3.7 3.8 4.2 3.5 4.2 3.8 2.8 Cognitive & expressive reassurance reassurance talking about pain touched/ hugged confidence in other 3.5 3.0 3.8 3.7 2.6 3.8 3.5 4.3 4.4 3.5 4.7 4.0 3.5 4.5 __ 4.0 General help help with activities 3.3 2.8 4.0 3.4 Miscellaneous other 4.0 4.0 4.0 4.3 Strategies with mean scores of > 3.9 were perceived as very or completely helpful in relieving pain. Those with mean scores of 3.0 - 3.9 were perceived as fairly helpful and those with scores of < 3.0 were not regarded as helpful. With the exception of the miscellaneous theme, no strategy was regarded as very or completely helpful across all of the groups. Miscellaneous strategies tend to be highly individualised, however, care should be taken with any additional interpretation as the numbers in this theme were extremely small. The table shows that the majority of strategies were perceived to be fairly, very or completely helpful. Of the 30 strategies listed, experimental subjects had a higher mean score than control subjects on 17 self-generated strategies and a lower score on 11. Of the nurse/doctor-generated strategies, experimental subjects scored higher on 18 strategies and lower on seven. Strategies that were scored higher by experimental subjects in both the self-generated and practitioner categories included: analgesics, distraction, warmth, cold, trying to sleep, mobilising, walking, changing position, reassurance, touching/hugging and general help. Strategies scored higher by control subjects in both groups were supporting the painful area and talking about the pain. In summary, subjects identified a wide range of strategies prompted by the self-completion questionnaire developed from the survey data. Both experimental and control subjects used self-generated pain relieving strategies postoperatively. The degree to which these strategies were actually perceived as promoting pain relief varied across the strategies and according to whether use was self or practitioner generated. The following section considers additional data from the field notes recorded during the RCT about individual approaches to postoperative pain and its management. 5.3.5 Researchers’ Reflections During the trial, field notes were recorded by the researchers on subjects’ comments about their pain and its management, as well as their comments about the trial design. Additional qualitative data were obtained from comments written by subjects on data collection instruments and from letters received by the researchers following the subject’s discharge from hospital. Three themes were identified relating to pain perception, strategy use and the trial design. Each is addressed below. 5.3.5.1 Pain Perception Comments elicited from subjects regarding pain perception covered areas such as definitions of pain, pain experiences and the influence of completing the pain intensity and distress scales. Reflecting McCaffery’s (1968) definition of pain one of the subjects suggested that: “Pain is a thing that cannot be defined specifically, as only the person who has it knows what it is.” Other subjects commented upon their past and present experiences of pain as follows: “The pain has taken over my life. It must be terribly boring for everyone else. It’s just the pain that is frightening me.” “Note: there are two types of pain - lengthy aching and short stinging!!” “Sometimes pain is nagging, aching or sharp jabs (like knives being stuck in) (My) “pain has not been believed. They lift my leg and just drop it. It’s very uncomfortable. I feel very upset. Seems I do everything wrong.” (Control subject). Several subjects from both groups commented upon the positive feedback that they received when completing their pain intensity and distress scales. For example: “Seeing the pain and distress go down is very positive”. (Control subject) “It’s really good how it gives you a visual picture of how you’re getting better. I really look forward to filling it in each day. Even the other day when it got worse I marked the distress less cos it was getting better overall and I could see that.” (Experimental subject) Another wrote on their scales when the scores reached zero: “Escape day. Success!”(Experimental subject) 5.3.5.2 Pain Relieving Strategies Many of the comments regarding the use of pain relieving strategies focused upon the impact of practitioners and were linked to the themes cognitive and expressive reassurance or seeking treatment. For instance: “My pain’s been very low indeed due to the attention I’ve received.” (Experimental subject) “Staff being there with advice and support when needed was completely helpful for my pain.” (Control subject) “The most helpful contribution to pain management is being able to discuss the treatment - diagnosis/prognosis with all medical staff, particularly the consultant and his assistants.” (Control subject) “The nursing and help were excellent for my pain.” (Experimental subject) “The compassionate approach by doctors and nurses was very helpful.” (Control subject) “The encouragement from the physios was excellent.” (Experimental subject) Other comments related to specific strategies, such as: My own pain threshold appears to be approx. 2 o’clock onwards the reason being joint positions due to sleeping movements which I find aggravate the joints and wake me up until a new position has been found.” (Experimental subject) “My faith in God is wonderfully relaxing and cheering and works very well for the pain.” (Experimental subject) “The pain wears off with exercise and seizes up with too much rest.” (Control subject) Several comments were also elicited regarding the ability of nurses to support self-generated strategies in the hospital setting. One subject who finds warmth very helpful for pain relief commented: “Warmth with blankets is okay, but I would have liked heat pads. They haven’t got any spare at the moment.” (Experimental subject) “The only time I am pain free with my hip is when I relax in a fairly hot bath. However, the pain returns about as soon as I have dried myself. I’d like to use baths in hospital, but I guess it will be difficult with the operation!” (Experimental subject) Finally, a number of statements were made regarding analgesic use as follows: “The choice of different analgesics at different stages is very helpful.” (Control subject) “Painkillers give me stomach and bowel problems and are therefore limited.” (Experimental subject) (I) “Cannot take painkillers, even codeine, cos they mess my system up. That’s possibly why I have a high threshold - it’s mind over matter and keeping busy that’s really important” (Control subject) Both of the latter two subjects went on to use analgesics and to score them as completely helpful on their postoperative strategies questionnaires. The next section considers the comments made regarding the trial design. 5.3.5.3 The Trial Design Most of those who commented were complimentary about the research design and were particularly interested to find nurses undertaking, not just research but, a randomised controlled trial. However, there were two specific exceptions. The first of these arose during a conversation between one of the researchers and a subject who was a retired anaesthetist. Although the researcher felt the subject was showing interest, the tone and approach were perceived to be challenging. The field notes were as follows: “Typical! Got into conversation with Dr ___. Typical doctor putting nursing research down. He seems interested in what we are doing, patronising that we are doing it at all and full of his own ideas on how we should have done it without even listening to me! Thank God he’s not an experimental patient - woops, I shouldn’t be biased. On the plus side, I think I “handled” him better than I’d have done before. I know more about pain and am more confident in myself”. The second subject was a practising Consultant Radiologist who wrote “rubbish” across the behavioural strategies theme on the postoperative questionnaire and suggested: “Translate into English.” It should be noted that several other health care practitioners, including medical staff, participated as subjects in the RCT and that the comments above were the exception, rather than the rule. The following section summarises the results and offers some initial interpretations. 5.4 Summary and Initial Interpretations The following discussion begins to highlight some of the key findings. The main discussion is presented in chapter seven. The literature on chronic pain management indicates that patients’ pain experiences may be decreased or controlled by the employment of self-generated strategies which modify pain perception through peripheral and central mechanisms (Craig, 1999; Weisenberg, 1999). In acute, postoperative settings, pain management has tended to focus upon reducing psychophysiological distress through the administration of analgesics and information giving. Consequently, postoperative pain management has, on the whole, been constrained to addressing the sensory and affective dimensions of pain (Melzack and Wall, 1996; Sofaer, 1998; McCaffery, 1999). Where a more multidimensional approach has been adopted, the emphasis has been upon taught rather than self-generated strategies. The efficacy of taught strategies is unclear (Sindhu, 1996; Grenside, 1997; Seers and Carroll, 1998). A few studies have suggested that it may be more useful to offer a choice of strategies rather than confining individuals to specific practitioner identified techniques (Stevens et al, 1987; Rokke and al’Absi, 1992). The work of Chaves and Brown (1987), Pick et al (1990), Burrows (1997) and Kashikar-Zuck et al (1997) highlights that patients employ self-generated strategies for acute pain relief. However, no previous study has tested whether identifying and supporting patients’ strategies postoperatively has the potential to influence postoperative outcomes. As well as enhancing individualised pain management, focusing upon self-generated strategies may encourage active participation in care - a move that successive governments have been increasingly keen to promote over the last 11 years (DH 1989, 1997, 1998, 2000). The emphasis upon patient-centred care relies upon practitioners increasing their understanding of individuals’ ideologies and health care actions (Nethercott, 1993; Wilkinson, 1999). Without this knowledge practitioners are unable to offer appropriate choices and involve patients in decision making. While the degree of involvement should be patient-led, and there is evidence to suggest that not all individuals want to assume responsibility for their health care actions (Biley, 1995; Caress, 1997), the premise of patient-centred care is one that requires a shift in the balance of power from practitioners to patients (Cahill, 1998). Copp (1974) identified that while self-generated strategies may be employed to relieve acute pain at home, patients may be disempowered in terms of extending their pain management practices into the hospital setting. The survey supported previous findings that most individuals do acquire a repertoire of self-generated strategies for dealing with pain (for example: Broome, 1986; Kashikar-Zuck et al, 1997). Strategies are learnt during childhood and developed through exposure to pain and the advice of family, friends and health care practitioners as the individual moves into adulthood. What is learnt during one pain episode, is re-tested in the next. Gradually, individuals identify those strategies that they think will either help relieve their pain, or at least promote a sense of self-control. Thus, most individuals use techniques based upon their ideologies surrounding pain and its management and regard themselves as active participants during pain episodes. The range of strategies employed is frequently wide and is linked to the intensity of pain, such that the more intense the experience, the greater the number of strategies used. There was also evidence that intensity influenced the type of strategies adopted. Relatively mild or short lasting pain is typically ignored or managed through cognitive control and behavioural strategies such as distraction. These techniques are used to help the individual disregard the sensory impact of pain and continue with their normal routines. As the intensity, or duration, begins to increase, strategies tend to become more focused. For instance, warm baths and going for walks are commonly recorded. This is followed by techniques such as mild analgesics and cognitive and expressive reassurance. Finally, resting alone, positioning, seeking treatment and medically prescribed analgesics are prevalent additions for more serious pain. Thus, as pain experiences demand more attention, self-management becomes increasingly multidimensional. The idea that mild everyday aches and pains should be ignored is culturally bound and, to some extent, linked to normal expectations of leading an active life and of ageing. Otherwise, the beliefs surrounding why particular strategies were adopted for everyday pain, while others were employed for acute pain, were difficult to expose during the relatively structured interviews. There was evidence to suggest that the majority of respondents developed their repertoires on the recommendation of third parties. With family and friends, recommendations had to be based upon personal use and proven efficacy for that individual. With practitioners, advice was taken on the premise that it was grounded in functional and/or empirical knowledge and must therefore work. When strategies were unsuccessful, individuals tended to absolve others of responsibility and simply regard themselves as different. However, comments were elicited which indicated that strategies, which had proved unsuccessful for everyday pain, might still be re-tested during acute pain episodes. This experimentation also worked in reverse. Having developed a repertoire that worked for them, many respondents were adamant that they wanted to use their preferred strategies in hospital, to be treated as individuals and to be fully engaged in their pain management. However, they also expressed concern that they would lose control to practitioners. Conversely, others expressed a preference for practitioners to adopt a nurturing role. These individuals wanted to use their strategies in hospital but either expected not to be allowed to do so, or stated that they would only do so if the doctors and nurses did not mind. Both groups appeared to be disempowered, be it all from contradictory perspectives. The survey was not designed to enable the co-action researchers or collaborative group to work towards empowering respondents in relation to their postoperative pain management. However, the advisory role adopted by respondents helped to raise consciousness regarding the possibilities of employing self-generated strategies postoperatively. Some participants might, therefore, have gone on to use their strategies in hospital. The survey findings enabled the collaborative group to discuss ideas for practice transformation both within the group and with the collaborative informants. The results were also used to generate the self-completed Strategies Questionnaires. The questionnaires were initially administered to experimental subjects on admission to hospital and subsequently to all subjects postoperatively. The RCT findings supported those of the survey by confirming that individuals develop a repertoire of pain relieving strategies and are open to employing these strategies in the hospital setting. Both experimental and control subjects used pain relieving strategies postoperatively. Although there were some differences observed in the type of strategies used, the total number employed by each group was similar. Most strategies were regarded as being fairly or very helpful, however there were a few, such as distraction, which were deemed unhelpful. Again this supported the survey findings which indicated that there was a hierarchy of strategies based upon pain intensity. Survey respondents regarded distraction as helpful for mild or short lasting pain rather than acute pain. The perceived lack of efficacy, in the postoperative context, was therefore not unexpected. Given the similarities between the number and type of strategies employed by subjects in the two groups, explanations for the statistically significant differences in postoperative pain intensity and distress need to be carefully considered. Experimental subjects had the opportunity to overtly identify their pain relieving strategies on admission and to enter into discussions with nursing staff regarding the necessary resources and mechanisms required to support them postoperatively. As a consequence of this discourse, staff became conscious of the patient’s repertoire and were able to focus practitioner generated strategies appropriately. Thus, for example, there was evidence of consistency in NSAID prescribing for experimental subjects. The preoperative intervention raised consciousness amongst both patients and staff of the possibilities for maintaining personal strategies postoperatively and enabled practitioner generated strategies to be individualised. The overt identification of strategies facilitated a patient-centred approach in which experimental subjects’ distinct pain management actions were valued. The preoperative discussions enhanced understanding of the choices available and enabled subjects to be actively involved in decision making. By approaching the patient as an equal partner, nurses sought to re-balance the locus of power. As a consequence, experimental subjects came to understand that they had “permission” to use their strategies in hospital. This simple action appeared sufficient to empower them to actively engage in their strategies postoperatively. Many subjects were relatively independent in the way in which they engaged in their strategies, however, some required active staff support. It is postulated that the latter group tended to be those who preferred staff to adopt a nurturing role. However, it should be noted that evidence for this hypothesis is based purely upon the researchers’ and nurses’ perceptions which were recorded in discussions following the RCT, not upon patient based empirical evidence. The hypothesis would therefore need testing. Although the above interpretations begin to offer an explanation of the results, they are insufficient in light of the fact that control subjects also used self-generated pain relieving strategies postoperatively. Thus further rationale is required. Discussion in chapter seven addresses this issue, linking the findings to theories about self-efficacy (see section 7.3). Walding (1991) hypothesised that powerlessness may be causally linked to increased postoperative pain and anxiety. Experimental subjects showed a statistically significant decrease in anxiety during their hospital stay. Enhanced self-efficacy and perceptions of control may provide the explanation for this observation. At just over 2 points on a 0 - 24 scale, the reduction in anxiety, although statistically significant, was small. In determining the sample size it may have been helpful to undertake additional power calculations on postoperative anxiety. To ascertain the clinical significance of the result, comparisons with perceived control may be informative. Unfortunately, the data collection instrument used to assess patients’ perceptions of where the balance of control lay, was flawed in relation to the question being asked. Subjects were asked to score, on a 0 - 10 scale anchored by the statements “not at all in control” and “completely in control”, their perceptions regarding self control and practitioner control. If a linear scale is to be employed, a more pertinent approach might be to use a scale anchored with “nurse/doctor in control” and “patient in control”. This could provide insight into the locus of control, patients’ perceptions regarding power balances and their perceptions of self-control. However, many factors influence power and control and a more sophisticated tool, or qualitative interviews, may offer better insight than a linear scale. Further work is needed to assess the relationship between perceptions of control and anxiety before clinical significance can be ascertained. Despite the statistical significance of the results related to pain intensity and distress, the mean differences were again small (1.21 for intensity and 1.17 for distress). One could argue that an improvement of one point, on a 0 - 10 scale, could be important to individual patients. Understanding more about the extent to which improvements are acceptable to patients is an area where further work is needed. The perceptible alteration from the nurse as expert in technological pain management immediately postoperatively, to patient as expert in self-generated pain relief once morphine had been discontinued, may be explained by the literature which suggests that patients who are acutely ill tend to be less participative in their care (Elliott and Turrell, 1996; Caress, 1997)). Equally, the effect of the anaesthetic and morphine may have lessened experimental subjects’ need and ability to utilise their non-pharmacological strategies in the first few days following surgery. As the anaesthetic wore off, and the use of morphine reduced, it appeared that subjects became actively engaged in their pain management by employing self-generated strategies and seeking personally relevant practitioner initiated strategies. The differences observed in analgesic use, such that significantly fewer experimental subjects received IM/oral opiates, while significantly fewer control subjects were administered NSAIDs may reflect the more individualised approach to subjects in the experimental group. The cost benefit of the difference needs to be analysed, however this was beyond the scope of the current study. The difference in length of stay between the two groups was not statistically significant. However, as previously mentioned, a larger sample might have obtained different results. Again, it would have been useful to undertake power calculations on length of stay when determining the sample size. One of the main differences between the experimental and control groups was the finding related to strategy mean helpfulness scores. A total of 30 strategies were listed on the Strategies Questionnaire and the experimental group had higher mean scores for 17 self-generated and 18 practitioner generated strategies. Overall, subjects from the experimental group perceived their strategies to be more effective in relieving postoperative pain than control subjects. This could explain the differences observed in pain intensity and distress. Again these issues are discussed further in chapter seven. 5.5 Conclusion Overall, the survey indicated that self-generated strategies are used to relieve acute pain. Some 43 strategies were identified and collapsed into 10 themes that were generated from the data rather than the literature. The results of the randomised-controlled trial indicated that identifying and supporting patients’ self-generated strategies significantly reduced postoperative anxiety, pain intensity and distress and altered analgesic consumption. Given that control subjects also used self-generated strategies it is concluded that patients need to be proactively engaged in their pain management if they are to be empowered to use their strategies beneficially. CHAPTER SIX: DEVELOPING NEW PRAXIS: ANALYSIS OF THE NURSES DATA 6.1 Introduction Chapter five presented the results related to patients’ pain coping strategies. This chapter now moves on to consider the nurses’ perspective. Specifically, the data were used to obtain insight into the management of postoperative pain, the potential for integrating patients’ self-generated strategies into this process, and the effects of adopting an action research and critical theory approach on both the nurses involved and on the nursing management of pain. The terms discourse, praxis and ideology are used throughout this chapter in accord with the definitions offered in Chapter three, sections 3.4. Thus discourse refers to a two way process of critical communication involving an exchange of ideas and opinions about ideology, contradictions and praxis. Praxis is regarded as a process of acting upon the circumstances one faces so that change is brought about, while ideology is viewed as the values, beliefs and norms of the culture and its institutions. During the course of the study a considerable amount of data were collected from the nurses. These data were obtained through observation, focus groups, collaborative group and co-action researcher meetings, workshops, questionnaires, researcher logs and field notes. The observation data were presented in chapter four (section 4.4.2) as they provided information regarding the study setting. The remaining data are presented here and are categorised according to three of the four key characteristics of action research, which were outlined in section 3.5. Thus, the focus group results and data from the appropriate field notes are discussed in relation to a category termed ‘the nature of the practice problem’. Data about the process of change in relation to pain and its management are incorporated into the category ‘creating change in practice’. Finally, the effects on the nurses of being involved in the study are considered in the category ‘collaboration’. To enhance clarity in the process of theory building, some discussion of the results is included within the chapter. The remainder is incorporated within chapter seven. 6.2 Nature of the Practice Problem: Analysis of the Focus Group Data and Related Field Notes The study was designed to enable a number of concerns to be considered in relation to pain and its management. Prior to agreeing to take part in the study the nurses were aware of my interest in exploring issues around the use of self-generated pain relieving strategies. During the observational stage, many of the nurses indicated that they were interested in adopting a more multi-dimensional approach to pain management as they found that analgesics were not always effective and pain was not always controlled. However, they commented that time was against them moving forward in practice. Curiosity was also expressed in considering the problem of effective pain management from the patients’ perspective. Although the observational data provided some insight into the nurses’ perspectives and ideologies, the main arena for determining the nature of practice problems related to pain management were the focus groups. The co-action researchers facilitated three focus groups, involving both qualified nurses and HCAs. Ten themes were identified from the data and relevant field notes. These are summarised in Table 48 overleaf. Table 48: The Themes Generated from the Focus Group Data and Related Field Notes Pain Management Goal Pain Assessment Communication and Teamwork Preoperative Information Giving Patient Behaviour  Pharmacological Pain Management Non-Pharmacological Pain Management Self-Generated Strategies Multidisciplinary Working and Organisational Issues Recent Improvements  Each theme is described in turn, with illustrative quotations being provided to highlight the range of beliefs and opinions, those which were commonly held, and those that proved more contradictory. Discussion is integrated with the findings. 6.2.1 Pain Management Goal Debate around the goal of pain management occurred in each of the three focus groups. As one participant acknowledged, all patients on the Ward experience pain, either as a result of surgery or because they are admitted with a rheumatoid flare. This universal experience led pain and its management to be viewed as “an important issue” on the Ward. Nevertheless, contradictory ideologies were observed regarding the goal for postoperative pain management. As one nurse stated: “Some… colleagues seem to think that it’s important that we should aim for the patient to be pain free, whereas other colleagues disagreed with that and thought we should aim for the pain to be manageable”. Those adopting the former perspective appeared to feel strongly that it was part of the nurses’ responsibility to: “Keep the patients as pain free as possible”. The emphasis was on the nurse being in control with the patient perceived as a recipient of practitioner determined pain management goals. Conversely, other nurses adopted a more patient-centred philosophy in which the goal for postoperative pain relief was decided by the patient. Thus it was the degree to which the patient deemed the pain to be at an acceptable level that was perceived as important. Typical comments, which illustrate this perspective, included: “My pain threshold may not be the same as somebody (else’s). I wouldn't like to go over the top. I normally would ask them”. “Manageable or bearable, I think that is up to the patient to say”. Several of the nurses, who appeared to be adopting a patient-centred stance, went on to suggest that it is: “Impossible to achieve total pain relief “ and (We) “will try and keep it back as much as it can, but you can't get rid of it”. These statements suggested that some of the nurses might have presented as being patient-centred when they were in fact reacting to the practical difficulties of managing pain effectively. Indeed, a range of comments were elicited which highlighted nurses frustrations in attaining both their own and the patients’ pain relief goals: “It isn’t very satisfying for a nurse to have a patient knowing they are in agony, your hands are tied and you can do nothing”. “It’s quite upsetting really”. “It was awful because she was twitching in so much pain. Her husband would be there as well and you could feel his pain as well and you know all I could say was I am really sorry, but can you imagine when you see her you actually feel how much pain she is going through”. These nurses appeared to be attempting to balance the limitations they faced in managing pain against an acceptable goal for pain relief. It could be conjectured that by placing responsibility on the patient, the nurses were able to abrogate any obligation to negotiate pain management goals in difficult situations. The literature on patient-centred care suggests that decision making should incorporate a process of negotiation. Within this process, nurses need to identify patient’s ideologies and provide them with sufficient information to enable informed choices to be determined (Avis, 1994; Wilkinson, 1999). Thus, postoperative patients need to be informed that it is important to relieve pain to the extent that it does not interfere with deep breathing and mobilisation (Cousins and Power, 1999). Beyond this, the patient should be empowered to determine the degree of relief they require to meet their individual needs. In patient-centred care, the goal of pain management should thus be negotiated in partnership with the patient. Interestingly, the focus group discussions revealed different ideologies in relation to the rheumatoid patients when compared with surgical patients. The former were regarded as having a higher pain tolerance as a result of the duration of their pain experiences and the consequent development of pain relieving strategies. As one participant stated: “Most of them say that the pain doesn't really go away, not completely even when they are sort of stable, the pain doesn’t go away completely. And they like it as long as they can sort of tolerate and they can move about, they seem to be quite happy with that”. Words such as “like”, “prefer” and “happy” were used regularly by those commenting upon the pain experienced by rheumatoid patients. However, on being challenged, participants were unable to explain why they used these terms. One or two suggested that they would: “…need to go away and think about that and decide if I really mean it. Maybe I might ask them.” The potential for focus groups to encourage reflexivity was demonstrated within this quotation. 6.2.2 Pain Assessment Data about pain assessment incorporated discussion on both informal verbal assessment and the formal use of pain assessment charts. With respect to verbal assessment, the point was made that: “Sometimes you look at patients and you think they are in pain and they are not, so I often say to them I cannot see your pain. I can only sort of think ‘oh you must be in pain but I don't really know for sure. So it’s up to you to really tell me’”. There was general agreement that patients were the sole authority on the presence of pain and that they therefore needed to be encouraged to inform nurses of any pain they were experiencing. These beliefs reflected McCaffery’s (1968) assertion that pain exists whenever the patient says it does. Verbal assessment relied mainly upon questions such as: “I ask them about their pain, how they are coping, if it hurts” “They say well it’s terrible, I say what do you mean by that?” One or two nurses commented that they would ask patients to imagine a 1 – 5 scale, with 1 being no pain and 5 severe pain, and to score their pain accordingly. Ideologically, this approach was similar to the formal pain assessment scale used with patients on PCAs and epidurals. Thus nurses were transferring practice from one specific pain management context to another. Their rationale for doing so was that without a scoring system the nurses felt that: “None of us have any proof about intensity, the patient’s pain is just interpreted, we just go on what we see and hear”. This comment was immediately succeeded by the following discourse: “We should have a general pain chart for each patient rather than just for postoperative PCAs and epidurals”. “I think we should have, it depends. We should have like different types of pain charts. There’s one where you have got the body that I have seen for use in palliative care and the patient can say exactly where their pain is and if it becomes worse at certain times of the day, and then...” “Nurses perceive it as a totally different area of the body sometimes” “But then it’s up to the patient, it’s up to the patient to actually mark it themselves because they are the people they know where the pain is, we can't see it”. Similar discussions were observed in the other focus groups so that overall there was general agreement that a formal approach to pain assessment was required. Most of the nurses were aware of the literature that highlights the benefits of regular, formal pain assessment (for example: Gould, 1992; Waterhouse, 1996; Melzack and Katz, 1999) and there was a feeling that they “ought” to be applying this evidence in practice. Pragmatic comments were also elicited in relation to the ability of pain assessment charts to enhance communication between patients, nurses and doctors. For example, several nurses suggested that if patients completed pain scales the following conflicts could be lessened: “I had one instance with a lady, she was newly diagnosed with cancer, and when the doctors arrived she would say ‘I'm not in pain’. As soon as he had gone, five minutes later she would say, ‘I am in pain’”. “If we had a chart you would have something to show them if they questioned your assessment”. “The doctors actually don't believe you, what you are saying because the patients always say ‘no I'm all right’ when the doctor arrives”. The issues and conflicts surrounding doctor/nurse communication are further addressed in section 6.2.9. Finally, there were some discrepancies regarding whether or not pain assessment charts were currently available on the Ward. Staff nurses who had been on the Ward for some time seemed to think that they were available, as they could recall using them in the past. No reasons were elicited as to why this practice had stopped. The following section considers communication within the nursing team. 6.2.3 Communication and Teamwork Overall, communication regarding pain and its management was considered to be good within the nursing team. However, for the HCAs there appeared to be some distortion in relation to teamwork. These distortions are considered following discussion of the more generally held opinions. Typically, nurses felt: “… very happy about the way we communicate within the team”. There was evidence to support the findings, noted during the observation period (see section 4.4.2), that pain and its management are discussed during nursing handovers: “It will come across in the handover. We say that so and so isn't coping with the pain, I've left the chart out”. Communication also took place via care plans, although the main avenue appeared to be through discussions during the shifts, as typified by the following comments: “If there is a particular problem with a patient then we will discuss it”. “If for some reason we cannot control somebody's pain we will discuss the best way to approach it”. Effective teamwork was evident and nurses frequently received feedback from patients that enhanced the team approach. For example, one nurse stated: “I think you can tell when you are working with somebody because obviously you work very closely together. And feedback from the patients: ‘so and so … offered me painkillers on the drug round but I didn't need them then, but they said I could have them now, is that all right?’ and things like this”. For the HCAs, many of the difficulties arose from their ability to recognise when a patient was in pain, combined with an inability to administer analgesics. They perceived the qualified nurses as caring and proactive in terms of pain management, but frequently distracted from prompt action due to the conflicting demands of a busy ward. Thus, HCAs would report the patient’s pain to the staff nurse and then be forced to wait for action while continuing to observe the pain and having their own work schedule disrupted. Comments included: “It's not my role to a certain extent, but I am aware of the patient having pain”. “You can only do so much by yourself. I sometimes find it frustrating - the situation where you are nursing somebody in pain who needs analgesia and the staff are busy”. “A nurse can only do so much, one thing at a time and she can't be everywhere, but they are. I find them very good on the ward, very caring”. There was also some indication that the emphasis upon pharmacological management disempowered HCAs from engaging in discussions regarding pain management or employing non-pharmacological strategies. For example, one HCA talked about feeling alienated from pain management, while others commented: “You can't always do something about it and you don't”. “You don't get to give anything”. “I can't talk about that because I'm not a nurse” It should be noted that qualified colleagues responded to this later participant with comments such as “of course you can!” The HCA then went on to offer further opinions regarding pain and its management. 6.2.4 Preoperative Information Giving The general perception, through all three focus groups, was that patients are now better prepared for surgery as a result of nurses providing both more information and increasingly focused information. As a consequence, the nurses believed that recovery has improved and that they are applying research to practice. There were indications that cognitive and sensory information were offered in a patient-centred manner through both verbal and written means. In addition, the influence of anxiety on patients’ ability to internalise information was recognised. The following quotations summarise the type of statements offered: “I feel they are better in a lot of ways because at least we spend time talking to them now, whereas before it was ‘you are having a major operation tomorrow, don't drink after midnight’, bang off you go. At least now you go through everything with them, who's going to see them, what's going to happen to them and how many drains they are going to have. So they know what they are going to have, what it is going to be like”. “They have post-operation information beforehand so they can sit and read quietly, what happens about the knee, what’s happening to them afterwards, at least they have got time to take it in”. “I say things like ‘when you have first had your knee done, and you get out of bed, you are going to have pain shooting up your leg’. I think that because we educate them better they are a bit more relaxed so perhaps that helps”. “Sometimes when people are anxious preoperatively they don't hear”. “You explain why you are doing it. You try not to do it too much because obviously that makes them anxious”. The literature is not clear about the impact of preoperative information giving on postoperative pain and anxiety, as factors such as personality and coping style may influence the outcomes (Weisenberg, 1999). Nevertheless, studies have indicated that patients who are given sensory information tend to have less pain, fewer analgesics and a shorter stay in hospital, than those who are either not prepared, or who are only provided with cognitive information (Cousins and Power, 1999). It could be argued, however, that the assertion made at the beginning of this section, namely, that the nurses offered information in a patient-centred manner cannot be claimed if coping styles are not assessed. Studies indicate that patients who use active coping styles tend to do better with more information, while those who prefer passive styles are better with less information (Weisenberg, 1999). If these findings are validated in the future, true patient-centred care will be dependent upon coping style being assessed prior to the provision of preoperative information. There was some awareness shown by the nurses that different coping styles may influence the degree to which patients seek information. For example, one nurse stated: “I don't know that people are getting enough preoperative information about you may have a catheter, you may have this, because it is difficult sometimes to determine whether people want to know about these things”. The following section considers data about patient expectations and behaviour in more detail. 6.2.5 Patients’ Attitudes, Perceptions and Responses to Pain This theme incorporated discussion about changing attitudes and expectations among patients regarding pain and its management. Generally, nurses believed that: “Patients are more aware (now) of pain and methods, not just medication methods”. “They realise the importance of rest and activities and elevating their legs and general positioning in a comfortable way”. “Patients are more open as regards the pain, while before patients… weren’t really complaining”. However, discrepancies were noted in relation to the age of the patient, with older patients regarded as less open and less likely to complain than younger patients. Several reasons were offered and these are neatly summarised in the following quotation in which a nurse offers her views from the perspective of a patient: “I've put up with this for years and I know you are very busy and I don't want to bother you”. The perception that most elderly patients admitted to the ward have already experienced pain for some years was therefore regarded as pertinent, as was the view that older people appear to be less assertive in seeking help from apparently busy staff. Associations between the type of surgery planned and expectations of pain were also linked to age. Thus, elderly adults undergoing hip and knee surgery were regarded as expecting pain after surgery, while younger patients having spinal surgery were said to expect to be pain free following the operation. The fact that they are frequently not was seen to add to anxiety in this group and was therefore an area requiring specific preoperative education. It could be conjectured that the observed difference in complaints between older and younger people is not only linked to past experience, psychological attributes (such as willingness to label stimuli as painful) and sociocultural issues, but also to the physiological perception of pain. A laboratory study by Chakour et al (1996) examined pain perception following radial nerve blocks in 15 young and 15 elderly subjects. Young adults demonstrated significant increases in threshold for thermal pain during A-fibre block compared to the elderly group. Together with other results from the study the findings suggested that elderly adults have decreased pain sensitivity and, in particular, are less likely to distinguish A-delta fibre pain. An earlier study by Gibson et al (1991) found that elderly adults show reduced central nervous system processing of painful information. Combined with Chakour et al’s (1996) findings, and those of other researchers which indicate the existence of peripheral changes in older people, it is feasible to suggest that elderly patients may be less inclined to complain of sharp A-delta pain and more inclined to describe pain as dull and aching. Practitioner interpretations of the language used may imply that older people experience less intense postoperative pain, or are less complaining of it. However, the clinical application of these laboratory findings requires further study. Discussion about patients’ behaviour surrounding analgesic management also took place. Many nurses expressed the opinion that: “A lot of them get in pain ‘cos they don't have tablets. You do your drug round and you say ‘do you want them’ and they say ‘no’. You see, you discuss it and if they finally decide that they are quite happy, they are not in pain and they don't want it, then 2 hours later in the middle of chaos they suddenly decide they are in pain. That's not so good. ‘Do you need it immediately?’ ‘Oh yes I really need it’. But it's not so easy to get back”. Conversely, other nurses appeared to have a more patient-centred, flexible philosophy: “I think the patients feel they can ask for pain relief as well because I know when I go round I ask them if they want analgesics, but if they refuse it then I say they can always ask for it at another time. And it’s not a problem to us”. The contradictions observed during these debates were associated with those that occurred when the pain management aim was discussed. Thus, nurses who demonstrated patient orientated ideologies with one topic were likely to re-confirm their stance as the discussions progressed. 6.2.6 Pharmacological Pain Management Not surprisingly, given the surgical nature of the study setting, a considerable amount of discussion took place regarding pharmacological pain management. In accordance with other comments, which highlighted the priority given to pain and its management, nurses emphasised the importance of administering analgesics promptly and regularly, both during and between drug rounds. Typical comments included: “Staff on the ward are keen to... make sure that (patients) have regular analgesia”. “(We) always ask the patients whether they have got any pain and what they would like for their pain, make it an issue on the drugs round”. “The majority might be given on the drug round but there's no constraint”. “I think it’s like we make that a priority. We say we need somebody out to check a drug or something and come and do it and we do. It’s almost like we leave everything else if we can and go and do that. Don't sort of hang around, get it done”. However, during the night shifts it was not always possible to maintain the same degree of immediacy. As one nurse commented: “Because there’s only ever two trained nurses on, when one goes for their break you are on your own, you can't check controlled drugs on your own”. Furthermore, the reduction in staff numbers at night makes it more difficult to offer analgesics flexibly: “That's where it falls down because when you are trying to do them at odd times that's when you fail the patient, because you get so many interruptions you don't get back immediately as you should. So by the time you get back they are in a lot more pain than they were to start with”. The organisational difficulties experienced in attempting to offer speedy and effective pain relief at night are considered further in section 6.2.9, while the suggestion that pain management may be professionalised is discussed in chapter seven. There was evidence that nurses tried to individualise analgesic management: “Obviously if there is someone who's in pain you give them analgesic. We've got one patient at the minute who likes his tablets the minute his eyes are open, so you are watching for him, from 5 o'clock, so the minute he opens his eyes he gets all his personal own drugs, 'cos that's what he wants”. This was taken further with rheumatology patients where the nurses actions in facilitating self-medication enabled normal routines to be continued and established the patient as the pre-eminent team member in their on-going pain management. “We do let them self medicate actually, quite a few in rheumatology do, and they do it quite well. The trouble with post-op patients is that it's not so easy”. “Some of them are a bit elderly and confused. You'd have to be a bit careful, but rheumatology patients they're sort of orientated. They'd prefer to do it because we muck up their regime. Basically, we change it to suit the ward routine and it doesn't suit them and they prefer to do it themselves”. Improvements in nurses’ pharmacological knowledge were regarded as an important development in enhancing pain management. Nurses felt that some of the distortions that had previously surrounded analgesic management had been corrected, that approaches such as combining analgesics and NSAIDs were understood, and that the fairly recent innovations of PCAs and epidural analgesic delivery systems were important developments. The impact of the latter in enabling circulating levels of analgesic to be stabilised was recognised as an improvement over the peaks and troughs that occur with intramuscular injections. Typical comments that highlight these perspectives included: “There’s not so many myths floating around as there were when I was a student. I think that has changed a lot”. “I think with the information coming out now that you can mix the two (analgesics and NSAIDs), that seems to have helped with pain management”. “I think the main thing has been the PCAs and the epidurals to be honest, because you are not back to this thing of repeat troughs and falls in analgesics”. There was a sense throughout the three focus groups that the Ward nurses were not frightened of administering analgesics, were assured in their knowledge base and were confident in approaching many of the medical staff to discuss prescription issues. However, the limitations of current pain management practices were also discussed. The range of comments elicited are summarised in the following four statements: “Sometimes they only write them up like 7.5 milligrams of morphine. I mean mainly because they think if they wrote 7.5 - 15 we might just go straight in and give 15 milligrams and not try the 7.5 first of all”. “If I bring a patient back from theatre I always check to see what they are written up for and I won't take them out of recovery until somebody has written up”. “We should be able to control pain, but it’s not always possible”. “It’s not always effective. We could do with using other things as well to treat the patient as a whole”. Three areas of difficulty were noted regarding pharmacological management. One of these is discussed in section 6.2.9 in relation to multi-disciplinary team working. The other two referred to the side effects of opiates and problems experienced with epidurals. The issue with opiates was the potential for patients to develop constipation. That this side effect was picked out was of some surprise to the co-action researchers, both of whom thought that nausea and vomiting would prove to be the more contentious issue. This premise was based upon the literature and personal clinical and research experience (Burrows and Whybourne, 1996), the fact that patients’ average length of stay was fairly short and the considerable emphasis upon mobilisation. Nevertheless, the following discourse, taken from one of the focus groups, highlighted the perceived problem. “A lot of our patients are worried about their bowels and being on analgesics and stuff. I do try and encourage them to drink plenty of fluids, but even so they will refuse analgesics if they are in pain because they will say ‘I have not had my bowels open’”. “Who was it now? Tom somebody. He said to me I am not having painkillers, I've not been to the toilet, and he didn't have them for days and he was in agony”. Laxatives and suppositories were prescribed and administered when necessary. However, the staff preferred to try and manage potential constipation by encouraging fluids, a high fibre diet (“within the constraints of the hospital”) and active mobilisation. These methods were perceived as being more consistent with patients’ normal routines, less physiologically intrusive and less likely to lead to problems in older adults following discharge from hospital. Two problems were mentioned in relation to epidurals, the first being pruritis and the second urine retention: “Pruritis, you find that the patients pain is well controlled but sort of itching all over because of the amount. You phone a doctor and they say ‘right take it down’ and you would find that they were taking down epidurals when it was at three mg an hour and then they would put the patient on to oral opiates and that wasn't really working at all. So I think that that is one of the difficulties”. A small study by Burrows and Whybourne (1996) on criteria for discontinuing PCAs, also found that pruritis was used as a reason to discontinue IV opiates. If immediate postoperative pain is to be effectively managed, anti-histamines should be prescribed and PCAs or epidurals left in place until more valid criteria, such as pain scores and the amount of opiate being used, warrant their discontinuation. Urine retention was probably the more difficult problem for the nurses to manage. Many patients who are administered epidural analgesics postoperatively are catheterised in theatres. However, at the time of the focus groups, the local consultants were reluctant to catheterise orthopaedic patients because of the potential for introducing infection. Consequently, anyone who developed retention had to be catheterised on the Ward by a nurse or junior doctor. Not only was this traumatic for patients and time-consuming for the staff, but there was a perception that catheterisation under these conditions offered a greater risk of infection than the controlled environment of the theatre suite. The nurses estimated that 80% or more of orthopaedic patients with epidurals were requiring catheterisation on the Ward. However, they were just about to commence a study with the Acute Pain Service and Audit Department to: “Prove to the orthopaedic surgeons that this is an ongoing problem”. “We will hopefully have the statistics to prove to the orthopaedic consultants that there is a need”. Chapter four highlighted that the Ward nurses were committed to research and research based practice (see section 4.4.2). The statements above offer a clear example of this commitment. It is interesting to note that the nurses were looking for “proof” from statistics. It may be that their determination to undertake the RCT was consistent with early recognition of the need to identify not only the best possible evidence, but also evidence that the medical staff would value. The following section considers the participants’ discussion on non-pharmacological approaches to pain relief. 6.2.7 Non-Pharmacological Pain Management The highest number of responses was offered in relation to non-pharmacological pain management. Although participants were directly asked “what non-drug interventions have you thought about using, or actually used when managing patients’ pain”, the range and number of responses were not anticipated in advance. Part of the reason for the interest was the potential of combining analgesics, with other strategies, in an effort to manage pain from a multidimensional perspective. One particular statement typifies the range of strategies discussed: “(You) need different therapies, you know, it’s not necessarily just drug therapy, its positioning, heat, warmth just someone to talk to, touch. It’s so we can sort of manage it from different ways”. Several of the themes identified during the patient survey were noted, in particular, cognitive and expressive reassurance, positioning, mobilising, strategies from the behavioural theme and a variety of complementary therapies which were incorporated into the miscellaneous theme during the survey analysis. The most frequently mentioned theme was cognitive and expressive reassurance. Table 13, which was located in section 5.2.3, highlights that this theme included verbal reassurance and expressive touch. These were the main interventions mentioned during the focus groups. Illustrative quotations related to verbal reassurance included: “Taking the time, reassuring them. They know then that you do care and they are not frightened then of telling you later on if they get pain. They know you are going to be there and listen to them again and do something about it. They feel a bit more secure like that”. “Try as much as I can to stay with the person, but if there are two or three that all need something, but I do try and stay with them as long as I can or I get one of my colleagues to sit with them”. “What do you think it is about sitting with a person that helps pain relief?” “I think they think hopefully that we are trying to do our best. They are on their own as well, frightened”. McCaffery (1981) discusses the importance of “being there” for patients in pain in an old, but insightful opinion paper entitled “Don’t just do something, sit there”. Based upon reflections on her own clinical experience, McCaffery (1981) advocates that nurses develop the ability to stay with patients in pain without necessarily performing nursing actions. She suggests that the benefits include keeping the patient in contact with other human beings, diminishing psychological isolation and anxiety and enhancing feelings of being cared for. The nurse’s presence may also act as a distraction from internal pain stimuli and the monotony of being in hospital. Nurses on the Ward recognised the potential of their physical presence and were in agreement regarding the use and perceived benefits of verbal reassurance. However, some dissonance was apparent in the discussion about touch as the following discourse highlights: “Touch I think it is very important, but again you have got to be aware of whether the patient wants you to do it or not. I think if you know the patient well then I think you just use the touch, its quite an effective way of trying to relax somebody”. “What sort of touch?” “Just a hand on their hand, you know just something very slight, you feel they are just not quite in so bad pain”. “I don't know about touching. You don't know how far we can go”. “I did it to a gentleman”. “I think you have to be careful with the opposite sex”. Touch was mentioned in each of the focus groups and was clearly viewed as a potentially useful strategy for enhancing pain relief. On the whole, it was the newly qualified, young, female nurses who were more concerned about applying touch with male patients. Older nurses and HCAs were much more confident. The RCT results indicated that experimental subjects recorded cognitive and expressive reassurance as the fifth most frequently used practitioner generated strategy, while control subjects recorded this theme sixth. Although other strategies were more frequently administered, items from the cognitive and expressive reassurance theme received some of the highest efficacy scores for practitioner generated strategies (see Table 47 and section 5.3.4). Nurses clearly recognised the potential benefits of these interventions. Positioning and mobilising were also frequently mentioned strategies during the focus groups, which was not surprising given the emphasis upon orthopaedic surgery and the medical management of patients with rheumatoid flares on the Ward: “We make them as comfortable as we can by giving them a soft mattress, letting them have bed rest if they need it or even, after the op we get them up to make sure they start using the new replacement they've had”. “We take care to position the patients to lessen their pain”. “Before it used to be 4 to 5 days in bed and it was so uncomfortable to do that, whereas, at least now it's in bed a day and then up. It helps them a lot, especially in that area, with their arthritis it stops them stiffening up”. The nurses’ awareness of the potential for mobilisation to reduce pain was built upon their experiences with both rheumatoid and osteoarthritis patients. Liaison with physiotherapists had also enhanced their understanding of the benefits of mobilisation. A variety of behavioural strategies were mentioned. These included warm baths which were perceived as relaxing, reducing joint stiffness (even if only temporarily) and enjoyable. Physiotherapist generated heat treatments and hydrotherapy were also commented upon. The use of cold was mentioned in terms of placing a cold facecloth across the individual’s eyes and/or forehead, or by using ice packs. Again the physiotherapists were seen as participating in the later pain relieving activity. Distraction was mentioned only briefly and even then its limitations were commented upon: “Obviously if their pain is overwhelming, it’s not going to work”. The RCT findings indicated that subjects in the control group did not view distraction as an effective strategy for postoperative pain management, while those in the experimental group perceived it to be only fairly helpful. The nurses regarded music more positively and patients were encouraged to bring in Walkmans from home, or listen to the hospital radio. The limitations of the latter were recognised and a suggestion was made which the Acute Pain Sister offered to follow up. The following discourse was extracted from the discussion: “There’s not a station on our unit with headphones that you can actually play the music through all the time”. “I think in the whole hospital we ought to have a relaxation channel”. “Yes that would be superb in the whole hospital, not just our ward”. “Perhaps that's something we can look into”. “I think it's good. I think it's excellent”. “I think it's a very good idea yeah”. “It would certainly be cheaper than painkillers! And you don't have side effects” Grenside’s (1997) systematic literature review, which was referred to in section 1.6.1, concluded that music might enhance the ability of patients to cope with their pain, relax and gain affective reassurance. However, she warned that as music preference is highly individualised, patients need to be offered choices. The nurse had noted that the hospital radio tended to play popular music and felt that a channel, which offered purely instrumental music, would widen choice. The involvement of physiotherapists in pain management was mentioned further in relation to the use of transcutaneous electrical nerve stimulation (TENS). Although there were three TENS machines available on the Ward, not all staff were aware of this and, in any case, regarded TENS as being within the physiotherapists rather than nurses remit. Thus, most nurses only used TENS once the physiotherapist had explained it to the patient and begun using it. The exceptions to the rule tended to be senior staff nurses who were confident practitioners and had learnt how to apply TENS correctly. A previous study by Burrows (1997b) identified that although staff were ideologically open to employing non-pharmacological pain relieving strategies, many lacked the knowledge and confidence to do so in practice. Carr (1990) suggests that nurses should be able to offer behavioural techniques as a matter of course. However, care must be taken with most interventions. For instance, ice placed too close to a wound may delay healing, while guided imagery may result in unpleasant associations if care is not taken to individualise the image. A certain amount of guidance is thus necessary for safe practice. The theme resting alone was acknowledged by staff but only to a minor extent in one of the focus groups. The purpose was regarded as promoting relaxation and no awareness was shown that resting alone could constitute a deliberate pain relief strategy for some patients. Conversely, considerable discussion took place regarding the use of complementary therapies, specifically aromatherapy and massage. Interestingly, survey respondents did not mention complementary therapies as frequently as had been anticipated. Indeed, only 1.5% of the survey sample stated that they had employed these techniques for pain relief. The amount of discussion in the focus groups may have occurred because aromatherapy and massage had previously been available on the Ward and the Consultant Rheumatologist had actively supported their use. Nurses appeared to feel frustrated that they were no longer able to offer these interventions because staff had left and no further training opportunities had arisen. Nevertheless, considerable awareness was shown of both the potential and risks of aromatherapy and massage: “We had a big chart on the wall and it explained herbs that were effective in different areas of the body. The patients found it quite relaxing even to look at the chart”. “It has got to be administered properly because you can overdose on a couple of the drugs”. One additional strategy theme was noted in the nurses’ focus groups in comparison to the patient data. Although expressed through activities, which could be linked to the general help theme, the interventions appeared to contain a psychological element related to enhancing patients’ morale as a deliberate pain relieving strategy: “We wash the hair and put their rollers in and they are so relaxed it does help sometimes”. “Especially if they have been very tense with their pain and the joints have been very tight, it really does help”. “We encourage them to get dressed so that gives them a boost to their ego”. The nurses were thus using strategies aimed at reducing pain distress and muscle tension and enhancing morale. Literature on chronic pain management highlights the relevance of such an approach (for example: Craig, 1999; Turk and Okifuji, 1999), while McGuire (1992) also comments upon the impact of emotional distress within the affective dimension of her multidimensional framework. Finally a number of general comments were elicited which linked non-pharmacological management to self-generated strategies: “There seems to be more methods, a lot of things both from nurses and patients”. “To ask them... what would they normally do if they ever had this kind of pain before, sometimes they have their own sort of strategies for dealing with it, which can be really helpful”. The following section considers nurses’ ideologies regarding patients’ self-generated strategies. 6.2.8 Self-Generated Strategies Again, a direct question was asked to elicit participants’ ideologies about patients self-generated pain relieving strategies. The nurses acknowledged that previous experience of pain prompted many patients with rheumatoid arthritis to develop self-generated strategies. As one nurse suggested: “You can try and help them sort of achieve whatever it is they normally do at home”. A variety of non-pharmacological techniques were mentioned, including music, distraction, homeopathic remedies, warm drinks and baths. Although nurses were not always convinced of the efficacy, or indeed safety, of some self-generated strategies, they were inclined to accept the patient’s perspective. For example, one nurse commented: “I talked to a patient the other morning who had quite a hot bath, which worried me a little, but she would have it and she actually walked so much better”. The use of creams, by rheumatoid patients, generated discussion in each of the focus groups as the following discourse highlights: “They love creams rubbed on their legs”. “A patient said to me, ‘Would you get that cream out of that draw, please?’ and I went and asked sister actually, 'cos I always do that, always cover myself. Took the cream to show it, that's fine, that's correct, and it was just massage it gently, and I massaged it and it hurt her, it was a little bit hard, I did it gently and that's fine”. “Was that her rheumatic cream, one of those really foul smelling balms, she uses that for deep heat”. “Oh so it creates heat does it?” “She uses it at home all the time”. “This patient really has been in quite a lot of pain recently and it really did help her”. Different ideologies appeared to surround pain management for patients undergoing surgery. Orthopaedic patients were regarded as more dependent upon nurse initiated strategies. In chapter two, it was argued that while consideration has been given to incorporating patients’ self-generated strategies into chronic pain management, little attention has been paid to their potential in acute pain management. The nurses seemed to believe that because surgery was a new and possibly unknown event, patients would be limited in their ability to transfer strategies to the acute setting. Equally, they seemed to think that surgical patients might not: “be conscious of what strategies (they) use”. Whether these perceptions arose from the nurses’ need to be in control of the more technological management of postoperative pain was difficult to ascertain from the data. Interestingly however, the potential impact on orthopaedic patients of altering their analgesics was acknowledged as the following quotation highlights: “Sometimes you find that the patient is happy with one type of analgesic before they come into hospital. They have confidence in that analgesic and then the doctor changes it for something completely different and, maybe it could be psychological, but the patient said it doesn't work”. Assessment of self-generated strategies was either carried out informally as part of general conversation, or not undertaken. The focus group discussions appeared to encourage nurses to critically reflect upon this approach and began to raise their consciousness regarding the potential for new praxis. The discourse offered below is taken from one focus group and typifies the direction of the reflexive dialogue: “I don't think people make a conscious effort to ask”. “It comes out in conversation, doesn't it? To be honest I don't know why we don't” (formally assess). “No idea. We had a lady in recently that had intermittent acute back pain and she was sitting absolutely rigid in a chair and I was asking her if she would like to lie down to help her pain. I asked her then how she coped with it at home, what she was doing. She would sit and wait till it went away. She didn't have any coping mechanisms for it at all, some people do”. “There is a possibility of using this in the future. Because in the assessment it would be quite good to say what they have used. Because if you are in severe pain, you are not going to be likely to communicate are you? You are just thinking about the pain”. “It’s like a man I had yesterday, he can hardly walk as it is, and I said to him do you take a painkiller he said ‘no I don't need them I have got used to it’. So I think you have to assess how they are coping, the way they cope with it”. (We need to) “try to ask them when they come in but it doesn't always get done, because you know the forms we use. I don't think people do them properly”. “Because you've only got a little space”. “I am hoping that the new forms will help”. “We could have questions about, "What if anything do they do at home for pain" that could be done on the self-assessment form couldn't it?” This discourse highlights the nurses’ recognition that if patients were to be enabled to use their self-generated strategies in hospital, the strategies would need to be systematically identified. Current informal practice and the limitations of the existing assessment forms, were regarded as frustrating attempts to engage patients in their own pain management by combining self-generated strategies, with patient-centred pharmacological management. Thus, existing ideologies began to be challenged and there was an increasing recognition of the contradictions and constraints within the organisational system. Other organisational issues are outlined and discussed in the following section. 6.2.9 Multidisciplinary Working and Organisational Issues Ideologies surrounding multidisciplinary working elicited the second highest number of responses. However, when combined with organisational issues, this category clearly generated the largest amount of data highlighting the passion and frustrations experienced by the nurses in relation to multidisciplinary working. The discussion considers positive elements within the system as well as the contradictions and oppressive features which were viewed as impacting upon patients’ pain management. Overall, nurses felt that communication had improved between different members of the team. They also believed that: “There could be a little bit more of a multidisciplinary approach”. There was a sense in the data that the contradiction expressed above arose from differences between multi and inter-disciplinary working. Multidisciplinary working involves members of different disciplines working relatively independently towards an apparently common goal (Miller et al, 1999). Conversely, interdisciplinary working involves a number of branches of learning collaborating in a mutually respectful manner (Parsell et al, 1998). A shared language and reciprocity typify this relationship (Richardson and Edwards, 1997). The data suggested that it was the latter style that the nurses would like to see in relation to pain management. The ensuing discussion commences with a section on the relationship between nurses and physiotherapists, before moving on to consider that with pharmacists and doctors. Data about organisational issues within the nursing team are then considered. 6.2.9.1 Physiotherapists Communication with physiotherapists was viewed as having improved recently, as the following discourse highlights: “Just recently I have said to the physios ‘you know this patient is having quite a lot of pain, we've given them something for pain, you know try mobilising them at certain times’. Usually they like to do it sort of late morning or early afternoon. Before, you know, they never used to say anything about pain. But you find the physiotherapists come in and say things like ‘I was mobilising the patient and they are in a lot of pain. Could you do something or look at their pain, their analgesia, just review’”. “Before that wasn't the case. I mean sometimes you find with the physios that they wouldn't take a lot of notice”. “Now they just sort of come and they say ‘well what’s happening about the patient’s pain’. I mean they will mention it themselves, whereas before they never did”. Data about why this improvement had occurred suggested that it arose as a consequence of shared working practices. Nurses had been working with physiotherapists to continue patients’ rehabilitation over weekends when the physiotherapists were unavailable. As a result, patients undergoing hip and knee replacements, in particular, were kept mobile and physiotherapists were able to continue progressing with rehabilitation when they returned to the Ward on Mondays. The one contentious issue, related to nurses’ comments that physiotherapists would not mobilise patients with PCAs. As one nurse stated: “They do tend though, if they see a PCA, (to say) ‘I'll mobilise tomorrow’”. Consequently, the nurses believed that they had a responsibility to educate physiotherapists regarding the importance of early mobilisation – a stance, they felt, the physiotherapists should already be aware of. As mentioned previously, the study by Burrows and Whybourne (1996) identified that although the pain relief afforded by a PCAs should enable early mobilisation, many health care professionals find the equipment impedes their ability to mobilise patients and thus rationalise their behaviour by discontinuing PCAs early. On the study Ward this was clearly not happening, the PCAs were left in place until patients no longer required them. However, the effect was a delay in rehabilitation and tension between nurses and physiotherapists. 6.2.9.2 Pharmacists The working relationship between pharmacists and nurses was viewed very positively as the following quotations indicate: “During the day having the pharmacist on the Ward so that you can ask her all sorts of things”. “The pharmacists now leave notes on the drug charts and say, you can't mix these two”. “They will tell you which drugs work well with the other and which you shouldn’t be giving, ‘you shouldn’t have this if they are on that’”. The suggestions offered by pharmacists were regarded as helpful and “not threatening”. Why these particular words were used was unfortunately not explored during the focus groups. It may have been that the pharmacist’s lack of challenge and their commitment to sharing knowledge that impressed the nurses in this situation. One organisational issue was identified as causing difficulties for nurses in their attempts to provide effective pain management, although again the pharmacists were viewed as sympathetic and helpful. The issue related to the need for certain patients’ drug charts to be looked at within the pharmacy department. Staff commented that: “When the charts go to pharmacy they take forever to come back and that has been a problem”. However, some dissonance was noted among the nurses, as several of them became quite defensive of the pharmacist and the system: “The pharmacy has said if we ring down, they will come back to us straight away”. “They have actually got somebody to bring the chart back to us in quarter of an hour”. Quite why this difference existed within the team was not clear from the subsequent discussions. 6.2.9.3 Doctors Data about nurses’ relationships with medical staff were typified by discourse that highlighted power balances and power games between the two disciplines. There was a sense that each was struggling for control and that the nurses were full, and surprisingly confident, participants within this struggle. Nevertheless there was a perception that the relationship with medical staff was gradually improving, that doctors were increasingly likely to listen to nurses suggestions and that nurses were particularly committed to welcoming new junior doctors. However, the language was indicative of potential tension and of nurses identifying the ground rules. For example: “We make them welcome and we tell them we will work with them if they work with us”. “We have the problem again in August. The pain nurse specialist will do her stuff and try and educate them but I think that each time there is a change of doctors we always have a problem”. Each new group of junior doctors was given the opportunity of integrating into the team before judgements were passed. They were expected to learn quickly and to listen and take the advice of nurses who tended to regard themselves as more knowledgeable: “He really doesn't have a clue. I feel he doesn't know what to prescribe and I think that is also a problem with the new doctors”. “He obviously did not know what to write, he couldn't even spell simple aspirin. You know, come on!” “We had the same problem with a doctor from X and it wasn't that he couldn't spell, but he was confused about the spelling of say aspirin and he was mixing like the languages and once you said to him ‘look this is what it is’ I got the BNF and I showed him. That’s what I did with him last week and I said ‘this is how we write this up and this is what it is, check for yourself’ and he said ‘oh yes, I know’". The education and training of new house officers was regarded as a responsibility which nurses shared with their more senior medical colleagues. To some extent, the nurses assumed this role because: “Sometimes when they (the junior doctors) arrive, the registrars aren't there. It’s only the consultant, and the consultant hasn't done these things for years”. “The background for the doctors isn't supportive enough for them. They don't get the information at all”. Consequently, nurses would attempt to work with junior doctors, advising and training them when they undertake their early morning and evening rounds. The doctors who do not learn and comply quickly, are sanctioned by the nurses contacting more senior medical staff: “It’s very difficult about not complying with our requests about analgesia”. “I had one problem before where the doctor wouldn't come out to change a prescription and I just said ‘I will just phone the registrar’ and he said ‘do that’, so I did and I got the result that I needed”. “I have been known to call the next one, the next one higher up and if not, the next one higher than that”. It was acknowledged that junior staff nurses might not have the same level of confidence to contact senior medical staff. However, this view was countered within the following statement: “You really have to push. You have to be quite assertive to get things done for your patients. That is quite hard because I am quite newly qualified, but you soon learn to be like the others, thank goodness, otherwise the patients will suffer”. Senior medical staff did not escape the power games as the following scenario highlights: “They had started her on MST and the consultant wanted to cross her off. I took (the drug chart) from him and I said ‘you are not doing that’, I said, ‘because if you had given the patient morphine that’s going to take a while to work so what has the patient got meanwhile?’ I said ‘you normally give a dose of morphine as well as giving the first tablet of MST’ and then afterwards I said to this doctor, I said, ‘if you are not sure why don't you ring up the pharmacist, or why don't you speak to the Florence Nightingale doctors, you know, they are happy to talk to us’”. Overall, the nurses’ perceptions were that most surgical and orthopaedic doctors do not regard pain management as a priority. They are perceived as viewing requests to increase analgesics as a failure of surgical treatment, which should see the patient consistently reducing their analgesic requirement, mobilising promptly and being discharged home. Thus, requests from nurses to increase or alter an analgesic were not welcomed. Furthermore, doctors who were busy elsewhere, for instance in theatres or casualty, were regarded as not having the time to consider pain management a priority. As there were no focus group discussions with doctors it was impossible to test out these beliefs. Doctors linked to the medical, rather than orthopaedic teams, were viewed as committed to high quality pain management. However, the nurses recognised that organisational systems were generally responsible for this difference: “They are not up all day and all night, whereas surgical doctors are up all day, all night and all the next day. The medical doctors only work during the day then another one takes over at midnight”. The night shift was a particular source of tension in relation to pain management. Nurses arrived on duty at 9 p.m. Following hand over and any immediately required action, they would commence the drug round which would often not be completed until 11 or 12 p.m. Doctors, however, would tend to make their last visit to the Ward between 9 – 10 p.m. Consequently, if prescriptions needed altering, nurses had to call the doctors back to the Ward. Frequently, doctors would: “Almost refuse to come out because we didn't foresee the problem earlier. When they get there it’s as if to say ‘why didn't you foresee the problem at 9 o'clock at night when I came round to do a round’”. Although the nurses tried to discuss these problems with new junior doctors and encourage them to either visit the Ward at 11 – 11.30 p.m. or telephone before going to bed, the system did not work well. As one nurse stated: “It is my biggest bugbear”. Tensions also existed within nursing organisational systems at night-time as the following section describes. 6.2.9.4 Nurses Two main areas of difficulty were revealed about nursing organisational systems. As mentioned above, one of these related to cover during the night shift. The other was linked to the team system used during the day shifts. At night-time, the main difficulty was the requirement for two qualified nurses to check controlled drugs. Occasionally the Ward was covered by just one registered nurse working with two or three HCAs. Even when two qualified nurses were on duty there would be at least two hours during the night when only one was available due to the other being on a break. In this situation, staff might attempt to get a nurse from another ward, or contact the night nurse practitioner. However, colleagues were not always available and, even when they were delays of 15 – 30 minutes were common. Nurses found the situation frustrating, time consuming and for the patient: “Well they are not getting their fair share of what they should be getting really. It’s awful”. Although more qualified nurses are available during the daytime, the Ward was frequently so busy that it was sometimes difficult to respond promptly to requests for analgesics. Team leaders felt that they were particular exposed to conflicting priorities. As one stated: “It’s getting so much that if you are team leader you really are running around, see the patients, checking up on whether this x-ray has been taken and this blood’s been taken. You say, ‘I'll be there in a minute’. You really mean to be there, and between picking up the card and getting to the drug trolley, you are stopped three times. It’s when people collar you ‘just do this, just tell me this’ and it can take you 15, 20 minutes to get back to them and that's not good”. Combined with their pharmacological knowledge about the advantages of taking regular analgesics, the nurses awareness of likely delays led them to actively encourage patients to take analgesics during the drug round. Some nurses mentioned an additional problem in relation to the team nursing system, as follows: “In the mornings it's very independent and split. And then if nurses from one team go off to coffee and you're sort of holding the fort for half an hour or whatever, you don't know that patient well enough, or their pain patterns, or their needs to really make good decisions on their care”. “And it's like in the afternoons… you might swap sides for a couple of hours, but you are not really on the ball with the patients, even though you get a quick verbal you don't know the patients well enough”. Other nurses disagreed that this was a problem. However, all agreed that the situation improved when the Ward Sister was on duty as she clearly adopted the role of co-ordinator allowing team leaders to plan, organise and deliver care. Quite why a co-ordinator was not identified in Sister’s absence was not ascertained as responses to this question resulted in comments such as: “I really don’t know, we should go through the off duty and nominate who is the co-ordinator”. Although a number of contradictions, imbalances and frustrations existed within the system, there was a general feeling that the management of pain was increasingly effective. The final section of data addresses nurses’ beliefs as to why this was the case. 6.2.10 Recent Improvements A variety of reasons was offered for the recognised improvement in pain management during the past few years. These included issues related to technology, education and resources. As previously mentioned, the development of PCAs and epidural analgesic delivery systems were viewed as enabling more active analgesic management which avoided the peaks and troughs experienced with intramuscular (IM) analgesics. Although the time saved in preparing IM injections had been replaced by the requirement to carry out regular assessments, the latter were conducted at the bedside rather than at the drug cupboard. The staff thus felt that they were able to spend more time with patients assessing and discussing pain and its management. The increased variety and availability of pressure relieving aids was regarded as an important improvement in relieving the pain of patients with rheumatoid flares. Other resources, such as the availability of pharmacists, the advice offered by staff in the local hospice and the development of the Acute Pain Service were also viewed positively. These personal contacts helped to improve individual patient’s pain management and also enhanced nurses’ knowledge. Attendance on study days and the pain management module at the local University had increased pharmacological knowledge and general awareness of pain management issues. There was also a sense that the overall “improvement and advancement” of nurses meant that they were more interested in the literature, more able to critically appraise research and increasingly able to apply their learning to pain management on the Ward. 6.2.11 Summary In summary, the nurses exposed a number of contradictions and frustrations within their ideologies, working practices and relationships with other disciplines. Pain was regarded as an important priority that should be managed through a multidimensional approach. There was a reasonable amount of consistency in terms of this management, although some dissonance existed in relation to the goal and the degree to which care could be said to be based on a patient-centred philosophy. The following section moves on to consider the data collected about praxis as the study progressed. 6.3 Creating Change In Practice: Pain and Its Management Data about the nature of the practice problem enabled the co-action researchers and collaborative group to work with nursing colleagues from a basis that did not threaten current good practice. Thus the foci in creating change included areas that were deemed problematical and those that linked to the research aims. As would be anticipated, given the length of the study, a number of contextual changes took place during the course of the research. These included the Ward’s focus altering from orthopaedic / rheumatology to orthopaedic / trauma, consequent re-allocation of staff within the surgical directorate and on-going developments within the Acute Pain Service. Some of the latter were informed by the study data but managed independently. Praxis related to pain management was thus influenced by a number of external factors. Nevertheless, the nurses believed that many of the practice transformations on the Ward arose either directly or indirectly from the study, with the latter emerging from increases in nurses’ awareness, understanding and confidence in relation to pain management. Data about changes in practice were collected at a variety of forums as mentioned in the introduction. These included the co-action researcher discussions, collaborative group meetings and the staff workshops. Data were also available from the staff questionnaires, researcher logs and general field notes. Three main themes were identified and each is discussed in turn below. 6.3.1 Pain Assessment During the study, the field notes, workshops and collaborative group discussions indicated that nurses were increasingly likely to ask patients whether they had any pain, rather than whether they wanted “anything for pain”. Assessment thus became more focused upon patients’ pain experiences rather than upon their analgesic requirements. With the exception of patients receiving analgesics via PCAs or epidurals, assessment remained relatively informal until the RCT was underway. The 0 - 10 Pain Intensity and Distress Scales, used during the RCT, provided a visual representation of subjects’ pain experiences from one day to the next. Not only did this have a positive impact upon patients (see section 5.3.4.1), but it also provided feedback to the nurses on how the patient was progressing with their pain management. One or two nurses commented that they had used the scales in discussions with medical staff and had found that: “They’re particularly useful when we’re talking to the docs about a difficult case. You know, we can see it and so can they. I think it makes them more willing to listen to us”. The field notes also indicated that nurses were continuing to use the APS 1 - 5 scales after PCAs and epidurals were discontinued. However, because documentation was not available to record patient responses, the assessments were verbal and the benefits of the visual image were unavailable. The possibility of adopting the Burrows pain assessment chart (Bpac) had been considered early on in the project. The Bpac, which was developed during a previous action research study, enables the assessment of patients’ self-generated strategies (Burrows, 1997b). However, the tool was developed without reference to patients and it became clear during the survey analysis that it was too restricted for use in the current study. Furthermore, its adoption could have been regarded as constraining by the collaborative group and informants. Instead, the Acute Pain Sister raised the issue of formal assessment at the Pain Link Nurse meetings. The group recommended that the APS scale should be incorporated into the re-design of the hospital observation chart that was currently underway. 6.3.2 Pain Management Data highlighted that the nursing team’s ideologies became increasingly patient-centred as the study progressed. In relation to the goal of pain management, the majority of nurses adopted the perspective that patients should be involved in determining the extent of pain relief they required. However, newly qualified nurses were initially inclined to believe that all patients should be pain free. While health care professionals might aim to provide full pain relief, the wishes of the patient should be considered if patient-centred care is to be a reality (Campbell, 1998). The increasing agreement amongst nurses regarding the goal of pain management was threatened one year into the study, when a surgical directorate ward was closed and management re-allocated nurses through-out the directorate. As a consequence, several of the Ward’s nurses were moved, including one of the collaborative group members. Not only was the team discouraged at losing valued colleagues, but they were also concerned that: “Incoming staff might have different views. We have come a long way and I’d hate to see that stopped”. The Acute Pain Sister was able to offer some reassurance, as she was aware of the pain management practices and apparent ideologies of the nurses who were being transferred into the Ward. She felt that most of these nurses held similar beliefs regarding pain relief, although the Ward nurses were more familiar with epidural management and the differences between acute and chronic pain. She suggested that the incoming nurses might have some concerns about the study and the fact that they would be: “Behind the rest of you in knowing what we are doing. Maybe as you fill them in, you can discuss things like beliefs and goals, where we have got to in our thinking, as well as practice. You know the sort of things we discussed a year ago at the focus groups and what we are planning now in these workshops. Remember, they might be quite nervous about the thought of a nurses led experiment”. These sentiments were taken on board and arrangements made for the Ward Sister and collaborative group members to: “Bring the new nurses up to speed”. In practice, no difficulties arose. The new staff were quickly socialised into the Ward team and, with the exception of one nurse who left the hospital within two months, expressed keenness to participate in the study. Data collected during the workshops offered the first empirical indication that nurses believed that considerable improvements had been made in the previous year. It was evident that praxis had been prompted by the exposure of nurses’ ideologies during the focus group discussions and the consequent increase in critical reflexivity within the Ward team. A collective consciousness was apparent at the workshops and examples were given of where nurses had acted to transform some of the contradictions that had been noted. As one participant commented, the improvements had occurred: “Mainly because the study has made us think in different ways and be less reticent in saying what we think should be done”. The continuing influence of the co-action researchers and the praxis of the collaborative group were regarded as providing the necessary: “Little bit of nudging to keep things moving forward”. The following sub-themes highlight the transformations that occurred in relation to pain management and expose the difficulties that continued to exist. 6.3.2.1 Patient Generated Strategies The focus groups raised consciousness regarding the possibility of identifying patients’ self-generated strategies and incorporating them into the plan of care. Several of the nurses who had participated in the discussions later tested out the ideas, reflected on their ‘findings’ and reported back to the co-action researchers and collaborative group members. The following examples typify this praxis. The first is taken from my field notes recorded after a discussion with one of the collaborative group nurses and the second from a comment made at one of the workshops: “A really exciting meeting with X today. She came in to discuss the analysis of the survey and ended up telling me the sort of things she is writing in her reflective journal as a result of being involved in the survey. I was especially interested to hear that she has been asking patients on admission whether they have any strategies that they use at home for pain. Where possible, she has started suggesting ways in which they could carry on using these in hospital after surgery. Of course, she is stressing that they must also take their analgesics. She said that she has also talked to students and some of the other staff about her discussions with patients and found out that quite a few of the qualified nurses are also asking patients! This is really incredible - the commitment of this team just bowls me over”. “It’s really interesting this patients’ own pain techniques thing. Since the focus group I’ve asked the odd one or two if they do anything and they have all told me things. It’s quite funny what some people do. But seriously, I am amazed at how many different things they use for pain. With a lady the other day, well she said she gets up to make a cup of tea at night, it’s not really the tea, it’s just the walking helps her pain and stiffness in her back and doing something is a good distraction. So when I was on nights the other night, I asked her if she wanted to do this and she did and she was so grateful. She just said it made her pain feel so much better and helped the tablets work better”. A summary of the survey results was presented at the workshops. The participants were particularly interested to hear about the range of strategies used by the sample as well as their colleagues’ experiences in assessing strategies informally. There was a feeling that self-generated strategies should be formally assessed upon admission. As one nurse stated: “We could ask them when they come in what they do at home and suggest that they do these things in hospital”. However, the nurses remained sceptical as to whether the employment of self-generated strategies could make any difference to postoperative outcomes and, as a consequence, were concerned about the time commitment in assessing and supporting patients’ strategies. As section 4.4.7.1 highlighted, the discussion concluded in the nurses deciding that an experiment was needed to ‘prove’ whether or not it was worthwhile identifying and supporting self-generated strategies. Several factors influenced this decision. Firstly, nurses wanted ‘hard’ evidence, not just for them, but also so they could be confident in discussing the potential of self-generated strategies with patients and medical staff. Secondly, they were becoming increasingly exposed to the concepts of evidence-based practice and the values of RCTs as the gold standard. Thirdly, it was important for the nurses continuing empowerment that they adopted an approach that their medical colleagues would understand and value. Finally, they needed to be sure of the impact upon outcomes if they were to expend time and energy on future praxis. The process followed in developing the RCT is further outlined in section 6.4.3 The collaborative group members and Acute Pain Sister were in agreement with the collaborative informants’ perspective. The situation for myself was more difficult. On the one hand, with a background in surgical nursing and a commitment to evidence based practice, the potential of an RCT to consider measurable outcomes was attractive. The possibility of producing generalisable results, whether the findings were positive or negative, was also appealing. However, the adoption of critical reflexivity and a collaborative action research approach appeared contradictory. Nevertheless, the arguments offered by the nurses were persuasive and although they agreed that an RCT would prevent patient-centred care being offered to all in the short-term, I agreed that it could enhance the implementation of evidence based patient-centred care in the long term. Unfortunately, but not unsurprisingly, some of the nurses were unable to attend the workshops. Following discussion within the collaborative group and liaison with the Ward Sister, letters and handouts were sent summarising the survey results, focus group findings and the workshop discussions (see appendix 17). Each nurse was also given the name of a collaborative group member with whom they could discuss queries and issues. Combining analgesic management with self-generated non-pharmacological strategies led the nurses to feel that pain management was increasingly multi-dimensional. During the RCT, nurses learnt about a variety of strategies, both through reading and working with the staff booklet and also through discussions with patients in the experimental group. It was clear that these patients often took the lead in the discussions and, although the nature of the RCT prevented them acting as pre-eminent team members (Elliott and Turrell, 1996), power balances were altered. As one nurse and an HCA stated: “I know this is a study at the moment, but you know if it works it will be great. I just feel there is so much more to pain management now and the patients really take the lead and teach us about what they do. It actually helps with other patients”. “It’s really nice when the strategy ones ask me to help them get something or help them with their pain because I can become involved. I don’t have to wait for the staff nurse to get the drug. If it’s one of the red ones (experimental patient) I can see what they put on their sheet and suggest they do what ever it is while they’re waiting for the painkiller”. This later statement was made by an HCA and typifies how this group began to become proactively involved in patients’ non-pharmacological pain management. Overall, the nurses felt that the shift in emphasis to a multidimensional approach had broadened their repertoire, enabling them to collaborate with and engage patients in their own pain management. As a consequence, the nurses appeared less frustrated than they had done during the focus groups. Nevertheless, problems were encountered during the RCT. The staff questionnaire and discussions, referred to in section 4.5.3.1, highlighted that while few difficulties were encountered in identifying experimental patients’ strategies and engaging these patients in preoperative discussions, time was not always available to proactively support patients postoperatively. A further, and to some extent linked, difficulty was that it was easy to become confused as to whether consented patients were in the experimental or control group. At the suggestion of one of the nurses squares of different coloured paper were subsequently pinned above each subjects bed so that staff could quickly and easily see which patients they should be implementing the trial intervention with. As the RCT data were analysed and shared with the nurses there was considerable interest that the time and effort involved in undertaking the study was showing statistically significant results. Although disappointed that length of stay and total analgesic consumption were not reduced, the nurses were delighted with the findings related to pain intensity and distress, anxiety and types of analgesics consumed. The collaborative group had already discussed the implications of the results for praxis (see section 6.4.2). However, it was notable that the collaborative informants came to many of the same conclusions: “I think we need to implement this with all patients straight-away. We could use the Strategies Questionnaire with everyone when they are admitted, that way we can give them their drugs and work with them to support their strategies”. “I think this is great, we need to get everyone doing that questionnaire. It means we can really get patients involved, you know, value what they do, empower them to be involved”. “If we talk to them about this as part of the pre-op info, we can really discuss a much more whole approach - drugs and the patients own techniques combined. I think that would be really helpful and it will really show patients we consider their pain a priority”. “We can be confident in telling patients it really does work for them to do their own thing, but they need to tell us so we can support it. The experiment has shown that, hasn’t it?” “Can we alter this booklet so that staff and patients have copies? What I mean is that it needs altering for patients”. One of the collaborative group members, who coincidentally shared the RCT findings with several of the most newly qualified nurses, expressed some concern that these nurses were a little confused regarding the differentiation between self-generated and taught strategies. She indicated that one or two had suggested that what they learnt from one patient they could “implement” with the next. Although the issue was discussed further and the differentiation made clear, she suggested that the Ward based collaborative group members would need to take responsibility for ensuring that new members of staff are empowered to understand the differences and act in concordance with the rest of the nursing team. The collaborative group had recognised that nurses’ perceptions of what might or might not work could influence patients’ strategy use, and had considered that this might lead to the development of a false ideology. However, confusion between self-generated and taught strategies had not been anticipated. The collaborative group member’s suggestion was deemed to be discerning under the circumstances. One of the on-going criticisms of action research has been the issue of permanency and whether practice transformations will be maintained once “the researcher” leaves the field and the team starts to change (Lewin, 1946; Holter and Schwartz-Barcott, 1993; Hart and Bond, 1995b). By working in a collaborative manner to challenge and alter ideology, praxis was invested in the nurses who were involved in the study. Meyer (1993) and Armitage et al (1991) describe the difficulties that can arise as result of staff turnover, however, the Ward team is relatively stable and my link lecturer relationship has been maintained. As a consequence, the chances of change being lasting were increased. Although new members of staff would need to be introduced to praxis on the Ward, the intention of the collaborative group to continue moving into a third cycle also suggested that the practice transformations might be maintained. Finally, the commitment of the Acute Pain Sister to implement the Pain Strategies Questionnaire throughout the hospital, should produce the required organisational transformation to support praxis. These issues are addressed further in chapter seven. 6.3.2.2 Pharmacological Management One of the interesting findings, within the RCT data, was that patients in the experimental group appeared to be enabled to continue their usual analgesic management once PCAs and epidurals had been discontinued (see section 5.3.3.3). Thus the patients in some sense, led prescribing and administration practices. Several months after the RCT was completed, one of the Pharmacists, in collaboration with the Acute Pain Service and relevant practitioners, developed and introduced an analgesic algorithm for use throughout the surgical directorate. The algorithm is underpinned by the systematic literature reviews carried out by the Oxford Pain Research team to establish the efficacy of various analgesics (McQuay and Moore, 1998). Analgesic prescribing at the study hospital has thus been rationalised and confined to named drugs which the reviews have shown to be effective for postoperative pain management. As a consequence, choice has seemingly been removed from the medical and nursing teams and also from the patients. On the surface, this loss of individuality would appear to be in conflict with patient-centred care. However, as Stower (1992) points out, health care decisions should be based upon the patient as an expert on their needs and the practitioner as an expert in health care. Practitioners therefore have a responsibility to share up-to-date knowledge with patients and negotiate care management from a sound evidence base. The introduction of the algorithm at the hospital is an example of good, evidence based practice, which has improved postoperative pain control and reduced the tensions between doctors and nurses. The impact of altering analgesics on patients self-efficacy beliefs, however, needs examining in an acknowledgement of the multidimensional factors influencing pain perception and management. 6.3.3 Multidisciplinary Working One of the main areas of transformation related to the relationship and working practices of the nurses and medical staff. The workshop data suggested that the increasing knowledge and confidence of the nurses was enabling them to lessen their power play with the doctors and adopt a more professional and empowered approach. The following statements highlight these changes: “I feel I can talk to them on a much more of a one to one now, whereas before it was like, well I suppose it was manipulation really”. “This doctor said to me the other day ‘what shall we do about Mrs so and so’s pain?’. He really wanted to know my opinion and I think it’s because we have this sort of mutual relationship with them now. I know what I think and I tell them sort of professionally. Before it was like we used different ways to get them to do what we wanted and that didn’t always work of course. Now it’s much better, much more of a two-way thing”. A number of on-going difficulties remained in the doctor / nurse relationship. However, these were generally related to specific individuals rather than overall working practices. For instance, one of the Senior Registrars, who was viewed as possessing an autocratic manner, was described as ‘ordering’ nursing staff to comply with his wishes such as discontinuing PCAs and epidurals when patients were still in pain. The nurses were concerned that this individual might prove obstructive during the RCT. However, all agreed that the consultants were continuing to demonstrate interest in the study and were therefore likely to maintain their active support. The nurses indicated that if they could not persuade the Senior Registrar to work with them, they would discuss the issues with the relevant consultants. It was notable that they did not suggest that the Ward Sister, senior members of the collaborative group or co-action researchers should act as advocates in this situation. On the whole, the medical staff and physiotherapists remained interested and supportive of the study. I was frequently asked how the study was progressing and one doctor offered assistance with data collection during the RCT. Once the RCT results had been analysed, the co-action researchers received invitations from the orthopaedic consultants, the rheumatologists, anaesthetists and surgeons to present the findings at academic sessions. The presentations were well received and discussion prompted further critical reflection within the collaborative group. Multi-disciplinary support for the hospital wide introduction of the Pain Strategies Questionnaire and a Patient Information Booklet were evident. Finally, it should be noted that the introduction of the algorithm did much to remove any remaining tensions related to prescribing and administration. 6.3.4 Summary The contradictions that existed in relation to pain management, at the beginning of the study, were linked to a variety of false ideologies in the generic, epistemic and functional senses (see sections 3.4.4 and 7.7). As the study progressed, the nurses were able to reflect upon and raise collective consciousness regarding their ideologies, test out new ideas and develop praxis to improve postoperative pain management. Most of the areas identified as practice problems were addressed by the study, Ward team or Acute Pain Service during the course of the project. The next section discusses the staff data from the perspectives of action research and critical theory, explaining and expanding these issues further. 6.4 Collaboration: The Action Research Process The theme ‘collaboration’ considers the experiences of the co-action researchers, collaborative group and collaborative informants, together with the influence of action research and critical theory on these experiences. As mentioned in section 4.4.3.1, the Acute Pain Sister and I worked as co-action researchers throughout the study. During the survey and RCT certain members of the collaborative group also operated at this level. For clarity, the two groups are treated separately in the discussion below. A separate section is also offered on the collaborative informants’ experiences. Finally, a theme, entitled ‘becoming empowered’ amalgamates specific experiences of those involved in the study. 6.4.1. The Co-action Researchers’ Experiences Data were recorded at, or immediately following, each co-action researcher meeting. The meetings were held to discuss practical issues related to the conduct of the study and philosophical issues regarding our ideologies, reflections and interpretations. Early on, we shared our beliefs and values on the meaning of patient-centred care and its implications for pain management. We agreed that we were interested in improving pain relief and enhancing patient control. We had both observed that: “People use a variety of strategies to cope with acute pain and believe that this should be investigated. As nurses we have a responsibility to promote individualism both for its own sake and to enhance acute pain management”. Interestingly, we also felt that: “We could become too focused on patient’s own strategies. We believe that professionals should act as professionals, using their knowledge & skills to identify what the patient wants and then, where possible, fulfilling these requests. If this means staff, rather than patient initiated strategies, are implemented so be it. i.e. if patients do not want control we should facilitate this also. This is particularly important given that although we believe that increasing perceptions of control will reduce pain perception, we do not know this for sure”. Nevertheless, we believed that we should work towards identifying: “The strategies that people use in their day-to-day lives and make use of these in the hospital setting”. At later meetings the data are far more patient orientated and the terminology more reflective of critical discourse, emancipatory action and theory generation. However, before this occurred we noted that our thinking was often chaotic, lacking sense and cohesion and yet flowing and creative. We would grasp a seemingly essential idea and then lose it again somewhere in the “swampy lowlands” between practice and theory. Clarity only developed as the study progressed and our critical reflections were supported or challenged by the research and action. For example, early on we discussed the tension between science and individuality: “Can it be scientific if individuality is allowed & encouraged rather than identifying ‘the best’ strategy to implement?” At a later stage we came to regard the identification and support of self-generated strategies as a ‘best strategy’. Thus it was not necessarily the specific strategy used by a patient that was important, but rather the engagement of patients in a patient-centred approach which valued the individual’s ideologies and pain relieving actions. A further example stems from our discussions regarding how much I needed to read, summarise and share the literature on coping. Issues about coping are addressed in chapter seven, the purpose of acknowledging the debate at this stage is to highlight the breadth and depth of our discourse. As advocated by Titchen and Binnie (1993a) we spent some time considering whose interests were being served by the study. First and foremost our own interests were patient-centred and focused upon whether identifying and supporting self-generated strategies could: “Promote coping behaviours, increase perceptions of control and reduce pain experiences.” Linked to this was an altruistic interest in decreasing nurses’ frustrations and enhancing their control in pain management. We recognised that these feelings stemmed from our own clinical practice during which we had been exposed to many of the same frustrations. We also both wanted to learn more about research and this was reflected in the way we organised the different aspects of planning, acting, observing and reflecting. For example, while I analysed the RCT data, the Acute Pain Sister designed and analysed the staff questionnaire and facilitated the researchers’ critical discussion. Titchen and Binnie (1993a) and Titchen (1994) suggest that co-action researchers work in a way that allows each to take the lead for specific and appropriate aspects of the study. Our experiences and the study data reflect this approach. Where possible, we tried to present ourselves as united co-workers. This meant that we often agreed agendas, priorities and who would take the lead in advance. However, it did not prevent us from challenging each other in front of colleagues. Indeed we regarded this as both an important freedom for ourselves and a necessity if we were to encourage critical reflexivity amongst all collaborators. Our friendship was undeniably helpful in allowing us to be critical as we were confident in each other’s inherent worth and knew that we could work through any dissonance. Discussions surrounding our individual strengths and weaknesses and how these impacted upon our roles within the study were an important on-going theme. The field notes highlight that I brought: “Research knowledge and organisational skills. Her self-motivation encourages and motivates others. Her facilitation skills create a relaxed atmosphere in which critical reflection is encouraged in a goal directed manner. Her communication & teaching skills enable others to develop research knowledge and have a clear idea of the aims and objectives of the study”. The Acute Pain Sister was viewed as bringing: “Knowledge of pain management which is up-to-date and clinically orientated. Effective communication skills enable her to liaise with all levels of staff in a friendly and professionally appropriate manner. Her assertiveness and commitment to ‘getting the job done’ maintains the pace of the study when dealing with consultants and nursing staff”. We recognised that many of our skills and abilities were complementary and that we tried to address mistakes and weaknesses. For example: “Recognising our propensity to deviate, we have attempted to improve our time and content management”. (I) “sometimes worries and feels guilty about adding to people’s workloads and is consequently less pro-active than she could be in encouraging people to meet deadlines, attend meetings and so on. This has a knock on effect on the overall progress of the study and may be one reason for the recent loss of impetus in ward staff. (For example, not pushing for the transcribing to be completed)”. (The Acute Pain Sister) “feels that she lacks time to commit to the study as fully as she would like to”. Nevertheless, we were: “Both enthusiastic about pain management (theoretically & clinically) and committed to improving pain control. We each believe that the other has sound facilitation and communication skills. Although our styles are similar, we are always ready to listen to each other’s suggestions for improvement. We use our past experiences to develop our roles and we believe that our friendship enhances our co-action research partnership. We are constantly helping each other to problem-solve whether in relation to the action or research”. As the study progressed we believe that we helped to empower each other. Rodwell (1996) argues that no one person can empower another. However, we felt that we encouraged each other to “be brave and try new things”. For example, I had always been reticent at approaching consultants with requests for support or assistance. The Acute Pain Sister knew that my reasoning that she had a working relationship with these doctors, and I did not, was “just an excuse”. She encouraged me to be more assertive. Equally, I was aware that her comments that I should lead on presentations, as I was more experienced, were also excuses for her nervousness in this area. By the end of the study, the Sister acted as an equal co-presenter when we shared the findings with a variety of internal audiences. When working with the collaborative group and collaborative informants, we developed techniques which allowed us to present draft ideas and facilitate critical discourse. We believed that: “Where possible staff should be assisted to initiate self-development. Our approach should be facilitative, aimed at staff empowerment and self-direction”. We used role modelling by questioning each other, overtly encouraged and valued contributions, sought others’ opinions and found ways of raising challenges without increasing vulnerability. My previous work on facilitation was informative in developing our techniques (Burrows, 1997a). An additional, but important point, was that we also developed strategies to encourage participation. For example, nurses attending events in their own time were offered bank staff rates or given time back by the Ward Sister, certificates of attendance were produced and letters, confirming involvement in the study, were written for individual’s portfolios. These activities valued and helped to maintain the nurses motivation. With regards to data collection, we shared and learnt from each other, although the responsibility for reviewing the literature and producing material to facilitate learning, fell to me. In preparation for the focus groups, the Acute Pain Sister read my publications (Burrows and Kendall, 1997; Burrows, 1998) and subsequently became increasingly participative as a co-facilitator. For the survey, we worked, with the collaborative group member who joined us as a co-action researcher, to develop our approach and undertake inter-rater reliability checks. Again I provided a summary of the literature and a list of points for consideration. However, discussion and decision-making were three-way. A particularly notable point, which highlighted the increasing confidence and critical ability of the Acute Pain Sister, occurred when, following a discussion with the study supervisors, I suggested we might aim for a stratified sample. The field notes summarise that the Acute Pain Sister: “Raised some practical & moral problems of doing this & it was agreed that we would basically keep to a random sample”. Reflective notes were kept throughout the survey. Three different excerpts, recorded during the interviews, are offered below to highlight the type of dialogue maintained: “He was physically relaxed and mirroring my own body posture. (NB/ I did deliberately mirror his to begin with)”. “Relaxed interview but client only seemed to have one pain relieving strategy - Alcohol!!! Tried hard to elicit further information from him unsuccessfully. I got the feeling from him that he was just passing the time by partaking in the survey but didn’t view it seriously. Having read the transcript we agree. On a serious note this was an honest account which reflects X interview style”. “Have had three patients in one clinic session who were either dyslexic or unable to read or complete the questionnaires on their own. Fortunately two of them explained their problem to me and asked me to fill in questionnaires for them while they gave verbal responses. The third one I picked up when I realised he was struggling. He was reluctant to admit his problem - another dyslexia!! Is this something we should be more aware of and perhaps inform them we will be available to help them fill in the form? I wonder how many others may have refused the survey purely because of subnormal literacy skills rather than disinterest in the project”. The field notes were not only important in helping us to critically reflect upon our research practice, but they also offered a communication forum. This became increasingly important as the survey became extended due to the non-attendance of booked patients. For instance: “What a wasted afternoon in the heat - 1 out of 7!!” “Cheer up, I managed another 8 yesterday”. In preparation for the first of the staff workshops, the two co-action researchers listed some of the ideas which had been discussed at the collaborative group and worked out the advantages and limitations of each in advance. We were then able to share these during the first session to prompt critical reflexivity. We then altered the presentation for the next workshop to take account of the discussion. This format was followed for each session. Although the Acute Pain Sister was fully participative in designing and implementing the RCT, she went on maternity leave soon after the trial commenced. Excerpts from my personal notes demonstrate how I missed her support and challenging. On return, however, the relationship took up where it had left off and, in fact, the Acute Pain Sister took the lead in relation to the final staff questionnaire and the researchers’ critical discussion. The following section moves on to consider the collaborative group’s experiences. 6.4.2 The Collaborative Group’s Experiences Members of the collaborative group were involved at all stages of the research: in planning, or formulating ideas for action; acting and critically examining how the objectives might be met; and reflecting upon the results. In addition, each was given the opportunity of participating in the observing stages, implementing strategy and collecting data. This offer was taken up by one member during the survey and by the group as a whole during the RCT. In many respects, the application of Titchen’s (1994) definition became inconsequential as all members therefore became co-action researchers during the RCT. Nevertheless, they were not involved to the same extent, in analysing findings or collecting data from the collaborative informants, as the Acute Pain Sister and myself. The first collaborative group meeting set the tone for the rest of the study. The meeting was friendly and participants were interested and enthusiastic. Notably, one member had arranged childcare to attend and this pattern continued throughout the course of the study. At this first meeting, members shared their rationale for volunteering. Some had done so because of their interest in pain and its management, while others were also curious to learn more about research. My willingness to work on the Ward and find out about current practices was positively regarded. It may have helped that I attended the first meeting in uniform, having just completed a morning shift on one of my link areas. Meyer’s (1993) point that action researchers need to be clinically credible was thus recognised. Of note, although my suggestions at the first meeting were taken positively,: “...everyone seemed able to express their own opinions. I believe this is because I achieved the right balance in articulating what I am an expert in and what I am a novice at. I left the meeting feeling excited and full of anticipation. The motivation is overwhelming, as is the confidence with which (members) make critical suggestions”. Following the meeting, the Acute Pain Sister and I reflected upon whether this was due to my facilitation skills or the characteristics of the collaborative group members. At this early stage it was impossible to determine the source of reflexivity, however, the atmosphere set the scene for the remainder of the meetings. Ultimately, we concluded that the ability of members to challenge, and be challenged, was partly linked to their own maturity, but also arose from the facilitation style to which they were exposed. This was confirmed by group members during the researchers’ critical discussion held towards the end of cycle two (see section 4.5.3.3). Participants stated that involvement in the group had led them to be more objective and critical. Confidence was enhanced so that they felt able to challenge ideas both within the group and also in relation to practice issues. The safety of the group and the valuing of ideas within the group had much to do with this. Early on in the study, it was acknowledged that the membership of the group was likely to alter over time. Interestingly, this did not cause a problem within the group, as it was perceived that there were sufficient senior members of staff who were unlikely to change jobs during the course of the study. New members commented upon how welcome they were made to feel and, as the study progressed, further comments were elicited which highlighted the extent to which working as a team was valued. For instance: “You knew you were not on your own”. “It stopped us just giving up”. The support obtained from colleagues was thus important in sustaining momentum. Members also welcomed the idea that the study would last several years as they felt that this would: “...enable us to set deadlines that are comfortable for us and realistic for the Ward”. The group added, however, that I would need to ensure that opportunities were maintained for participants to feel involved and increase their knowledge and skills regarding pain management and research. To facilitate this process, I made a considerable effort to communicate research theory to the group. Thus meetings were often spent discussing design, conduct and analysis. As the study progressed, attendance at the meetings waxed and waned as a consequence of individual’s personal commitments, the stage of the study and how busy the Ward was. At one point, there was a period of some five months where meetings were constantly cancelled. The Acute Pain Sister suggested that she contacted members one week in advance of each meeting to check their attendance and rang them again on the morning of the meeting to remind them to attend. This idea was subsequently accepted and generally worked well. The collaborative group was encouraged to maintain reflective journals throughout their involvement in the study. Although reflections were shared and discussed at meetings, the written records were never made available to the co-action researchers. There was some reticence in identifying why this occurred, however, it appeared to be related more to participant’s ‘newness’ at completing reflective journals than a problem with sharing the contents. Indeed, on a number of occasions members brought their journals to meetings or appointments with me. As the journals were not made available for analysis, a considerable amount of written data were ‘lost’ to the study. However, no pressure was applied as it was deemed more important that individuals should have the freedom to keep journals and be assured of privacy and voluntary consent. The authority to implement change on the Ward lay firmly with members of the collaborative group. Although I was the Ward’s link lecturer, and the Acute Pain Sister held responsibility for developments within the Trust related to the nursing management of pain, the collaborative group retained power for initiating change upon the study Ward. As a consequence, the co-action researchers were reliant upon the motivation and commitment of the group. Fortunately, despite staff changes, the group was always committed to advancing the study and instituting practice transformation and research activity. Although not all members were involved in every aspect of the study, they requested copies of any information that I produced. Thus the information sheet about inter-rater reliability, for example, was distributed to them all. A second example related to the group’s interest in data analysis. For instance, following the focus group discussions, members requested information regarding qualitative analysis. Some also attended additional meetings with the co-action researchers to participate in the analysis strategy. This process was followed through with both the survey and RCT data. As a result, although the group was not directly involved in basic data analysis, their understanding allowed them to challenge the process as well as critically reflect upon the results. Nevertheless, participants’ lack of involvement meant that most retained their wariness of undertaking statistical analysis, although aspects such as the meaning of statistical significance became well understood. The group was particularly active in keeping their collaborative informant colleagues informed of progress. Throughout the study, the nurse who had initially been on night duty offered to take the lead in keeping night staff informed and involved. Equally, those on day duty shared responsibility for ensuring that all colleagues had opportunities to give their opinions. Sadly, and despite attempts to the contrary, none of the HCAs could be persuaded to join the group. Nevertheless, certain HCAs, who attended the focus groups and staff workshops, ensured that other HCAs were fully informed. The collaborative group was proactive in designing information packs for the collaborative informants. For instance, at the request of their colleagues, members worked with the co-action researchers to determine the content of the staff strategy booklet used during the RCT. They also suggested that they took responsibility for sharing and discussing the results with the collaborative informants. Thus each group member had named individuals with whom they discussed the findings over a three-week period. They were provided with a results pack and posters were placed on the Ward to prompt discussion. Group members fed back the comments to the co-action researchers for consideration regarding future praxis. (These comments were summarised in section 6.3.2.1). One of the major concerns expressed within the collaborative group was the dual role that members would adopt during the RCT. As nurses on the Ward and researchers within the RCT, there was a danger that rigour would be compromised. These issues were discussed at group meetings and strategies identified to reduce bias and ensure the maintenance of critical reflective notations. For example the field notes stated that: “When we are acting as researchers we need to minimise all inter-actions e.g. do not ask “how are you feeling”. If interactions do occur we need to be as natural as possible and record them. There are real tensions here for the collaborative group between RCT rigour and AR collaboration”. Communication between the researchers involved in the RCT was excellent. Rarely a day went by without someone recording in the log. Although many of the entries summarised details regarding the number of booked admissions, versus the number admitted and consented, other notes reflect the constant communication between the researchers and reflexive notes regarding experiences and conversations with subjects. The agreed strategy appeared to work well and there was little evidence of any tension or compromise within the roles. Several specific comments were made regarding the level of stress experienced as a result of the RCT: “...going on and on”. Originally, given the pattern of admission to the Ward, it had been anticipated that data collection would be completed within four months. In fact, it took some seven months due to a fall in booked admissions as a consequence of winter pressures. Maintaining motivation became increasingly difficult and, as time went on, members frequently felt demoralised, particularly if a consented patient did not complete. There was also a sense of responsibility to maintain the motivation of ward colleagues. This added further to the stress of the group members. It was a reflection on the cohesion and commitment of the collaborative group that they supported each other and kept the study going. During the study, participants became increasingly reflexive. Although they had demonstrated an ability and confidence in challenging the co-action researchers from the beginning, their was evidence that the level of thinking altered from intuitive disagreement to one in which participants were able to expose not only the rationale for their dissension, but frequently the contexts behind this rationale. For example, at the end of the second cycle I had suggested facilitating a critical reflexive discussion on our experiences in the RCT. The idea was accepted with the caveat that the Acute Pain Sister should act as facilitator. The rationale included that as the Sister had been on maternity leave for most of the data collection period she could feel excluded from the discussion, asking her to adopt the facilitators role would thus ensure her active involvement. Furthermore, this would mean that I could be a full participant in sharing my experiences rather than trying to undertake a dual role as facilitator and ‘reflector’. Finally, there was a perception that the Acute Pain Sister would be more probing and challenging and less biased precisely because she had not been an ‘insider’ to the experience. The critical discussion revealed that on the whole, collaborative group members enjoyed the experience of being involved in the study and found that the exposure removed many of their fears regarding research. As one stated: “It’s so nice to experience it. It’s not as frightening as you think it’s going to be”. Another suggested that although they needed “some time out”, they would “certainly be prepared” to participate in research again. Finally, a third indicated that she felt “empowered to tackle” a small research project on her own. As a central feature of a critical theory approach to action research, it is interesting that this member used the term empowerment. An interesting, and unanticipated consequence of members’ involvement in data collection, was the perceived improvements in individual’s communication skills. The self-consciousness experienced at the beginning of each new data collection method, resulted in increased self-awareness in communication. For example, during the survey, one of the interviewers commented that she had to actively listen to subjects in order to ensure that their responses related to what was required and to enable appropriate supplementary questions to be asked. This member commented that she had subsequently paid much more attention to her communication skills when talking to patients on the Ward. The critical discussion was followed, at a later date, by a collaborative group meeting to consider praxis and ideas for the third cycle. These were outlined in section 4.6, however, some additional detail, highlighting the level at which the group were operating by this stage, is provided in the following excerpt which was taken from the meeting notes: “We discussed whether the revised Patients Strategy Booklet should just be introduced, or should it be trialed? If the latter, could we conduct very brief interviews with patients regarding their views on the booklet (or do we want to do a randomised controlled trial!)? Either approach would need to be discussed with ward staff (X and Y said they would do this). If there is to be any data collection (the author) would like to see staff taking the lead on this with support and assistance from herself and (the Acute Pain Sister), rather than the other way around. This was agreed by those present as it would help to consolidate the skills that have been developing in relation to both pain management and research”. Section 4.6 highlights that, with the agreement of the collaborative group, the content of the booklet was incorporated into a new Pain Management Booklet that was about to be piloted by the Acute Pain Service. The group continued other activities during this pilot. Additional comments were made regarding the need to audit practice changes on an on-going basis as a: “… continuing ‘test’ of the RCT results”. The point was a valid one as although we had tried to be as rigorous as possible, practice was being transformed on the basis of a single RCT. Carr and Kemmis (1986) state that all practitioners are capable of becoming action researchers. However, the practitioners of whom they write are teachers and thus tend to work more independently than nurses. Ward nurses not only work in teams, but are also exposed to patients through-out their working hours. For the purpose of this study, it was impossible to involve all of the collaborators as researchers, however, the collaborative group members could be regarded as fully participative action researchers. The following section considers the experiences of the collaborative informants. 6.4.3 The Collaborative Informants’ Experiences Titchen (1994) regards collaborative informants as being subjects of the research and participants in the change process. In considering the study design, it was recognised that this approach could undermine the critical theory stance by excluding the majority of staff from reflecting upon their ideologies, challenging false beliefs and transforming practice. Thus attempts had to be made to maximise collaboration. As previously highlighted, most of the Ward nurses were involved in the focus group discussions which provided opportunities to expose and challenge ideologies. Critical reflexivity originated at this stage through the raising of collective consciousness regarding pain and its management. Evidence also existed that discussions continued on the Ward. This process was enhanced, at the suggestion of the Acute Pain Sister, by the provision of a Ward file containing the notes of the collaborative group meetings and pertinent literature which the group: “... would like staff to be aware of / comment upon”. One of the group members also proposed that summary posters of the research activity and findings could be displayed on the office notice-board, while another suggested that verbal updates could be given at team and Ward meetings by the collaborative group. I communicated with the informants both on visits to the Ward and through the distribution of letters which summarised activity on a regular basis and sought ideas and opinions from the nurses regarding both the research design and praxis. It was constantly stressed that: “As you are the ones involved in nursing patients in acute pain, your insights and opinions would be helpful”. A further strategy to enhance collaboration, critical reflexivity and participation in the research design and praxis transformation, was the involvement of collaborative informants in the staff workshops. Informants received letters of invitation, in advance, which stated that the workshops would provide an: “Opportunity to hear the results of both the survey and last year’s focus groups, and to offer your opinions on how we can best develop practice and move the research forward. (We) will present some of the ideas we have developed during the (collaborative) group meetings and ask you for your opinions and own ideas. We will also need to consider the practical advantages and disadvantages of each suggestion. Staff involvement in any data collection will be addressed and the implications discussed”. The ideas for research ranged from a qualitative, interview based study (see chapter seven), to a change experiment. The collaborative group had favoured considering whether it might be feasible to introduce change in one team, using the other team as a control. Their rationale was that: “It would enable us to evaluate how the changes affect patients’ pain management and any implications for staff”. Although qualitative approaches were mentioned at the staff workshops there was a clear preference for an experiment. As previously mentioned, collaborators wanted to test the effect of incorporating patients own pain relieving strategies into the plan of care against identified, measurable outcomes, such as pain intensity, pain distress and analgesic consumption. If there were no effects upon outcomes, the cost effectiveness of the additional assessment time required to identify individuals’ strategies would be regarded as questionable. Several different experimental approaches were considered at the first workshop. These included the collaborative group’s suggestion of splitting the teams, with the caveat that once the data had been analysed a decision could be made on the basis of the findings regarding whether or not the changes should be discontinued or introduced across the ward. This independent group design was initially the most popular suggestion as the data collection period would be shorter than for other designs and the approach could encourage night staff to work in teams. The issue of keeping the groups separate was thought to be possible. However, upon further reflection concern was expressed that although all consented patients would have data collected from them, the informants working in the ‘control’ team would not be able to experience the intervention. As a consequence, they would be unable to gain insight into the impact of adopting new working practices. There were also issues regarding rigour, particularly in relation to randomisation and the extraneous variables related to the different teams. Consideration was then given to a pre-test, post-test design, in which baseline data would be collected through-out the Ward from consenting patients. Following this, the changes would be implemented across the Ward. After a few weeks, in which nurses could become familiar with the changes, post-test data would be collected from a second group of consenting patients. This design was immediately favoured by one of the informants as she felt that it would enhance collaboration and participation. Following further discussion, the other participants at this workshop also came to favour this design, as it would mean that subjects from the control and experimental groups would not be able to mix. The disadvantages of the design were considered, together with a brief discussion of RCTs. At the second workshop, a summary of the three designs was given, together with an outline of the advantages and limitations of each. Participants were told that informants in the first workshop preferred the pre-test/post-test design. However, those in the second workshop were concerned that this might not be rigorous enough and a more traditional RCT approach was chosen. Discussion then focused upon possible outcome measures and the way in which both the collaborative informants and the collaborative group might be involved in data collection. Those attending the third and final workshop arrived with the assumption that the time would be spent teasing out the details of the proposed RCT. Clearly, discussion had continued on the Ward and agreement been reached that this was the most appropriate design. Fortuitously, the co-action researchers had come to much the same conclusion. The nurses’ involvement in data collection and their responsibilities in supporting the intervention were discussed and agreed. Thus, not only did the informants collaborate in designing the RCT, but they were able to do so from an informed perspective regarding the implications for their working practices. One of the issues identified during these discussions is summarised in the following excerpt from the field notes: “The need to teach staff about different strategies and ways to support patients using their own strategies was briefly addressed. It was decided that the (collaborative) group should develop a booklet of strategies that could be actively supported on the ward. Commentary should include suggestions such as: ‘if the patient wanted peace & quiet one possible way of at least simulating this would be to draw the bed curtains’. The remainder of the meeting focused upon other ideas of how strategies could be supported. It will be important to emphasise that staff will be expected to support existing strategies, rather than teach new ones”. An unanticipated occurrence at the first workshop was the suggestion by a participating HCA that qualified nurses should be given priority to attend the two remaining workshops. This view was supported by a second HCA who was also present at the time. The following excerpt summarises the discussion and decisions made: “X offered to take on the role of filling in the other HCAs on the workshop content and inform (the co-action researchers) of any questions or comments. It is unlikely that the other HCAs will attend future workshops as X and the others felt it was more important that the trained staff attended. (The author) disagreed as she felt everyone should feel able to be involved and HCAs are likely to have a major role in supporting patients’ strategies. However, X made a good case for the trained staff needing to comment on the design of the trial, while the HCAs would be more involved with the intervention”. Although uncomfortable with the apparent withdrawal of the HCAs from collaboration in the workshops, I had to concede that this suggestion was reflexive and relevant. The Ward needed to maintain nursing cover and yet maximise the involvement of informants in the design. The HCA’s suggestion was a sensible one and her offer to liaise with colleagues demonstrated her commitment to the study. By the conclusion of the workshops, all G and F grades had attended, together with 66% of E grades, 75% of D grades and 17% of HCAs. The remaining nurses were written to and, as previously mentioned, collaborative group members took responsibility to ensure all colleagues had the opportunity to ask questions and share ideas and opinions. Thus by the time the RCT commenced, everyone was confident that they understood their role and had given informed consent to being involved. As the trial progressed, collaborative informants were involved in critical reflections regarding any problems they were experiencing, both with the overall design and their ability to identify and support experimental subjects interventions. At one stage, extra copies of the Patient’s Strategy Questionnaire were made available for the nurses to: “Have some fun (?!) filling it in for yourself”. They were kept informed of progress and, on completion of data collection, were thanked: “… for all your support, activity and interest during the experiment. Given how busy the ward has been, I think we all did remarkably well to cope with the trial as successfully as we did”. Once the RCT had concluded, the nurses completed the Acute Pain Sister’s questionnaire and participated in the collaborative group’s reflective sessions regarding implications of the results for praxis. Those who asked were given individualised letters confirming their exact involvement in the study since its commencement. A summary handout of the RCT results was also made available and a full report was given to the Ward, collaborative group members and one or two of the informants who requested one. Carr and Kemmis (1986) believe that involvement in action research should result in enlightenment and transformation of ideologies and praxis. By adopting Titchen’s (1994) collaborative model, there was a danger that many of the nurses could have been excluded from this activity. However, the strategies adopted to enhance collaborative informants participation promoted their involvement in both the research and action. 6.4.4 Becoming Empowered Through-out the study there was evidence that the nurses became increasingly reflexive and freed from constraining ideas and approaches to pain management, both as individuals and as a team. Analysis of the second staff questionnaire indicated that the nurses’ involvement in the study had considerably improved their awareness of pain and knowledge of pain management. As they became more confident in their knowledge base, their need to retain power over patients lessened. Rather, they were able to adopt a more patient-centred approach by inviting experimental subjects to participate in their pain management. This was achieved by identifying patients’ pain management goals and their values, beliefs and actions surrounding the use of self-generated strategies. Choices were offered according to patients’ preferences and the resources available to support strategy use on the Ward. Postoperative pain management was negotiated and involved combining evidence based pharmacological and non-pharmacological approaches in which patients and nurses worked in partnership. These interventions formed the basis for emancipatory actions following completion of the RCT. Rodwell (1996) suggests that empowerment relies upon self-determination. As nurses’ knowledge and confidence increased, they were able to see that they had the authority to act. Knowledge confers power and this power can either be used or abused. The ability of the nurses to use their power to both improve relationships with the medical staff and to engage patients in their own pain management arose from an increasing willingness and ability to critically reflect on praxis. Knowledge was also gained through involvement in the research process. The outcome was increased confidence and critical ability. 6. 5 Conclusion Meyer (1995) highlights that difficulties can arise during action research projects as a consequence of participants’ reluctance to change, lack of time, motivation and leadership and the routine nature of nursing practice. Furthermore praxis may threaten participants’ ideologies and current working styles, leaving them feeling criticised and vulnerable (Meyer, 1995). Careful maintenance of field notes highlighted that, although the Ward nurses were exposed to their fair share of stressors related to organisational change, poor staffing and lack of time and resources, they rarely lost their motivation for the study. They became increasingly participative, and were able to challenge and reflect upon their ideologies. During the course of the study ideas for action were identified and acted upon. As a consequence, approaches to pain management altered from a unidimensional focus upon analgesic management to one that combined pharmacological strategies with management based upon identifying and supporting patients’ self-generated, non-pharmacological, pain relieving strategies. The next chapter moves on to consider the fourth key characteristic of action research, that of developing theory. CHAPTER SEVEN: A CRITICAL THEORY FOR ACUTE PAIN MANAGEMENT: DISCUSSION AND CONCLUSIONS 7.1 Introduction Chapters five and six presented the data collected from the patients and nurses involved in the study. Some discussion was offered in these chapters to enhance clarity in the process of theory building. This chapter now moves on to consider the fourth key characteristic of action research, that of developing theory. Specifically, the discussion explores the contribution of the study to a critical theory of acute pain management and considers the impact upon both patients and nurses of adopting an emancipatory approach to praxis. Limitations of the study are recognised within the discussion and recommendations are made for further research and practice. The starting point for theory development was a consideration of the impact of acknowledging patients’ strategy preferences. Related to this, were the ideas of freeing patients and nurses from contradictory ideologies and enabling them to work in partnership in managing postoperative pain. Concepts such as pain coping, strategy preference, self-efficacy and empowerment are relevant to this debate as discussed in the following three sections. 7.2 Self-Generated Strategies: An Approach to Coping with Pain The concepts and issues surrounding coping were considered as a possible theoretical framework early on in the study. However, a decision was subsequently taken to adopt the Gate Control Theory (Melzack and Wall, 1965, 1996) and McGuire’s (1992) multidimensional framework. Hence, the review of coping offered in chapter one was brief. The basic reasons for not using a coping framework were threefold. First and foremost, in a useful review of the literature on coping in health care, Endler et al (1993) point out that health psychology research has tended to be underpinned by a systems approach, which requires complex multivariate analysis and interpretation. At this early stage in examining self-generated coping strategies for acute pain relief, such complexity was felt to be inappropriate. Rather, there was a need to explore the range and types of strategies employed and whether their use could, at a fairly concrete level, influence patients’ postoperative experiences. Endler et al (1993) themselves, caution against adopting complex approaches in new areas of research until conceptual and measurement questions have been clarified. Secondly, the action research and critical theory stance aimed to elucidate a social action theory for praxis, rather than one that was located within a more abstract, theoretical understanding of coping. This was an important issue if the emphasis was to remain upon patient-centred care and freeing both patients and nurses from constraining ideologies with the intention of adopting new praxis. Finally, the collaborative nature of the action research approach was deemed pertinent, in so far as participants needed to be able to understand not only the practice and basic research elements, but also be involved during the reflection stages in on-going attempts at theory development. Using a theoretical framework located in the pain management rather than health psychology literature was more accessible to the nurses. Despite these issues it is important to consider the relationship of the study findings to coping theories. The survey findings confirmed that individuals develop pain coping strategies as a result of exposure to everyday pain. Significant others and health care practitioners may influence the type of strategies an individual attempts to use. Where the attempts are evaluated as being successful, they are incorporated into an evolving repertoire of techniques which individuals believe work for them. It appears that the development of a repertoire is needed to support two commonly held ideologies: firstly, that the more intense the pain the greater the number of coping strategies needed and, secondly, that certain strategies are, over-time, hierarchically ordered as being best suited to different pain intensities. The ability of individuals to differentiate between strategies used for everyday pain and those employed for more intense pain, may be explained by state coping theories. In other words, it is the stressor within the situation that determines the type of coping strategy employed (Folkman and Lazarus, 1985). For instance, strategies from the cognitive control and mobilising themes were particularly popular amongst survey respondents for everyday pain, while positioning and resting alone were commonly used for more serious pain. Several other studies have also found a relationship between pain intensity and the type of strategies used, although this is not always a consistent finding. For example, McCaul and Malott (1984) demonstrated that distraction was perceived to be more effective for low intensity stimuli, while focusing appeared to work better for higher intensities. Conversely, Kashikar-Zuck et al (1997) found no correlation between type of strategy and intensity of pain. The evidence to support the premise that coping with pain is, in part, state dependent is thus currently somewhat controversial. Nevertheless, it is postulated that situation dependency may have some relevance for acute pain coping. Weisenberg (1999) argues that situation dependence, or state coping, is complicated by personality, or trait coping styles which may influence individuals’ strategy preferences. In the survey, respondents were inclined to express their strategy preferences in connection with a characteristic repertoire, which remained relatively stable across time and circumstance. For example, those who commonly engaged in strategies from the cognitive and expressive reassurance themes indicated that they would prefer health care practitioners to adopt nurturing roles in the postoperative period. In contrast, individuals who sought control tended to use a range of behavioural and resting alone strategies. These were person focused and generally required no active support from others. Given that these data came predominantly from the small interview sub-sample, care must be taken with the interpretation. However, the RCT also demonstrated that experimental patients’ strategy use was consistent across time from admission to discharge. The suggestion is that some pain relieving strategies developed for use at home may be successfully adapted for postoperative pain management. This finding offers some support for the idea that acute pain coping styles may be linked to trait as well as state elements. The literature highlights that psychologists have been arguing for some time about the relative importance of state versus trait coping and that a number of studies have been undertaken to explore this relationship. However, the results are inconclusive with contrary findings and weak associations being suggested at best (Endler et al, 1993; Shaw, 1999). In this study, there was evidence to indicate that both state and trait styles may have relevance for acute pain coping. Coping style is not only related to issues about trait and state coping, but may also be distinguished into emotion and problem-focused coping as outlined in section 1.8 (Folkman and Lazarus, 1985). The type and range of strategies used by a UK out-patient based population were, following the survey, classified into ten themes which appeared repeatedly across the three main groups of self, family and practitioner generated strategies. As individuals appraised pain, they dealt with the stimulus through cognitive and behavioural efforts designed to alter its impact (problem-focused), or attempted to reinterpret the pain episode so that it was not perceived as threatening (emotion-focused). Strategies from the seeking treatment and positioning themes were typical of problem-focused coping, while those from the cognitive control and cognitive and expressive reassurance themes were emotion-focused. Cohen and Lazarus (1994) have considered the relationship between problem and emotion-focused coping on the one hand, and trait coping styles, such as avoidance and vigilance, on the other. There is some indication that, as well as commonly using distraction techniques, avoiders may also employ more emotion-focused coping behaviours. If this is the case, it may offer some explanation for the link, made in chapter five, between cognitive and expressive reassurance strategies and patients adopting a passive role in response to nurturing practitioners. Interestingly, Endler et al (1993) suggest that in situations where personal control is not possible, avoidance styles may be more effective in terms of psychological well-being, whereas vigilance may be efficacious where personal control is feasible. Consequently, empowering patients to use their own strategies postoperatively may alter individuals’ pain coping behaviours. This point is returned to in section 7.4. Further research could be undertaken to investigate whether patients who seek empowerment in postoperative pain management are vigilant by style and more inclined to use problem-focused, self-generated pain coping strategies, compared to those who seek nurturing, are avoidant in style and inclined to employ emotion-focused pain relieving strategies. One area that was not examined in detail in the study, and yet is recognised within both the coping and chronic pain management literature, as well as in McGuire’s (1992) multidimensional framework, is the potential for significant others to effect the individual’s ability to cope with pain. Work carried out by, for example, Affleck et al (1988), Revenson and Majerovitz (1990) and Keefe et al (1996) has demonstrated that spousal behaviour can impact, either positively or negatively, on the chronic pain experience. The outcome depends upon the type of support offered and the resulting influence on the patient’s pain coping behaviours. This is particularly the case in terms of the type of role the chronic pain sufferer seeks from others. The role and remit of significant others was considered within the survey in so far as subjects were asked what strategies families and friends employed to help relieve the individual’s pain. All strategy themes were mentioned with the exception of mobilising and cognitive control. The most frequently mentioned theme was cognitive and expressive reassurance. Unfortunately, the influence of significant others was not specifically examined during the RCT and this was noted by the researchers as a limitation of the study. Subsequent discussion has led to the development of a pilot project by a health psychologist, in collaboration with the Consultant Rheumatologist, a Clinical Psychologist, CNS Rhematology and the author. The pilot, is an RCT with a control group receiving what is now “normal” pain management on the Ward, that is identifying and supporting patients’ self-generated strategies, and an experimental group in which the influence of significant others is to be further examined. A number of tools are being used in the study to gain further insight into pain coping. These include the Social Support Questionnaire (Sarason et al, 1983), the Pain Management Inventory (Davis and Atwood, 1996) and the Pain Strategies Questionnaire (see section 7.5). The current study has demonstrated that it is possible to identify patients’ acute pain coping strategies and encourage their use following surgery. The impact upon pain intensity and distress, anxiety and analgesic consumption was significant in the experimental group. Interestingly, as highlighted in chapter five, subjects in the control group employed a similar range of coping strategies. Explanations other than those related directly to the coping literature therefore need to be considered. The following section examines the relationship between acute pain self-generated strategies and self-efficacy. 7.3 The Role of Self-Efficacy Both the Gate Control Theory (Melzack and Wall, 1965, 1996) and McGuire’s (1992) multidimensional framework highlight the interplay of factors associated with pain perception. It is not just the intensity of pain which needs to be taken into account in considering postoperative pain experiences, but also psychosocial processes. Perceived self-efficacy is one psychological mechanism that has been shown to moderate physiological perception and health related actions (Bandura, 1986, 1977, 1997). As was explained in section 1.5, self-efficacy refers to the beliefs that persons have in their abilities to engage in particular actions, such that desired goals are achieved. It is related to perceptions of control, the particular actions that individuals choose to take, how long they will engage in that action and how much effort they are prepared to expend. Bandura et al (1987) suggest that there are a number of ways in which self-efficacy theory may be relevant to pain management. Individuals who believe that they can moderate their pain experiences are likely to employ strategies, which they have previously learnt, may help relieve pain. Those who question their abilities, will either not engage in such activity or are likely to give up quickly if their attempts lack success. Self-efficacy is thus an important factor in both initiating and continuing pain coping strategies, whether state or trait dependent (Bandura, 1986). Beliefs in one’s ability to influence pain may also reduce anxiety and thereby physiological stimulation and muscle tension (Bandura et al 1987). Diverting attention away from pain and into attempts to cope with the experience has been shown to further reduce pain perception (Williams and Kinney, 1991). The implication of self-efficacy theory, for the present study, is that patients who believe in the ability of their self-generated strategies to reduce postoperative pain are likely to experience less anxiety, pain and distress and require fewer analgesics. On the whole, this was the case for patients in the experimental group. However, subjects in the control group were also employing self-generated strategies in attempts to relieve their postoperative pain. The difference between the groups is that experimental subjects’ strategies were overtly identified and supported by the nurses. Bandura (1997) believes that self-efficacy can be enhanced through skills mastery, vicarious experiences, verbal persuasion and the provision of information. Vicarious experiences refers to the action of observing other’s success or failure and processing this information so that it impacts upon one’s own perceived ability. The differences between the control and experimental group could not be explained on this basis. However, the other three concepts may offer some insight as the following discussion explains. Subjects in the experimental group were “informed” that they could continue to use their pain relieving strategies in the hospital setting and were shown where and how to access any resources they required. Suggestions were also made as to how strategies could be adapted from home to hospital. For instance, hot water bottles are not allowed on the Ward, however, heat pads could be provided through the physiotherapists. The ability of these patients to initiate and maintain their strategies in a hospital setting was thus facilitated through the provision of information. It could be argued that skills mastery was also enhanced through this approach. At a more abstract level, it is conjectured that overtly identifying and supporting individual’s strategies values the patient as an expert in their pain management. In this sense, patients are regarded as possessing social power which has arisen as a result of increasing knowledge, skills and experience (Buchmann, 1997) in the employment of pain coping strategies. If patients are to be able to exercise their social power, nurses need to act as “significant others” who hold the patient and their actions in genuine high regard. Buchmann (1997) outlines how the work of Rodin and Janis (1979) may be applied to enhance skills mastery and, thereby, self-efficacy. The following discussion applies some of these ideas to the study findings and the concepts of skill mastery, verbal persuasion and patient-centred care. A number of actions are seen as providing the essential building blocks to allow practitioners to acknowledge patients as experts and enable them to exercise their social power. Firstly, patients need to be encouraged to self-disclose. Identifying patients’ self-generated strategies is therefore an essential element of the process. The action of identification enables nurses to gain insight into the individuals’ thoughts and actions surrounding pain coping and to indicate that engaging in self-generated pain coping is a worthwhile activity. On its own, this action grants ‘permission’ for engaging in coping strategies and may reduce the problem found by Copp (1974), whereby individuals stopped using their strategies on the basis that they feared practitioner disapproval. Following strategy identification, the nurse needs to demonstrate genuine interest in providing the support that the patient requires. This will necessitate discussion and negotiation in equal partnership. Thus, the nurse uses her knowledge of the setting, and postoperative pain management, to work with the expert patient in developing a mutually agreeable plan of activity. It is feasible that this may entail no more than showing patients who, for example, employ distraction, where the day room and television are, or informing those who prefer peace and quiet that they may draw the curtains around the bed. Buchmann (1997) further outlines that selective positive feedback is important in the maintenance of action. In addition to verbal enquiries regarding continuation of strategy use and efficacy, the visual impact of reductions on pain intensity scales, which was positively commented upon by some of those involved in the RCT, may be relevant. Visualisation may also help to underscore the dual importance of self-determination and responsibility in the use of expert, self-generated pain coping strategies. The premise of the above account is that if the nurse follows this type of action, the patient’s belief in their pain coping strategies can be enhanced and individuals will be enabled to exercise their expert power. Although verbal persuasion is regarded as a relatively weak form of enhancing self-efficacy, Bandura (1997) believes that if people are ‘persuaded’ that they possess the skills to master difficult situations, and are provided with the necessary resources to take action, they will engage in strategy use with greater effort and for longer periods. If the pain coping strategy works, skills mastery and perceived ability to control the situation will be further enhanced. It is suggested that the experimental patients were exposed to this type of support. Conversely, patients in the control group were not exposed to this activity and potentially empowering relationship. As a consequence, they were unaware of whether they could use their strategies in hospital and, if they did so, of the ensuing value for postoperative pain management. Without practitioner encouragement, it could be conjectured that control patients may have given up using their strategies at the first sign of failure. However, it should be noted that this was not assessed and would need to be investigated in any future study. In summary, it is postulated that self-efficacy and personal power were not promoted in the control group. Conversely, patients in the experimental group were regarded as possessing a degree of mastery in pain coping and, through information giving, negotiation and verbal persuasion, were enabled to perceive their strategies as potentially efficacious in the management of postoperative pain. This argument is consistent with the direction of findings in the chronic pain literature, whereby perceived self-efficacy can reduce perceptions of powerlessness, enhance control and, as a consequence, improve pain tolerance (Barlow et al, 1996; Keefe et al, 1996, 1997a; Arnstein et al, 1999). Self-efficacy theory may therefore offer an explanation for the different outcomes seen between patients in the control and experimental groups. Further research is needed to test the relative importance of self-efficacy in acute pain coping. This would require the employment of measures of self-efficacy, such as the Pain Self-Efficacy Questionnaire (PSEQ) (Williams et al, 1993). However, although tools such as the PSEQ have undergone relatively rigorous psychometric testing with chronic pain populations, they may not be applicable to acute pain. It may therefore be necessary to develop tools specifically for this group before self-efficacy can be measured in acute pain settings. If self-efficacy theory is accepted as providing one explanation for the differences between the control and experimental group, it is possible to conjecture a further proposition: namely, that by enhancing self-efficacy and personal power, patient-centred care in relation to pain management will be promoted. Identifying and supporting self-generated pain coping strategies appears to overtly endorse patient choice, encourage partnerships and involve patients as active participants in decision making and care management related to postoperative pain control. These concepts are all central features of patient-centred care (Stower, 1992; Campbell, 1998; Cahill, 1998) and, given the study findings, appear to fit well with Department of Health exhortations that involving patients in care decisions and management constitutes good practice (1989, 1998). However, three cautionary points must be noted in connection with the study. Firstly, Bandura et al (1987) have examined perceived self-efficacy in relation to cognitive coping strategies and endogenous opioid mechanisms. They concluded that although self-efficacy can enable people to tolerate high levels of pain, increasing reliance on coping strategies can, in some cases, result in pain reaching such a level that at the point of perception it is immediately unbearable. Relying solely upon non-pharmacological strategies would therefore seem unwise when pain is expected to be severe. At several stages in this thesis, the point has been made that analgesics are a unidimensional approach to pain management. That is not to say that they are not useful and, indeed, it was clearly articulated in section 2.7 that the aim of this study was to add to, rather than detract from, the therapeutic approaches available for relieving postoperative pain. Thus, the identification of self, family and practitioner-generated strategies incorporated analgesics as one of the themes, while experimental patients were enabled to maintain their self-generated analgesic management postoperatively. An arrangement, which combines pharmacological and non-pharmacological techniques, is therefore regarded as appropriate for managing the multidimensional phenomenon of postoperative pain. This should ensure that best use is made of both analgesics and patients’ expert non-pharmacological strategies. In this way, patient-centred care clearly relies upon a partnership of patient and professional expertise. The second caution relates to the evidence, outlined in chapter two, that not all patients want to adopt active roles. The ability of this study to differentiate between self and practitioner generated strategies, as determined by the patient, allowed those who preferred more passive roles to be accommodated. For example, those choosing to rely upon practitioner generated cognitive and expressive reassurance, were able to seek nurturing rather than empowerment. The implication of this action is that patient-centred care in postoperative pain management can be achieved without forcing individuals to take responsibility for their actions. This was regarded as important given the surgical nature of the setting and the age range of those admitted to the Ward. As a number of authors have noted, acutely ill and older patients have been found to prefer less active participation (Biley, 1992; Thompson et al, 1992; Beaver et al, 1996) . Any future study would need to consider undertaking an analysis of variance to explore the degree to which variables, such as age, influence both strategy use and outcomes. An additional point, and one that may have had a positive impact upon the study, is the evidence which suggests that actions not requiring medical expertise are viewed more favourably by patients as opportunities for active involvement (Biley, 1995; Caress, 1997). With the focus of the study being patient expertise, medical expertise for action was not an issue. In section 7.2, it was suggested that enhancing personal control could alter pain coping styles by empowering patients to be proactive and vigilant in their strategy use. The third cautionary point relates to how such an action might affect postoperative outcomes. The literature indicates that avoidant styles may in fact enhance outcomes in acute settings (Endler et al, 1993). Approaches that emphasise vigilance could, therefore, be detrimental. Further research is needed in this area in relation to the use of self-generated strategies. However, it is proposed that if patients are treated as individuals by emphasising their expertise and freedom to articulate strategy preferences, any potential negative impact may be negated. This section has considered the relationship of the study findings to self-efficacy and certain elements of patient-centred care. The next section takes this latter perspective further by examining issues such as individuality, partnerships, emancipation and empowerment. 7.4 Individuality and Partnerships, Emancipation and Empowerment Acknowledging patients’ self-generated pain coping strategies as pertinent to postoperative pain management, provides a benchmark for understanding the degree to which participation in decision making and pain management can be individualised. As was highlighted in chapter two and commented upon in the section above, one of the difficulties with patient-centred care is the degree to which individuals want, or are able, to be active participants. If the patient’s pain coping repertoire is exposed through the use of a self-complete questionnaire, followed by discussion, the nurse can gain insight into how the patient wishes to behave on a continuum from wholly passive to highly active. As Elliott and Turrell (1996) suggest, adopting a philosophy of individualism allows ‘active’ participation to be passive, if that is what the patient wants. However, this model assumes that patients will be willing to identify their strategies to the nurse in the first place. Administration of the Pain Strategies Questionnaire (see section 7.6) could in itself be seen as paternalistic and controlling, as it forces the patient to share their strategies with the professional. As practitioners learn more about the person’s experiences, the danger is that they could professionalise the activity and seek to take control. It is interesting to hypothesise whether this philosophy may be behind the attempts to teach non-pharmacological strategies for acute postoperative pain management, despite awareness in the literature that individuals possess spontaneous strategies. As a consequence, unique personal experiences could be gradually denied by professionals who institutionalise the original person-centred behaviours. Data obtained during the focus groups, in relation to pharmacological pain management, suggested that the nurses saw themselves as being in control of the patient. For example, statements were made such as “you give them analgesics” and “we do let them self-medicate” (see section 6.2.6). If verbal persuasion, regarding non-pharmacological management, is added to this scenario in an attempt to enhance self-efficacy, the result may be a legitimisation of power back in the hands of the professional. Thus, not only would there be a danger of paying lip service, at best, to the ideals of patient-centred care, but the original intention of the study to free patients and nurses from constraining ideologies and enable empowered action would also be put at risk. The importance of working with nurses to enable them to critically challenge their beliefs and attitudes and to be involved in identifying and testing new ideas for praxis, cannot therefore be under-estimated in terms of the study findings. The generalisability of the RCT outcomes must therefore be located within the overall methodological approach of focusing upon exposing ideologies, critically reflecting upon thoughts and actions and viewing patients as partners in postoperative pain management. Although the NHSME (1993) regard information giving as the most crucial factor in empowering patients, Elliott and Turrell (1996) argue that it is the type of partnership negotiated between the practitioner and individual which is the key element. In the current study, the partnership style certainly appeared the more important issue as it determined the style of communication adopted and was antecedent to information exchange. At the start of the study, it was clear that while nurses attempted to develop partnerships with rheumatology patients, that could have been located at Elliott and Turrell’s (1996) equal partner or most important team member levels, orthopaedic patients tended to be viewed as advisors or collaborators. It is suggested that the survey findings and the critical discussions which took place during the cycle one staff workshops, enabled consciousness to be raised regarding the potential of working in equal partnerships with orthopaedic patients. According to Elliott and Turrell (1996), the nurses’ role in an equal partnership is to trade ideas and views and ensure that the patient is treated as an equal. Contrasting this level with collaborator partnerships, equal partners are invested with the right to participate. Furthermore, their influence in decision-making is regarded as equal, rather than subsumed, to that of practitioners. As patients came to be regarded as experts in pain coping, pain management relied upon developing partnerships which commenced with an invitation for the patient to disclose their strategy use preoperatively. The exchange was thus patient, rather than nurse, centred. Postoperatively, the equal partner philosophy allowed patients to make best use of their self-generated strategies in conjunction with practitioner generated pain management strategies. Had the most important team member level been adopted, patients would have been expected to lead, direct and control the decision-making process. This was felt to be inappropriate in a setting where practitioners need to determine pain control while patients are still recovering from anaesthesia. It may be that in the immediate postoperative period, a collaborator partnership is necessary. As the patient recovers, equality becomes possible. Whether patients can ever become pre-eminent members in acute surgical settings is questionable, not only from the patient’s perspective, but also from the practitioners. Considerably more emphasis would need to be placed upon working with staff and patients to identify patients’ wishes, examine practitioners’ ideologies and concerns and critically consider the possibilities for patient-led decision-making. Certainly, it would be difficult to confine such activity to one aspect of care. Rather the totality of the recovery episode would need to be deliberated. One of the difficulties with a study such as the one undertaken here, is the issue of whether one can actually empower people in just one aspect of their lives. Saunders (1995) believes that it is appropriate to view patient-centred care from the perspective of patients taking responsibility for specified actions, thus in this instance, for self-generated, non-pharmacological pain management. While this perspective may be reasonable, tensions arise if a jump is then made to claiming that the social action of taking responsibility equates to empowerment. For empowerment to occur, individuals need the freedom to make choices, to believe that they have the power to act and be satisfied that the expenditure of effort is worthwhile. The patients involved in the study, therefore, needed to be freed from the constraints of wondering whether practitioners would disapprove of their self-generated strategy use and be encouraged and enabled to engage in their strategies postoperatively. The nurses’ role was to enhance self-efficacy beliefs while respecting patients’ individualism and right to self-determination. In this way, patients could be empowered to decide whether or not they wished to act. While this clarification may provide insight into relationships between freeing people to act, pain coping and pain self-efficacy, it still does not fully address the issue of whether empowerment in one aspect is reasonable. Traditionally, postoperative pain management has focused upon the administration of analgesics by nurses, who are able to control when and how assessment takes place, the timing of administration, the dosages given and the process of evaluating pain relief. With the exception of attempting to maintain experimental patients’ usual analgesic prescriptions, the nurses in the study retained control for pharmacological management. Nevertheless, the findings indicated that patients, who were encouraged to participate in their care and accept the authority to engage in self-generated, non-pharmacological strategies, experienced less anxiety, pain and distress. It is therefore conjectured that it is possible to empower individuals to act in a specified context and achieve measurable improvements in health related outcomes. The psychosocial impact of this activity, however, needs additional investigation. The discussion above has begun to link the study findings to certain concepts related to critical theory. The following section takes this discussion further. 7.5 Acute Pain Management: The Beginnings of a Critical Theory As previously discussed in chapter three, critical theories focus upon the meanings given by individuals to events within the contexts of the events. The experience of pain prompts people to develop coping strategies based upon past learning, present context, the intensity of the stimulus and other multifaceted issues. Historically, patients have tended not to engage in self-generated strategies during hospitalisation, because, according to Copp (1974), they fear practitioner disapproval. More recently, Chaves and Brown (1987) indicated that some patients continue their strategy use without practitioner knowledge. The current study found that all patients involved in the RCT had a repertoire of pain coping strategies that were used in the hospital setting with, or without, the nurses’ knowledge. The apparently increasing trend for patients to use their own strategies may be linked to the growing social and political focus upon consumerism and individual rights (see section 2.5). Nevertheless, practitioners are, on the whole, unaware of patients’ activities and the emphasis in postoperative pain management has remained upon medical staff prescribing and nursing staff administering practitioner determined analgesics. The contention of this study is that patients can be freed from constraining ideologies to act in partnership with practitioners in the management of postoperative pain. The emancipation of patients in this activity relies upon nurses valuing the patients’ expertise and patients’ valuing non-pharmacological strategies. Through discussion and information, which focuses upon the potential of self-generated strategies to influence pain related outcomes, nurses may enable patients to accept the authority to determine their preferred role in postoperative pain management. This role may be active, with patients engaging in, for example, autonomous behavioural strategies and patient determined analgesic management goals, or more passive, for instance, where patients seek cognitive and expressive reassurance and analgesic management from others. The social reality is reconstructed as the power for determining the patient’s role is transferred from the nurse to the patient. The patient is encouraged to think critically about their strategy repertoire and identify the actions that they wish to engage in during the postoperative period. Through discussions with the nursing staff, resources are identified which will enable the patient to employ their strategies within the twin contexts of the hospital setting and pharmacological postoperative pain management. As a consequence, practice can be transformed, as self-efficacy and coping are enhanced and patients participate in a dual approach that combines patient-centred non-pharmacological management with practitioner facilitated analgesic management. Section 7.7 continues the discussion of the study processes and findings in light of critical theory and from the nurses’ perspective. However, the next section briefly considers issues surrounding the development of the Patient Strategies Questionnaire in order to complete the discussion of findings from the patients’ standpoint. 7.6 The Pain Strategies Questionnaire The Pain Strategies Questionnaire (PSQ) was originally developed from the survey data as a tool to examine strategy use during the RCT. The survey identified some 43 different strategies. These were reduced to 30 by, for example, amalgamating all pharmacological items into one strategy theme termed ‘analgesics’. A similar pattern was followed for cognitive control, general help and miscellaneous strategies. The remaining items were listed by theme. The themes themselves were generated from the data rather than by using existing classifications. This was deliberate as many of the classification systems have been developed in North America and it is possible that cultural differences exist in the types of pain coping strategies used by UK patients. Furthermore, most existing systems relate to strategy use in chronic pain. The application of taxonomy from acute to chronic pain was deemed inappropriate at this early stage in generating theory about acute pain coping. This premise was later supported by the apparent lack of catastrophising in the survey sample when discussing acute pain strategies. Examples of other questionnaires, which have developed classifications that could potentially have been used for the present study, include the Cognitive Coping Strategies Inventory (CCSI) and the Coping Strategies Questionnaire (CSQ). The CCSI, developed by Butler et al (1989), was generated to identify acute pain coping styles. However, the categorisation focused upon cognitive coping and was specifically linked to trait dependency. As previously argued, there was evidence in this study that coping with everyday and acute pain has state, as well as trait, elements. Furthermore, the survey highlighted that acute pain coping is a combination of cognitive and behavioural techniques. Comparisons between the CCSI classifications and those generated for the PSQ were thus felt to be limited. Conversely, the CSQ was developed to assess both cognitive and behavioural pain coping and linked to trait and state dependency. Although commonly used in chronic pain management and research, the CSQ has been tested for utility in acute pain contexts. For instance, Buckelew et al (1992) used an adapted version when examining pain coping in patients undergoing electrodiagnostic studies. The issue with the CSQ is that it does have to be adapted as some statements are not relevant to acute pain. Furthermore, categorisation includes catastrophising and praying, neither of which were frequently used by the acute pain population in this study. Thus comparisons were again felt to be limited. The classification used in the PSQ was deemed to provide an accurate picture of the range and type of strategies identified during the survey. The survey sample was representative of the population attending surgical and orthopaedic out-patient clinics in preparation for booked surgery at a rural, district general hospital, in the south of England. The area is relatively affluent and class distribution is skewed towards the higher social classes. Furthermore, just 3% of the local population are non-Caucasian. Application of the PSQ to UK based black, Chinese or Asian communities may therefore be limited. In administering the original PSQ to the RCT sample, several factors were considered. Firstly, as the survey had involved mainly surgical, rather than orthopaedic outpatients, it was felt to be important to record the miscellaneous category as ‘other’ and ask patients to state what they use when in pain. In this way, if the survey findings were not generalisable to orthopaedic in-patients, a tool that was unable to accurately record individuals’ self-generated strategies would not invalidate the RCT. Equally, this approach enabled the researchers to examine the RCT data for additional strategy themes. In fact, few miscellaneous strategies were recorded and those that were had already been mentioned in the survey. No new themes were identified through the RCT data. The PSQ allowed nurses to develop a patient-centred approach to care by enabling experimental subjects to identify their pain relieving strategies. The listed items acted as a prompt in discussions between the patient and nurse on the best way to utilise the strategy in hospital, any resources required in terms of equipment and the support needed from nursing or other staff. An integral part of this process was mutual sharing and learning. Thus patients informed nurses of their needs and nurses informed patients of the potential for supporting strategies. Actual strategy use could then be negotiated and implemented. The PSQ can therefore be used not simply to identify self-generated pain coping strategies, but also as a catalyst for patient-centred pain management. The previous five sections have considered the study findings from the patients’ perspective and presented a new action theory related to the employment of self-generated coping strategies for patient-centred postoperative pain management. The chapter now moves on to consider the nurses perspective and methodological issues. 7.7 Emancipatory Praxis: Furthering the Critical Theory Perspective The intention of this section is to highlight the critical social theory that was generated as nurses collaborated in the research study and sought to transform practice from a practitioner centred concern, to one that reflected the tenets of patient-centred care. At the start of the study, nurses were interested in improving pain management on the Ward and addressing some of the surrounding practice and organisational constraints that they regarded as existing within the setting. Titchen (1993) emphasises that action research must be grounded in practitioners concerns. Although the nurses were encouraging rheumatology patients to maintain their self-generated strategies, it had not occurred to them that orthopaedic patients might also have pain coping repertoires. Consequently, they had no reason to address this aspect of pain management and the catalyst for the research was thus an externally identified idea. Nevertheless, the nurses were concerned about the limitations of managing a multidimensional phenomenon from a unidimensional perspective and were keen to examine the potential for widening their approaches to pain relief. They were therefore open to focusing upon questions about the use of self-generated strategies and to considering the potential of working in partnership with orthopaedic patients. In their recent review of the action research literature, Meyer et al (2000) highlight lack of ownership as one of the main factors inhibiting change. The fact that the team was open to taking ownership of the problem is viewed as highly relevant to the success of this study. Entering the setting with a previously identified research idea, appears to locate the initial phase of the study within a technical approach to action research. As previously mentioned in section 3.6.4, technical approaches are based upon the natural sciences and linked to positivism (Holter and Schwartz-Barcott, 1993). The fact that a survey and RCT were subsequently employed, added strength to the argument that the study was technical in nature. However, the incorporation of consciousness raising, reflexivity and praxis, placed the study within an enhancement approach where the emphasis was upon emancipatory philosophies and critical theory. Attempting to work in this way on an extremely busy acute orthopaedic/trauma ward was demanding and not always feasible. Models, such as Titchen’s (1994) collaborative framework, had to be employed to enable the study to move forward within the realities of practice. This, in its self, raised philosophical and practical tensions, as not all nurses could be involved in on-going critical discussions regarding either the practice transformations or research protocols. It would therefore be unwise to claim that each participant internalised new ideologies and approaches to practice. Nevertheless, considerable evidence was available that many nurses did reconstruct their thinking and practice and a new team ideology was definitely manifest in the way in which praxis developed. Altering ideologies depends upon exposing values, beliefs and norms so that false ways of thinking can be challenged and new ideologies developed. The focus groups and early field notes highlighted that contradictions existed throughout the three senses: namely, the generic sense which is concerned with the origins of and motivation for beliefs, epistemic sense which incorporates empirical evidence or self-fulfilling prophecies and, the functional sense, which is based upon social justification of beliefs (Geuss, 1981). In the generic sense, the nurses were constrained to viewing postoperative pain management from a traditional, pharmacologically led perspective. Although the Ward philosophy was based on the current view that patients should be treated as individuals, normal routines and specific needs tended to be subsumed to the interests and priorities of professional practice. Combined with this, the professionalisation of acute pain management meant that practitioner determined interventions, whether pharmacological or non-pharmacological, were assumed to offer the normal route to pain relief. Frustrations were apparent as the nurses felt that they should be able to teach non-pharmacological strategies to patients but were unable to find the time to do so. These frustrations were added to by beliefs that organisational constraints could not be challenged as routines were too entrenched and, in comparison to the nurses, both the organisation and medical staff were too powerful. As the study progressed, the process of critical reflexivity led to a raised consciousness, challenging of some of the constraints and the formulation of ideas for praxis. For example, relationships with the medical staff were improved as nurses began to feel more confident in their knowledge base and able to communicate with doctors as equal professionals. In addition, the nurses came to appreciate that patients enter hospital with a repertoire of self-generated strategies which could be employed by the patient to enhance postoperative pain management. Encouraging strategy use was not regarded as challenging the nurses’ power. Rather, it was viewed as promoting a patient-centred approach to pain which was consistent with the Ward’s written philosophy, the expressed preference of many of the staff and the current NHS agenda. Thus new motivations were developed for practice. In the epistemic sense, there was, from the beginning, a commitment to identifying an empirical basis for practice and praxis. Even the nurses who were not particularly taken by the idea of pain management, were overtly interested in being involved in a research project. Interest clearly waxed and waned over the four years, but as the discussion in sections 6.4.2 and 6.4.3 highlight, the nurses were committed to the study and motivated by the twin focus upon pain management and patient-centred care. They were also specifically committed to testing their ideas, believing that this would provide an evidence base for praxis and empower them to transform practice. This issue is addressed further in section 7.8, however, an important point here is that the motivation for this epistemic belief lay within the acute care context and the increasing focus upon evidence based practice within the NHS which has been previously referred to. Nurses’ functional ideologies also altered as the study progressed. Initially, roles in pain management were clearly delineated with doctors prescribing, nursing staff both influencing prescribing and taking responsibility for administration and patients requesting or receiving treatment. Different power relationships, as outlined in chapters five and six, underpinned this practice. As the study progressed, and nurses became increasingly critical of their ideologies and assured in knowledge base, the power games began to lessen and patients were increasingly regarded as equal partners. The study thus promoted emancipation and informed action. Nurses were freed from constraining beliefs and, through critical reflexivity and praxis, were able to develop the knowledge and confidence to accept the power to act. As a consequence, they were then able to develop more equal partnerships with both patients and medical staff. The ability of emancipatory designs to produce such transformations has been recognised within the nursing literature by, for example, Barker (1992), Hart and Bond (1995), Meyer, (1995a) and Bellman (1996). However, other authors have encountered difficulties with using this approach in health care settings, suggesting that factors such as lack of time, energy and resources, reluctance to change and an unstable workforce mitigate against change taking place. In addition, lack of multidisciplinary teamwork, talking, support, leadership, knowledge about the study and commitment to the study have been shown to be inhibiting factors (Meyer et al, 2000). The context of an action research study is, therefore, highly pertinent. Several of the problems listed above were relevant for the study setting. The Ward has a reputation for being one of the busiest in the hospital and time and energy are therefore always an issue. Furthermore, despite communication with the multidisciplinary team and the interest shown by the medical staff, in particular, the study was focused purely upon patients and nurses. This issue is referred to again in section 7.8. On the positive side, commitment was high and leadership strong. There was an explicit interest in practice and personal development, the workforce was relatively stable and the team exceptionally supportive. Communication was good, within both the nursing team and within the multidisciplinary team. The co-action researchers and collaborative group were enthusiastic and the informants were interested. The length of the study allowed for periods of reduced activity if stress levels were high or staffing was low. Both alone and in combination, these factors have been shown to facilitate practice change during action research projects. (Meyer et al, 2000). Slightly different factors have been found to influence change when the topic, rather than setting is considered (Meyer et al, 2000). Knowledge, skills and confidence, commitment, patient support and role modelling have been found to facilitate change, while lack of knowledge, skills and confidence, lack of suitable patients and lack of patient support have been found to inhibit change. At the beginning of the study the two co-action researchers and two of the collaborative group members were able to share their knowledge and skills in pain management through discussions and role modelling. As the Ward nurses began to talk about pain and its management, and consciousness was raised, knowledge and confidence increased exponentially. The process was viewed as developmental and emancipatory and was thus part of the study process. The nurses’ commitment to improving pain management was evident and patient recruitment was satisfactory in terms of the interest shown by patients in the study topic. The current worldview on pain is one that regards patients as experts in their pain experiences (Carter, 1998). If nurses can be freed from constraining ideologies and practices to acknowledge patients as experts not only in terms of pain perception, but also in self-generated pain coping strategies, postoperative pain management can be improved. This improvement can be seen in relation to measurable outcomes such as pain and anxiety and in the more abstract concept of promoting patient-centred care. The final section of this chapter considers some of the methodological issues contained within the study. 7.8 Evidence Based Praxis: Methodological Issues With its situation specific nature, action research does not aim to develop generalisable knowledge in the positivistic sense. However, findings may be transferred to different contexts if they are viewed by the people within those contexts as fitting their experiences (Appleton, 1995). The process findings related to the factors which inhibited, or facilitated, change in the present project, were not dissimilar to those of other action research studies. It is therefore suggested that the process outcomes support the existing literature and provide a good functional example of the factors that promote change. The study is thus claimed to have external validity in relation to change theories (Meyer et al, 2000). The nature of the survey and RCT should provide external validity in the generalisable sense. Previous research into self-generated strategies for acute pain has, on the whole, been limited to describing strategies rather than testing their effect on postoperative outcomes. The claim of this study is that where patients strategies are identified upon admission and subsequently supported through valuing patients’ expertise and negotiating the required resources to enhance self-efficacy and strategy use, anxiety, pain intensity and pain distress will be reduced and analgesic consumption will be altered. The RCT intervention contained two discreet elements. Firstly, patients’ strategies were systematically identified preoperatively through the administration of the PSQ, followed by discussion and negotiation between the patient and nurse. This intervention was rigorously maintained throughout the study. Conversely, the second intervention, which focused upon nurses actively supporting patients’ strategies, as and when, required postoperatively, was less systematic. Two main reasons existed for this, namely lack of time and lack of need. It is plausible that nurses were covertly controlling patients by not finding the time to support strategies postoperatively. However, there was no indication that this was the case. Indeed field note data from the first eight weeks of the RCT highlights the concerns of both the researchers and nurses that strategies were being successfully identified, but not systematically supported. Critical discussions offered some insight into the problem. Although time was an undeniable issue, the main reason for the apparent lack of systematic support related to the independence of patients in engaging in their strategies. Indeed the amount of active support required by patients was relatively minimal. Put simply, in most cases all that was necessary was for nurses to value the patients actions, ‘give permission’ for strategies to be used in hospital and point people in the right direction in terms of safe access to resources. This might entail little more than showing patients or relatives where the tea machine or day room was, how to get out onto the balcony and where to find bowls and warm water. Added to this, nurses requests to patients to, for example, “remind us, if you are in pain, that you would like to be go to the day room to watch TV, rather than just be walked around the Ward”, appeared to be sufficient. The simplicity of this approach should not be underestimated in terms of its impact. Not only were improvements seen in several of the outcome measures, but patients were also actively involved in a patient-centred approach encompassing choice and participation in care. By recognising that patients frequently required little active support, the team developed a sense that enforcing support was in itself controlling. Wade (1995) warns that paternalism can undermine patient-centred approaches, while several other authors comment upon the failure of practitioners to enable choice and a shift of power. If information is provided to enable patients to determine whether or not they wish to use their strategies in hospital, nurses offer support at the patient’s request. On the one hand it appears that administration of the PSQ, combined with overt valuing of patients’ strategies, could be sufficiently generalisable to enable nurses in other acute pain settings to enable patients to engage in their self-generated strategies in order to lessen postoperative pain and anxiety. The involvement of the Acute Pain Sister at the study hospital is enabling the PSQ to be applied in other wards. The impact of this action will be audited and the results will enable further consideration of the degree to which praxis developed in an emancipatory action research study can be transferred to areas where the nurses have not been exposed to an on-going critical reflexive process. Thus the wider implications regarding evidence-based practice remain open to question. Nevertheless, the study has added to the literature about postoperative pain management in a new and fundamentally different way. Furthermore, the emphasis upon patient-centred care and the adoption of the RCT, reflects current NHS concerns regarding the need to ensure that practice is both clinically and cost effective while emphasising patient choice and the right to participate in decision making and care management. In light of the subjective nature of both pain management and patient-centred care a further caution needs mentioning in relation to the study methods. Pain is a subjective phenomenon and while pain intensity and distress are undeniably important to postoperative patients, the study has only told part of the story regarding the use of self-generated pain coping strategies. Previous sections of this chapter have advocated that further work is undertaken in relation to pain coping and self-efficacy. The danger is that this work could be confined to further RCTs and, as a consequence, focus solely upon objective measures and statistical significance. It is suggested that additional research should incorporate the completion of daily diaries as advocated by Keefe et al (1999). These could be used to gain insight in when and how patients engage in their strategies, coping strategy maintenance issues, perceived control and perceived clinical efficacy. The latter point is particularly pertinent given that the RCT findings, although statistically significant, averaged just one point difference on the 0 - 10 scale. The final area, which should be considered in future research, is the need to adopt a multi, or inter-professional approach. Carter (1998) suggests that pain management provides a good example of practice where professionals should and could work more closely together. Opportunities existed within this study to actively involve the Ward’s medical staff. However, these were not take up due to time constraints. Although the nurses’ perceptions and the field note data suggested that relationships improved with medical staff and that the doctors were genuinely interested in the study, it is difficult to ascertain whether this was truly the case. The relevance of this is whether practice transformations can be maintained when only one professional group has been involved in considering an interprofessional issue. 7.9 Recommendations and Summary Recommendations for future research have been made throughout this and previous chapters. In summary, further work is needed regarding: acute pain coping in relation to trait and state dependency the influence of significant others in postoperative pain management the relationship, if any, between empowerment and vigilance on the one hand and nurturing and avoidance on the other the relevance of self-efficacy to acute pain coping and the development of scales to measure this psychological mechanism in acute pain contexts development and refinement of the PSQ the cost benefits of including patients self-generated strategies in postoperative pain management Clinically, nurses need to consider widening their approaches to postoperative pain management in an attempt to reflect the multidimensional nature of pain. The PSQ could be implemented relatively easily in many postoperative settings. However, if this is undertaken without a concurrent focus upon nurse empowerment, the transferability of the patient related findings would need to be carefully assessed. This could be managed through nurse-led audits. Education needs to be focused upon freeing nurses from their constraining ideologies around pain management. In addition, nurses need to be offered specific examples on how to operationalise the concepts of patient-centred care. The findings from this and other similarly focused studies need to find their way into the educational setting. In summary, individuals develop a repertoire of pain coping strategies through exposure to everyday pain experiences. When faced with more serious pain, they re-appraise the situation and their strategy repertoire. Techniques that have worked well in the past are re-employed, with the caveat that there appears to be a hierarchy of strategies according to pain intensity. If nurses identify patients’ preferred strategies on admission to hospital and encourage them to engage in self-generated strategies postoperatively, they can enhance patient-centred care and improve postoperative outcomes. By participating in emancipatory action research, the nurses themselves can become empowered to consider their ideologies and work in partnership with patients. CONCLUSION Traditionally, acute postoperative pain management has been dominated by a way of thinking that places the practitioner in control through the provision of analgesics and taught non-pharmacological strategies. Within the last 15 years, there has been an increasingly strong suggestion that it may be more appropriate to match pain relieving interventions with patients’ coping repertoires, than to teach strategies without regard to individuals’ preferences (Chaves and Brown, 1987; Stevens et al, 1987; Butler et al, 1989; Pick et al, 1990; Buckelow et al, 1992; Rokke and al’Absi, 1992; Burrows, 1997b; Carr and Thomas, 1997; Kashikar-Zuck et al, 1997; Miro and Raich, 1999). Similarly, there has been an increasing political focus upon the ideals of patient-centred care with its central tenets of increasing patient choice, involvement in decision-making and participation in care management (DH, 1989, 1998; NHSE, 2000). It could be argued that these two notions are mutually compatible, in so far as they value the patient as an expert and aim to empower them to be actively involved in health care. Added to this, the evidence based practice movement emphasises the importance of clinical effectiveness and the involvement of nurses in clinical governance initiatives (DH, 1998) The study involved nurses and patients, action research and critical theory, a survey and RCT, focus groups, critical discussions and questionnaires. The multiplicity of approaches was required to address a complex issue with different components. The collaborative action research stance enabled a group of nurses to be actively involved in undertaking research, producing evidence for practice and adopting new praxis. The outcomes included improvements in postoperative pain management and in patient-centred care. The uniqueness of the study is that it has demonstrated that if patients are engaged in their postoperative pain management, through a focus upon self-generated pain coping strategies, postoperative anxiety, pain intensity and distress can be reduced. By facilitating nurse empowerment, the study has enabled nurses to adopt a broader approach to pain management and locate their practice within patient-centred care. At the beginning of the 21st Century, patient-centred approaches should form part of all nurses’ pain management practice. References Abude, F.E., Asare, H. and Adiamah, C. (1999) The effect of health education to surgical patients individually about anaesthesia and post operative pain. Abstracts 9th World Congress on Pain, Vienna, August Adelman, C. (1993) Kurt Lewin and the origins of action research. Educational Action Research 1(1), pp.7 – 24 Affleck, G., Tennen, H. Pfeiffer, C. and Fifield, J. (1988) Social support and psychosocial adjustment to RA. Arthritis Care and Research 1, pp.71 – 77 Ahles, T.A., Blanchard, E., and Ruckdeschel, J.C. (1983) The multidimensional nature of cancer-related pain. Pain 17, pp.277 - 288 Allcock, N. (1996) The use of different research methodologies to evaluate the effectiveness of programmes to improve the care of patients in postoperative pain. Journal of Advanced Nursing. 23, pp.32 - 38 Anand, K.J.S. and Craig, K.D. (1996) Editorial: new perspectives on the definition of pain. Pain 67(1), pp.3 - 6 Appleton, J.V. (1995) Analysing qualitative interview data: addressing issues of validity and reliability. Journal of Advanced Nursing 22, pp.993-997 Aristotle De Anima, Book II, Chap 2, Book III, Chap 1. Translated by Smith, J.A. and Ross, E.D. (1931) Clarendon Press, Oxford Armitage, P., Champney-Smith, J. and Andrews, K. (1991) Primary nursing and the role of the nurse preceptor in changing long-term mental health care: an evaluation. Journal of Advanced Nursing. 16, pp.413 - 422 Arnstein, P., Caudill, M., Mandle, C.L., Norris, A. and Beasley, R. (1999) Self efficacy as a mediator of the relationship between pain intensity, disability and depression in chronic pain patients. Pain 80, pp.483 - 491 Ashworth, P. (1997) The variety of qualitative research. Part one: Introduction to the problem. Nurse Education Today 17, pp.215 - 218 Astley, A. (1990) A history of pain. Nursing: The Journal of Clinical Practice, Education and Management 4, 23 Aug., pp.33 - 35 Atkinson, F.I. (1996) Survey designs and sampling. In Cormack, D.F.S. Ed. The Research Process in Nursing. 3rd ed. Blackwell Science, Oxford, pp.202 - 212 Avis, M. (1994) Choice cuts: an exploratory study of patients views about participation in decision making in a day care surgery. International Journal of Nursing Studies 31(3), pp.189 - 298 Aylesbury Vale Community Health Council (1995/6) Elephants Never Forget: A Study of Outpatient Non-Attendance. Buckinghamshire Health Authority, Aylesbury Bandura, A. (1977) Self-efficacy theory: toward a unifying theory of behavioural change. Psychological Review 84 (2), pp.191 - 215 Bandura, A. (1986) Social Foundations of Thought and Action: A Social-Cognitive Theory. Prentice-Hall, New Jersey Bandura, A., O’Leary, A.. Barr Taylor, C., Gauthier, J. and Gossard, D. (1987) Perceived self-efficacy and pain control: opioid and nonopioid mechanisms. Journal of Personality and Social Psychology 53(3), pp.563 - 571 Bandura, A. (1997) Self Efficacy: The Exercise of Control. W.H. Freeman, New York Barker, P.J. (1996) Questionnaire. In Cormack, D.F.S. Ed. The Research Process in Nursing. 3rd ed. Blackwell Science, Oxford, pp.237 - 249 Barker, W. (1992) Health visiting: action research in a controlled environment. International Journal of Nursing Studies. 29(3), pp.251 - 259 Barlow, J.H., Williams, B. and Wright, C. (1996) The generalised self-efficacy scale in people with arthritis. Arthritis Care and Research 9, pp.189 - 196 Basford, L. and Slein, O. (1995) Theory and Practice of Nursing. Campion, Edinburgh Beauchamp, T.L. and Childress, J.F. (1994) Principles of Biomedical Ethics. 4th ed. Oxford University Press, Oxford Beaver, K., Luker, K.A. and Owens, R.G. (1996) Treatment decision-making in women newly diagnosed with breast cancer. Journal of Clinical Epidemiology 45(9), pp.944 - 950 Beckham, J.C., Keefe, F.J., Caldwell, D.S. and Roodman, A.A. (1991) Pain coping strategies in rheumatoid arthritis: relationships to pain, disability, depression, and daily hassles. Behavioural Therapy 22, pp.113 - 124 Behi, R. and Nolan, M. (1996) Experimental designs. British Journal of Nursing 5(12), pp.754 - 756 Beisecker, A.E. (1988) Ageing and the desire for information and input in medical decisions: patient consumerism in medical encounters. Gerontologist 28, pp.330 - 335 Bellman, L.M. (1996) Changing nursing practice through reflection on the Roper, Logan and Tierney model: the enhancement approach to action research. Journal of Advanced Nursing 24, pp.129 - 138 Benner, P. (1984) From Novice to Expert: Excellence and Power in Clinical Nursing Practice. Addison-Wesley, Menlo Park Biley, F.C. (1992) Some determinants that effect patient participation in decision making about nursing care. Journal of Advanced Nursing 17, pp.414 - 421 Biley, F.C. (1995) Encouraging patients to take part in their own care. Nursing Times 91(9), pp.42 - 43 Blacktop, J. (1996) A discussion of different sampling techniques. Nurse Researcher. 3(4), pp.5 - 15 Blomqvist, K. and Hallberg, I.R. (1999) Pain in older adults living in sheltered accommodation - agreement between assessments by older adults and staff. Journal of Clinical Nursing 8, pp.159 - 169 Boore, J. (1978) Prescription for Recovery. Royal College of Nursing, London Boud, D., Keogh, R. and Walker, D. (1985) Reflection: Turning Experience into Learning. Kogan Page, London Briggs, M. and Dean, K.L. (1998) A qualitative analysis of the nursing documentation of post-operative pain management. Journal of Clinical Nursing 7, pp.155 - 163 Brody, D.S., Miller, S.M., Lerman, C.E., Smith, D.G. and Caputo, C. (1989) Patient perception of involvement in medical care: relationship to illness attitudes and outcomes. Journal of General Internal Medicine 4, pp.506 - 511 Brookfield, S.D. (1986) Understanding and Facilitating Adult Learning. Open University Press, Milton Keynes Broome, A. (1986) Coping with pain. Strategies for relief. Nursing Times 16th Apr., pp.43 - 44 Bruster, S., Jarmean, B., Bosanquet, N., Weston, D., Erens, R. and Delbanco, T.L. (1994) National survey of hospital patients. British Medical Journal 309, pp.1542 - 1546 Buckelew, S.P., Conway, R., Shutty, M.S., Grafing, M., Anderson, S., Hewett, J. and Keefe, F.J. (1992) Spontaneous coping strategies to manage acute pain and anxiety during electrodiagnostic studies. Archives of Physical Medical Rehabilitation. 73, pp.594 - 598 Buchmann, W.F. (1997) Adherence: a matter of self-efficacy and power. Journal of Advanced Nursing 26, pp.132 - 137 Burnard, (1991) Improving through reflection. Journal of District Nursing 9(11), pp.10, 12 Burrows, D. (1990) An Evaluation of the Effectiveness of a Course of Preparation in the Use of a Pain Assessment Tool. Unpublished BSc(Hons) dissertation, Royal College of Nursing/University of Manchester Burrows, D. (1997a) Facilitation: a concept analysis. Journal of Advanced Nursing 25, pp.396 - 404 Burrows, D. (1997b) Action on pain. In Thomson, S. Ed. Nurse Teachers as Researchers: A Reflective Approach. Edward Arnold, London, Chapter 5, pp.86 - 117 Burrows, D. (1998) Using focus groups in nursing research: a personal reflection. Social Sciences in Health: The International Journal of Research and Practice. 4(1), pp.3-13 Burrows, D. and Kendall, S. (1997) Focus groups: what are they and how can they be used in nursing and health care research. Social Sciences in Health.: The International Journal of Research and Practice. 3(4), pp.244-253 Burrows, D. and Whybourne, N. (1996) The Discontinuation of Patient Controlled Analgesia Systems: An Action Research Study. Management Report, University of Luton Butler, R.W., Damarin, F.L., Beaulieu, C., Schwebel, A.I. and Thorn, B.E. (1989) Assessing cognitive coping strategies for acute post-surgical pain. Psychological Assessment 1, pp.41 - 45 Cahill, J. (1998) Patient participation - a review of the literature. Journal of Clinical Nursing 7, pp.119 - 128 Campbell, J.P. et al (1982) What to Study: Generating and Developing Research Questions. Sage, Beverly Hills Campbell, P. (1990) Mental health self-advocacy. In Winn, L. Ed. Power to the People: The Key to Responsive Services in Health and Social Care. Kings Fund Centre, London Campbell, T. (1998) Patient-focused care: primary responsibilities of research nurses. British Journal of Nursing 7(22), pp.1405 - 1409 Caress, A.L. (1997) Patient roles in decision-making. Nursing Times 93(31), pp.45 - 48 Carr, E. (1990) A no-touch technique. Nursing: The Journal of Clinical Practice, Education and Management 4(11), pp.9 - 11 Carr, E. (1997) Overcoming barriers to effective pain control. Professional Nurse 12(6), pp.412 – 416 Carr, E.C.J. and Thomas, V.J. (1997) Anticipating and experiencing post-operative pain: the patients’ perspective. Journal of Clinical Nursing 6, pp.191 - 201 Carr, W. and Kemmis, S. (1986) Becoming Critical: Education, Knowledge and Action Research. Falmer Press, London Carson, T. (1990) What kind of knowing is critical action research? Theory Into Practice xxix (3), pp.167 - 173 Carter, B. (1998) Researching pain: paradigms and revolutions. In Carter, B. Ed. Perspectives on Pain: Mapping the Territory. Arnold, London, Chapter 6 Chadderton, H. (1995) An analysis of the concept of participation within the context of health care planning. Journal of Nursing Management 3, pp.221 - 228 Chakour, M.C., Gibson, S.J., Bradbeer, M. and Helme, R.D. (1996) The effect of age on A delta- and C-fibre thermal pain stimulation. Pain 64, pp.143 - 152 Chally, P.S. (1992) Empowerment through teaching. Journal of Nursing Education 31(3), pp.117 - 120 Chalmers, I. and Altman, D.G. (1995) Systematic Reviews. BMJ Publications, London Chavasse, J.M. (1992) New dimensions of empowerment in nursing - and challenges. Journal of Advanced Nursing 17(1), pp.1 - 2 Chaves, J.F. and Brown, J.M. (1987) Spontaneous cognitive strategies for the control of clinical pain and stress. Journal of Behavioral Medicine 10(3), pp.263 - 276 Chein, I., Cook, S.W. and Harding, J. (1948) The field of action research. American Psychologist 3, pp.43 - 50 Clayton, S. (1988) Patient participation: an underdeveloped concept. Journal of the Royal Society of Health 108(2), pp.55 - 56 Closs, S.J. (1992) Surgical Patients Experiences of Sleep, Night-Time Pain and Analgesic Provision. Unpublished PhD thesis. Department of Nursing Studies, Edinburgh University Closs, S.J. and Cheater, F.M. (1996) Audit or research - what is the difference. Journal of Clinical Nursing 5, pp.249 - 256 Cohen, F. and Lazarus, R.S. (1994) Active coping processes, coping dispositions, and recovery from surgery. In Steptoe, A. and Wardle, J. Eds. Psychosocial Processes and Health: A Reader. Cambridge University Press, Cambridge, pp.348 - 368 Cohen, L. and Manion, L. (1985) Research Methods in Education. 2nd ed. Routledge, London Collier, J. (1945) United States Indian administration as a laboratory of ethnic relations. Social Research. 12, pp.275 - 276 Collins Paperback English Dictionary. (1999) 4th ed. Harper Collins, Glasgow Copp, L. (1974) The spectrum of suffering. American Journal of Nursing 74, pp.491 - 495 Coulter, A. (1997) Partnerships with patients: the pros and cons of shared clinical decision-making. Journal of Health Services Research and Policy 2(2), pp.112 - 121 Cousins, M. and Power, I. (1999) Acute and postoperative pain. In Wall, P. and Melzack, R. Eds. Textbook of Pain. 4th ed. Churchill Livingstone, Edinburgh, pp.447 - 491 Crabtree, B.F. and Miller, W.L. Eds. (1992) Doing qualitative research. Sage, London Craib, I. (1992) Modern Social Theory. From Parsons to Habermas. 2nd ed. Harvester Wheatsheaf, Hemel Hempstead Craib, I. (1992) Modern Social Theory. From Parsons to Habermas. 2nd ed. Harvester Wheatsheaf, Hemel Hempstead Craig, K.D. (1999) Emotions and psychobiology. In Wall, P. and Melzack, R. Eds. Textbook of Pain. 4th ed. Churchill Livingstone, Edinburgh, pp.331 - 343 Critical Appraisal Skills Programme (CASP) (1997) Workshop Pack. The Oxford Institute of Health Sciences, Oxford Davis, G.C. and Atwood, J.R. (1996) The development of the Pain Management Inventory for patients with arthritis. Journal of Advanced Nursing 24, pp.236 - 243 DeCrosta, T. (1984) Relieving pain: four noninvasive ways you should know more about. Nursing Life 4(2), pp.29 - 33 Department of Health (1989) Working for Patients: The Health Service Caring for the 1990s. HMSO, London Department of Health (1997) The New NHS: Modern, Dependable. HMSO, London Department of Health (1998) A First Class Service: Quality in the new NHS. HMSO, London Department of Health (1999) Making a Difference: Strengthening the Nursing, Midwifery and Health Visiting Contribution to Health and Healthcare. HMSO, London DePoy, E. and Gitlin, L.N. (1994) Introduction to Research: Multiple Strategies For Health and Human Services. Mosby, St Louis Derbyshire, S.W.G. (1996) Comment on editorial by Anand and Craig. Pain 66(1), pp.210 - 211 Descartes, R.(1644) Treatise of Man. Translated by Hall, T.S. (1972) Harvard University Press, Cambridge DeWolf, L. (1985) Focus groups: Assessing patient satisfaction and targeting new services. Hospital Topics Mar/Apr., pp.24-28 Dey, I. (1993) Qualitative Data Analysis. Routledge, London Donnan, P.T. (1996) Quantitative analysis (descriptive). In Cormack, D.F.S. Ed. The Research Process in Nursing. 3rd ed. Blackwell Science, Oxford, pp.302 - 315 Donovan, M., Dillon, P. and McGuire, L. (1987) Incidence and characteristics of pain in a sample of medical-surgical inpatients. Pain 30, pp.69 - 78 Dunbar, R. (1995) The Trouble With Science. Faber and Faber, London East, L. and Robinson, J. (1994) Change in process: bringing about change in health care through action research. Journal of Clinical Nursing 3, pp.57 - 61 Ekblom, A. and Rydh-Rinder, M. (1998) Pain mechanisms: anatomy and physiology. In Rawal, N. Ed. (1998) Management of Acute and Chronic Pain. BMJ Publications, London, Chapter 1, pp.1 - 22 Elliott, J. (1991) Action Research for Educational Change: Developing Teachers and Teaching. Open University Press, Buckingham Elliott, M. and Turrell, A. (1996) Understanding the conflicts of patient empowerment. Nursing Standard 10(45), pp.43 - 47 Endler, N.S., Parker, J.D.A. and Summerfeldt, L.J. (1993) Coping with health problems: conceptual and methodological issues. Canadian Journal of Behavioural Science 25(3), pp.384 - 399 Entwistle, V.A., Sowden, A.J. and Watt, I.S. (1998) Evaluating interventions to promote patient involvement in decision-making: by what criteria should effectiveness be judged? Journal of Health Services Research Policy 3(2), pp.100 - 107 Fay, B. (1987) Critical Social Science. Polity Press, Oxford Field, P.A. and Morse, J.M. (1985) Nursing Research: The Application Of Qualitative Approaches. Chapman and Hall, London Fielding, N. (1994) Varieties of research interviews. Nurse Researcher. 1(3), pp.4 - 13 Fish, D. (1991) But can you prove it? Quality assurance and the reflective practitioner. Assessment and Evaluation in Higher Education 16(1), pp.22 - 36 Flaherty, G.G. and Fitzpatrick, J.J. (1978) Relaxation technique to increase comfort level of postoperative patients: a preliminary study. Nursing Research. 27(6), pp.352 - 355 Flor, H. and Turk, D.C. (1988) Chronic back pain and rheumatoid arthritis: predicting pain and disability from cognitive variables. Journal of Behavioral Medicine 11, pp.251 - 265 Folkman, S. and Lazarus, R. (1980) An analysis of coping in a middle-aged community sample. Journal of Health and Social Behaviour 21, pp.219 - 239 Folkman, S. and Lazarus, R. (1985) If it changes it must be a process: study of emotion and coping during the three stages of a college examination. Journal of Personality and Social Psychology 48, pp.150 - 170 Folkman, S., Lazarus, R.S., Dunkel-Schetter, C., DeLongis, A. and Gruen, R.J. (1986) Dynamics of a stressful encounter: cognitive appraisal, coping and encounter outcomes. Journal of Personality and Social Psychology 50(5), pp.992 - 1003 Forbes, D.A., King, K.M., Kushner, K.E., Letourneau, N.L., Myrick, A.F. and Profetto-McGrath, J. (1999) Warrantable evidence in nursing science. Journal of Advanced Nursing 29(2), pp.373 - 379 Fordham, M. and Dunn, V. (1994) Alongside the Person in Pain. Balliere Tindall, London Foster, C. Ed. (1997) Manual for Research Ethics Committees. King’s College, London Foucault, M. (1988) Power/Knowledge: Selected Interviews and Other Writings 1972 - 1977. Harvester, London Friere, P. (1972) Pedagogy of the Oppressed. Translated by Bergman-Ramos, M. Penguin, Harmondsworth Fulford, K.W. (1996) Concepts of disease and the meaning of patient-centred care. In Fulford, K., Ersser, S., Hope, T. Eds. Essential Practice in Patient-Centred Care. Blackwell Science, Oxford. pp.1 - 14 Fulton, Y. (1997) Nurses’ views on empowerment: a critical social theory perspective. Journal of Advanced Nursing 26, pp.529 - 536 Getliffe, K. (1998) Developing a protocol for a randomised controlled trial: factors to consider. Nurse Researcher. 6(1), pp.5 - 17 Geuss, R. (1981) The Idea of a Critical Theory: Habermas and the Frankfurt School. Cambridge University Press, Cambridge Gibson, C. (1991) A concept analysis of empowerment. Journal of Advanced Nursing 16, pp.354 - 361 Gibson, S.J., Gorman, M.M. and Helme, R.D. (1991) Assessment of pain in the elderly. In Bond, M.R., Charlton, J.E. and Woolf, C.J. Eds. Proceedings 6th World Congress on Pain, Elsevier, Amsterdam, pp.527 - 533 Gil, K.M., Thompson, R.J., Keith, B.R., Tota-Faucette, M., Noll, S. and Kinney, T.R. (1993) Sickle cell disease pain in children and adolescents: change in pain coping strategies over time. Journal of Paediatric Psychology 18, pp.621 - 637 Gil, K.M., Wilson, J.J. and Edens, J.L. (1997) The stability of pain coping strategies in young children and adults with sickle cell disease over an 18-month period. Clinical Journal of Pain 13, pp.110 - 115 Glenister, D. (1994) Patient participation in psychiatric services: a literature review and proposal for a research strategy. Journal of Advanced Nursing 19, pp.802 - 811 Good, M. (1996) Effects of relaxation and music on postoperative pain: a review. Journal of Advanced Nursing 5, pp.905 - 914 Gould, T. (1992) Policy for controlling pain after surgery: effect of sequential changes in management. British Medical Journal 305, pp.1187 - 1193 Graffam, S. (1990) Pain content in the curriculum - a survey. Nurse Educator 15(1), pp.20 - 23 Greenwald, H.P. (1991) Interethnic differences in pain perception. Pain 44, pp.157 - 163 Greenwood, J. (1984) Nursing research: a position paper. Journal of Advanced Nursing 9, pp.77 - 82 Grenside, S. (1997) Music for Postoperative Pain and Anxiety: A Systematic Review of the Effects of Music on Postoperative Pain and Anxiety. Unpublished BA(Hons) dissertation, University of Luton Grubb, A. (1997) The Law relating to consent. In Foster, C. Ed. Manual for Research Ethics Committees. The Centre of Medical Law and Ethics, Kings College, London Guba, E.G. and Lincoln, Y.S. (1981) Effective Evaluation. Jossey-Bass, San Francisco Guba, E.G. and Lincoln, Y.S. (1998) Competing paradigms in qualitative research. In Denzin, N.K. and Lincoln, Y.S. Eds. The Landscape of Qualitative Research: Theories and Issues. Sage, Thousand Oaks Habermas, J. (1972) Knowledge and Human Interests. Heinemann, London Habermas, J. (1974) Theory and Practice. Heinemann, London Habermas, J. (1979) Communication and the Evolution of Society. Beacon Press, Boston Habermas, J. (1988) On the Logic of the Social Sciences. Translated by Nicholson, S.W. and Stark, J.A. Polity Press, Cambridge Hall-Lord, M.L., Steen, B. and Larsson, G. (1998) Postoperative Experiences of Pain and Distress in Elderly Patients - An Explorative Study. Paper presented at the Royal College of Nursing International Nursing Research Conference, England Hallett, C. (1997) The use of descriptive statistics in nursing research. Nurse Researcher. 4(4), pp.4 - 16 Halsey, A.H. Ed. (1972) Educational Priority: Volume 1 Problems and Policies. HMSO, London Harden, J. (1996) Enlightenment, empowerment and emancipation: the case for critical pedagogy in nurse education. Nurse Education Today 16, pp.32 - 37 Hart, E. and Bond, M. (1995a) Developing action research in nursing. Nurse Researcher 2(3) Mar, pp.4 - 14 Hart, E. and Bond, M. (1995b) Action Research for Health and Social Care: A Guide to Practice. Open University Press, Buckingham Harvey, L. (1990) Critical Social Research. Unwin, London Hawks, J.H. (1992) Empowerment in nursing education: concept analysis and application to philosophy, learning and instruction. Journal of Advanced Nursing 17, pp.609 - 618 Hayward, J. (1973) Information - A Prescription Against Pain. Royal College of Nursing, London Heinneman, E.A., Lee, J.L. and Cohen, J.I. (1995) Collaboration: a concept analysis. Journal of Advanced Nursing 21, pp.103 - 109 Heron, J. (1977) Dimensions of Facilitator Style. University of Surrey, Guildford Heron, J. (1981) Philosophical basis for a new paradigm. In Reason, P. and Rowan, J. Eds. Human Inquiry: A Sourcebook of New Paradigm Research. John Wiley and Sons, Chichester Heron, J. (1996) Co-operative Inquiry. Research into the Human Condition. Sage, London Hewison, A. (1995) Nurses’ power in interaction with patients. Journal of Advanced Nursing 21, pp.75 - 82 Heye, M.L. (1991) Pain and discomfort after coronary artery bypass surgery. Cardiovascular Nursing 27(4), pp.19 - 24 Hillier, V. and Gibbs, A. (1996) Calculations to determine sample size. Nurse Researcher. 3(4), pp.27 - 34 Holter, I.M. and Schwartz- Barcott, D. (1993) Action research: what is it? How has it been used and how can it be used in nursing? Journal of Advanced Nursing 18, pp.298 - 304 Hughes, T. and Foster, C. (1997) Communicating with the potential research subject. In Foster, C. Ed. Manual for Research Ethics Committees. The Centre of Medical Law and Ethics, Kings College, London International Association for the Study of Pain Subcommittee on Taxonomy (1979) Pain terms: a list with definitions and notes on usage. Pain 6, pp.249 - 252 Jadad, A.R., Moore, R.A., Carroll, C., Jenkinson, C., Reynolds, D.J.M. and Gavaghan, D.J. (1996) Assessing the quality of reports of randomised controlled clinical trials: is blinding necessary? Controlled Clinical Trials 17, pp.1 - 12 Jenkins, R. (1997) Issues of empowerment for nurses and clients. Nursing Standard 11(46), pp.44 - 46 Jensen, M.P., Karoly, P. and Braver, S. (1986) The measurement of clinical pain intensity: a comparison of six methods. Pain. 27, pp.117 - 126 Jenson, M.P., Turner, J.A. and Romano, J.M. (1991b) Self-efficacy and outcome expectancies: relationship to chronic pain coping strategies and adjustment. Pain 44, pp.263 - 269 Jenson, M.P., Turner, J.A., Romano, J.M. and Karoly, P. (1991a) Coping with chronic pain: a critical review of the literature. Pain 47, pp.431 - 438 Johns, C. (1994) Centre for Reflective Practice: Model of Structured Reflection. 10th ed. Faculty of Health Care and Social Studies, University of Luton Jones, S. (1996) An action research investigation into the feasibility of experienced registered sick children’s nurses becoming children’s emergency nurse practitioners. Journal of Clinical Nursing 5, pp.13 - 21 Kahn, D. and Steeves, R. (1988) Caring and practice: construction of the nurses’ world. Scholarly Inquiry for Nursing Practice 2, pp.201 - 216 Kashikar-Zuck, S., Keefe, F.J., Kornguth, P., Beaupre, P., Holzberg, A. and Delong, D. (1997) Pain coping and the pain experience during mammography: a preliminary study. Pain 73(2), pp.165 - 172 Keefe, F.J., Caldwell, D.S., Baucom, D., Salley, A., Robinson, E., Timmons, K., Beaupre, P. Weisberg, J. and Helms, M. (1996) Spouse assisted coping skills training in the management of osteoarthritis knee pain. Arthritis Care and Research 9, pp.276 - 291 Keefe, F.J., Kashikar-Zuck, S., Robinson, E., Salley, A., Beaupre, P., Caldwell, D., Baucom, D. and Haythornthwaite, J. (1997a) Pain coping strategies that predict patients’ and spouses’ ratings of patients’ self-efficacy. Pain 73(2), pp.191 - 199 Keefe, F.J. and Lefebvre, J.C. (1999) Behavioral therapy. In Wall, P. and Melzack, R. Eds. Textbook of Pain. 4th ed. Churchill Livingstone, Edinburgh, pp.1445 – 1461 Keefe, F.J., Lefebvre, J.C., Maixner, W., Salley Jr., A.N. and Caldwell, D.S. (1997b) Self-efficacy for arthritis pain: relationship to perception of thermal laboratory pain stimuli. Arthritis Care Research 10, pp.177 - 184 Keefe, F.J., Salley, A.N. and Lefebvre, J.C. (1992) Coping with pain: conceptual concerns and future directions. Pain 51, pp.131 - 134 Kendall, S. and Sturt, J. (1996) Negotiating access into primary health care using action research: insights from critical theory. Social Sciences in Health 2(2), pp.107 - 120 Kim, H.S. (1993) Putting theory into practice: problems and prospects. Journal of Advanced Nursing 18, pp.1632 - 1639 Kim, H.S., Holler, I.M., Leovenson, M.L., Inayoshi, M., Shimaguchi, S., Ryder, R.S., Kawaguchi, Y., Hovi, R., Takezaki, K., Kilpi, H.L. and Ununen, M.M. (1993) Patient nurse collaboration. A comparison of patients and nurses attitudes in Finland, Japan, Norway and the USA. International Journal of Nursing Studies 30(5), pp.387 - 401 Kincheloe, J.L. and McLaren, P.L. (1998) Rethinking critical theory and qualitative research. In Denzin, N.K. and Lincoln, Y.S. Eds. The Landscape of Qualitative Research: Theories and Issues. Sage, Thousand Oaks Kitson, A. (1996) Perspectives on patient-centred care: a nurse’s perspective. In Fulford, K., Ersser, S., Hope, T. Eds. Essential Practice in Patient-Centred Care. Blackwell Science, Oxford. pp.x - xi Kitzinger, J. (1994) The methodology of focus groups: the importance of interaction between research participants. Sociology of Health and Illness 16(1), pp.103-121 Knodel, J. (1993) The design and analysis of focus group studies: A practical approach. In Morgan D.L. Ed. Successful Focus Groups: Advancing the State of the Art. Sage, London Koch, T. (1996) Implementation of a hermeneutic inquiry in nursing: philosophy, rigour and representation. Journal of Advanced Nursing 24, pp.174 - 184 Krueger, R.A. (1994) Focus Groups: A Practical Guide for Applied Research 2nd ed. Sage, Thousand Oaks, California Kuhn, T.S. (1962) The Structure of Scientific Revolutions. University of Chicago Press, Chicago Kuokkanen, L. and Leino-Kilpi, H. (2000) Power and empowerment in nursing: three theoretical approaches. Journal of Advanced Nursing 31(1), pp.235 - 241 Lankshear, A.J. (1993) The use of focus groups in a study of attitudes to student nurse assessment. Journal of Advanced Nursing 18, pp.1986-1989 Large, R. and Strong, J. (1997) The personal constructs of coping with chronic low back pain: is coping a necessary evil? Pain 73(2), pp.245 - 252 Lather, P. (1986) Issues of validity in openly ideological research: between a rock and a soft place. Interchange 17 (4), 63-84 Lathlean, J. (1996) Practice-Based Research: A Case for Action Research? Paper presented at the national conference: A vision for the future: facing the research challenge. Missenden Abbey, Buckinghamshire Lazarus, R.S. and Folkman, S. (1984) Stress, Appraisal and Coping. Springer, New York le May, A. (1999) Evidence-Based Practice. Nursing Times Clinical Monographs No 1. NT Books, Guildford Lester, N. and Keefe, F.J. (1998) Coping with chronic pain. In Baum, A., McManus, C., Newman, S., Weinman, J. and West, R. Eds. Cambridge Handbook of Psychology, Health and Medicine. Cambridge University Press, Cambridge Lewin, G.W. Ed. (1948) Resolving Social Conflicts: Selected Papers on Group Dynamics by Kurt Lewin. Harper and Brothers, New York Lewin, K. (1946) Action research and minority problems. Journal of Social Issues 2, pp.34 - 46 Lloyd, L. and McLauchlan, A. (1994) Nurses' attitudes towards management of pain. Nursing Times 90, 26 Oct., pp.40-43 Macleod Clark, J. and Latter, S. (1990) Working together. Nursing Times 86(48), pp.28 - 31 Madjar, I. (1985) Pain and the surgical patient: a cross-cultural perspective. The Australian Journal of Advanced Nursing 2(2), pp.29 - 33 Mahler, H.I. and Kulik, J. (1990) Preferences for health care involvement, perceived control and surgical recovery. A prospective study. Social Science Medicine 31, pp.743 - 751 Malin, N. and Teasdale, K. (1991) Caring versus empowerment: considerations for nursing practice. Journal of Advanced Nursing 16, pp.657 - 662 Manne, S.L. and Zautra, A.J. (1992) Coping with arthritis: current status and critique. Arthritis and Rheumatology 35, pp.1273 - 1280 Marteau, T.M. and Bekker, H. (1992) The development of a six-item short-form of the state scale of the Spielberger State-Trait Anxiety Inventory (STAI). British Journal of Clinical Psychology 31, pp.301-306 May, C. (1995) Patient autonomy and the politics of professional relationships. Journal of Advanced Nursing 21, pp.83 - 87 McCaffery, M. (1968) Nursing Practice Theories Related to Cognition, Bodily Pain, and Man-Environment Interactions. University of California, Los Angeles McCaffery, M. (1981) Don’t just do something: sit there. Nursing81 June, pp.58 - 61 McCaffery, M. (1999) Pain: Clinical Manual. 2nd ed. Mosby, St Louis McCaffery, M. and Beebe, A. (1994) Pain: Clinical Manual for Nursing Practice. UK ed. Mosby, London McCaul, K.D. and Malott, J.M. (1984) Distraction and coping with pain. Psychological Bulletin 95, pp.516 - 533 McCracken, L.M. (1991) Cognitive-behavioural treatment of rheumatoid arthritis: a preliminary review of efficacy and methodology. Annals of Behavioral Medicine 13, pp.57 - 65 McDonald, D.D. (1994) Gender and ethnic stereotyping and narcotic analgesic administration. Research in Nursing and Health 17, pp.45 - 49 McGuire, D. (1984) The measurement of clinical pain. Nursing Research. 33, pp.152 - 156 McGuire, D. (1992) Comprehensive and multidimensional assessment and measurement of pain. Journal of Pain and Symptom Management 7, pp.312 - 319 McNiff, J. (1988) Action Research: Principles and Practice. Macmillan Education, London McQuay, H. and Moore, A. (1998) An Evidence-Based Resource for Pain Relief. Oxford University Press, Oxford Melville, M. (1997) Consumerism: do patients have power in health care? British Journal of Nursing. 6(6), pp.337 - 339 Melzack, R. (1973) The Puzzle of Pain. Penguin, London Melzack, R. and Casey, K.L. (1968) Sensory, motivational and central control determinants of pain: a new conceptual model. In Kenshalo, D. Ed. The Skin Senses. Thomas, Springfield, pp.423 - 443 Melzack, R. and Katz, (1999) Pain measurements in persons in pain In Wall, P. and Melzack, R. Eds. Textbook of Pain. 4th ed. Churchill Livingstone, Edinburgh, pp.409 - 426 Melzack, R. and Wall, P.D. (1965) Pain mechanisms: a new theory. Science 150, pp. 971 - 979 Melzack, R. and Wall, P.D. (1996) The Challenge of Pain. 3rd ed. Penguin, London Merskey, H. (1980) Some features of the history of pain. Pain 9, pp.3 - 8 Merskey, H. (1996) Response to editorial: new perspectives on the definition of pain. Pain 66(1), pp.209 - 229 Meyer, J. (1993) New paradigm research in practice: the trials and tribulations of action research. Journal of Advanced Nursing 18, pp.1066 - 1072 Meyer, J. (1995a) Introducing Lay Participation in Care. Unpublished PhD thesis. Kings College, University of London Meyer, J. (1995b) Stages in the process: a personal account. Nurse Researcher 2(3) Mar, pp.24 - 37 Meyer, J., Spilsbury, K. and Prieto, J. (2000) Comparison of findings from a single case in relation to those from a systematic review of action research. Nurse Researcher 7(2) Winter, pp.37 - 57 Miller, C., Ross, N. and Freeman, M. (1999) Shared Learning and Clinical Teamwork: New Directions in Education for Multiprofessional Practice. ENB, London Miro, J. and Raich, R.M. (1999) Effects of a brief and economical intervention in preparing patients for surgery: does coping style matter? Pain 83, pp.471 - 475 Mobily, P.R., Herr, K.A. and Kelley, L.S. (1994) Cognitive-behavioral techniques to reduce pain: a validation study. International Journal of Nursing Studies 30(6), pp.537 - 548 Mogan, J., Wells, N. and Robertson, E. (1985) Effects of preoperative teaching on postoperative pain: a replication and expansion. International Journal of Nursing Studies 22, pp.267 - 280 Monat, A. and Lazarus, R. Eds. (1991) Stress and Coping: An Anthology. 3rd ed. Columbia University Press, New York Montes-Sandoval, L. (1999) An analysis of the concept of pain. Journal of Advanced Nursing 29(4), pp.935 - 941 Morrison, R., Jones, L. and Fuller, B. (1997) The relation between leadership style and empowerment on job satisfaction of nurses. Journal of Nursing Administration 2, pp.27 - 34 Muir Gray, J. (1997) Evidence-Based Healthcare: How to Make Policy and Management Decisions. Churchill Livingstone, Edinburgh Munro, B.H. and Page, E.B. (1993) Statistical Methods for Health Care Research. 2nd ed. Lippincott, Philadelphia Nagy, S. (1999) Strategies used by burns nurses to cope with the infliction of pain on patients. Journal of Advanced Nursing 29, pp.1427 - 1433 National Health Service Executive (1996) Promoting Clinical Effectiveness. A Framework for Action In and Through the NHS. NHSE, Leeds National Health Service Executive (2000) The National Plan for the NHS. April 20th National Health Service Management Executive (1993) Patient Empowerment. NHSME, Leeds National Institute of Nursing Research (1994) Symptom Management: Acute Pain: A Report of the NINR Priority Expert Panel. U.S. Department of Health and Human Services, Bethesda Nethercott, S. (1993) A concept for all the family, family-centred care: a concept analysis. Professional Nurse 8(12), pp.794 - 797 Newton, C.A. (1995) Action research: application in practice. Nurse Researcher 2(3) Mar, pp.60 - 71 National Health Service Management Executive (1993) Patient Empowerment. NHSME, Leeds Numbers, L. (1986) Pain: an introduction. Nursing: The Add-On Journal of Clinical Nursing 3(10), pp.358 - 359 O’Hanlon, C. (1994) Reflection and action in research: is there a moral responsibility to act? Educational Action Research 2(2), pp.281 - 289 Oldham, J. (1994) Experimental & quasi-experimental research designs. Nurse Researcher 1(4), pp.26 - 36 Paley, J. (1996) How not to clarify concepts in nursing. Journal of Advanced Nursing 24, pp.572 - 578 Parker, J.C., Smarr, K.L., Buescher, K.L., Phillips, L.R., Franck, R.G., Beck, N.C., Anderson, S.K. and Walker, S.E. (1989) Pain control and rational thinking: implications for rheumatoid arthritis. Arthritis Rheumatology 32, pp.984 - 990 Parsell, G., Spalding, R. and Bligh, J. (1998) Shared goals, shared learning: evaluation of a multiprofessional course for undergraduate students. Medical Education 32, pp.304 - 311 Peters, M. and Robinson, V. (1984) The origins and status of action research. Journal of Applied Behavioral Science 20(2), pp.113 - 124 Pick, B., Pearce, S. and Legg, C. (1990) Cognitive responses and the control of postoperative pain. British Journal of Clinical Psychology 29, pp.409-415 Pill, R., Rees, M.E., Stott, N.C.H. and Rollnick, S.R. (1999) Can nurses learn to let go? Issues arising from an intervention designed to improve patients’ involvement in their own care. Journal of Advanced Nursing 29(6), pp.1492 - 1499 Piper, S.M. and Brown, P.A. (1998) Psychology as a theoretical foundation for health education in nursing: empowerment or social control? Nurse Education Today 18, pp.637 - 641 Plato Dialogues, 2 Gorgias, 2 Timaeus, 4 Philebus. Translated by Jowett, B. (1871) Oxford University Press, London Playle, J.F. (1995) Humanism and positivism in nursing: contradictions and conflicts. Journal of Advanced Nursing 22, 979 - 984 Polit, D.F. and Hungler, B.P. (1999) Nursing Research: Principles and Methods. 6th ed. Lippincott, Philadelphia Poole, K. and Jones, A. (1996) A re-examination of the experimental design for nursing research. Journal of Advanced Nursing 24, pp.108 - 114 Porter, S. (1996) Qualitative research. In Cormack, D.F.S. Ed. The Research Process in Nursing. 3rd ed. Blackwell Science, Oxford, pp.113 - 122 Presly, A.S. (1996) Common terms an concepts in nursing research. In Cormack, D.F.S. Ed. The Research Process in Nursing. 3rd ed. Blackwell Science, Oxford, pp.14 - 21 Price, D.D. (1988) Psychological and Neural Mechanisms of Pain. Raven Press, New York Price, D.D., Mao, J. and Meyer, D.J. (1994) Central neural mechanisms of normal and abnormal pain states. In Fields, H.L. and Liebeskind, J.C. Eds. Pharmacological Approaches to the Treatment of Chronic Pain: New Concepts and Critical Issues. Raven Press, New York, pp.61 - 84 Rapoport, R.N. (1970) Three dilemmas of action research. Human Relations 23, pp.499 - 513 Rappaport, J. (1981) In praise of paradox: a social policy of empowerment over prevention. American Journal of Community Psychology 9(1), pp.1 - 25 Rawal, N. Ed. (1998) Management of Acute and Chronic Pain. BMJ Publications, London Reason, P. and Rowan, J. Eds. (1981) Human Inquiry: A Sourcebook of New Paradigm Research. John Wiley & Sons, Chichester Reason, P. Ed. (1988) Human Inquiry in Action: Developments in New Paradigm Research. Sage, London Reiskin, H. (1992) Focus groups: A useful technique for research and practice in nursing. Applied Nursing Research 5(4), pp.197-201 Revenson, T.A. and Majerovitz, S.D. (1990) The effects of illness on the spouse: social resources as stress buffers. Arthritis Care and Research 4, pp.63 - 72 Richardson, J. and Edwards, M. (1997) An undergraduate clinical skills laboratory: developing interprofessional skills in physical and occupational therapy. Gerontology and Geriatrics 17(4), pp.33 - 43 Rodin, J. and Janis, I. (1979) The social power of health-care practitioners as agents of change. Journal of Social Issues 35, pp.60 - 81 Rodwell, C.M. (1996) An analysis of the concept of empowerment. Journal of Advanced Nursing 23, pp.305 - 313 Roe, B. (1994) Is there a place for the experiment in nursing research? Nurse Researcher 1(4), pp.4 - 12 Rogers, C. (1979) Carl Rogers on Personal Power. Delacorte, New York Rogers, C. (1983) Freedom to Learn for the 80’s. Bell and Howell, Columbus Rokke, P.D. and al’Absi, M. (1992) Matching pain coping strategies to the individual: a prospective validation of the cognitive coping strategy inventory. Journal of Behavioral Medicine 15(6), pp. 611 - 625 Roper, N., Logan, W. and Tierney, A. (1990) The Elements of Nursing. 3rd ed. Churchill Livingstone, Edinburgh Rosential, A.K. and Keefe, F.J. (1983) The use of coping strategies in chronic low back pain patients: relationship to patient characteristics and current adjustment. Pain 17, pp.33 - 44 Royal College of Surgeons of England and the College of Anaesthetists (1990) Report of the Working Party on Pain After Surgery. Royal College of Surgeons of England and the College of Anaesthetists, London Ryles, S.M. (1999) A concept analysis of empowerment: its relationship to mental health nursing. Journal of Advanced Nursing 29(3), pp.600 - 607 Sackett, D., Richardson, W.S., Rosenberg, W. and Haynes, B. (1996) Evidence-Based Medicine. Churchill Livingstone, London Salmon, P. and Manyande, A. (1996) Good patients cope with their pain: postoperative analgesia and nurses’ perceptions of their patients’ pain. Pain 68, pp.63 68 Sandelowski, M. (1986) The problem of rigor in qualitative research. Advances in Nursing Science 8 (3), pp.27 - 37 Sandelowski, M. (1990) Rigor or rigor mortis: the problem of rigor in qualitative research revisited. Advanced Nursing Science 16 (2), pp.1 - 8 Sanford, N. (1981) A model for action research. In Reason, P. and Rowan, J. Eds. Human Inquiry: A Sourcebook of New Paradigm Research. John Wiley and Sons, Chichester pp.173 - 181 Sarason, I.G., Levine, H.M. and Basham, R.B. (1983) Assessing social support: the Social Support Questionnaire. Journal of Personality and Social Psychology 44, pp.127 - 139 Saunders, P. (1995) Encouraging patients to take part in their own care. Nursing Times 91(9), pp.42 - 43 Schanberg, L.E., Lefebvre, J.C., Keefe, F.J., Kredich, D.W. and Gil, K.M. (1997) Pain coping and the experience in children with juvenile chronic arthritis. Pain 73(2), pp.181 - 189 Schon, D. (1991) The Reflective Practitioner: How Professionals Think and Act. Avebury, Aldershot Schroeder, C. and Neil, R.M. (1992) Focus groups: A humanistic means of evaluating an HIV/AIDS programme based on caring theory. Journal of Clinical Nursing 1, pp.265-274 Schwengkmezger, P., Lang, J. and Landsiedel, W. (1990) Coping skills and perceived threat: an examination in haemodialysis patients. In Schmidt, L.R., Schwengkmezger, P., Weinman, J. and Maes, S. Eds. Theoretical and Applied Aspects of Health Psychology. Harwood Academic, London pp. 413 - 421 Scientific Software Development (1998) Atlas.ti Version 4. Scientific Software Development, Berlin Seers, K. (1987) Pain, Anxiety and Recovery in Patients Undergoing Surgery. Unpublished PhD thesis, University of London Seers, K. (1994) Foreword. In Fordham, M. and Dunn, V. Alongside the Person in Pain. Balliere Tindall, London, pp.vii - ix Seers, K. and Carroll, D. (1998) Relaxation techniques for acute pain management: a systematic review. Journal of Advanced Nursing 27, pp.466 - 475 Serpell, M.G., Makin, A. and Harvey, A. (1998) Acute pain physiology and pharmacological targets: the present and future. Acute Pain 1(3), pp.31 - 47 Shaw, C. (1999) A framework for the study of coping, illness behaviour and outcomes. Journal of Advanced Nursing 29(5), pp.1246 - 1255 Silverman, D. (1993) Interpreting Qualitative Data. Sage, London Simmons, S. (1995) From paradigm to method in interpretative action research. Journal of Advanced Nursing 21, pp.837 - 844 Sindhu, F. (1996) Are non-pharmacological nursing interventions for the management of pain effective? - A meta-analysis. Journal of Advanced Nursing 24, pp.1152 - 1159 Skelton, R. (1994) Nursing and empowerment: concepts and strategies. Journal of Advanced Nursing 19, pp.415 - 423 SPSS Incorporated (1994) Statistical Package for Windows. Release 6.1, SPSS Incorporated Sofaer, B. (1984) The Effect of Focused Education for Nursing Teams on the Postoperative Pain of Patients. Unpublished PhD thesis. University of Edinburgh, Edinburgh Sofaer, B. (1998) Pain: Principles, Practice and Patients. 3rd ed. Stanley Thornes, Cheltenham Sofaer, B. and Walker, J. (1994) Mood assessment in chronic pain patients. Disability and Rehabilitation 16(1), pp.35 - 38 Spradley, J.P. (1980) Participant Observation. Holt, Rinehart and Winston, New York Spreitzer, G. (1996) Social structural characteristics of psychological empowerment. Academy of Management Journal 39, pp.483 - 504 Sternbach, R. (1968) Pain: A psychophysiological analysis. Academic Press, New York Stevens, M.J., Pfost, K.S. and Rapp, B.J. (1987) Modifying acute pain by matching cognitive style with cognitive treatment. Perceptual and Motor Skills 65, pp.919 - 924 Stevens, P.E. (1989) A critical social reconceptualization of environment in nursing: implications for methodology. Advances in Nursing Science 11(4), pp.56 - 68 Stevenson, C. (1995) Non-pharmacological aspects of acute pain management. Complementary Therapies in Nursing and Midwifery 1, pp.77 - 84 Stower, S. (1992) Partnership caring. Journal of Clinical Nursing 1, pp.67 - 72 Street, A.F. (1992) Inside Nursing: a Critical Ethnography of Clinical Nursing Practice. State University of New York Press, Albany Strong, J. (1995) Self-efficacy and the patient with chronic pain. In Shacklock, M. Ed. Moving Into Pain. pp.97 - 102 Sturt, J. (1997) The Implementation of Self-Efficacy Theory into Health Promotion Practice in Primary Health Care: An Action Research Approach. Unpublished PhD thesis. Faculty of Health Studies, Brunel University, Middlesex Susman, G.I. and Evered, R.O. (1978) An assessment of the scientific merits of action research. Administrative Science Quarterly 23, pp.582 - 601 Taal, E., Seydel, E.R., Rasker, J.J. and Wiegman, O. (1993) health status, adherence with recommendations, self-efficacy and social support in patients with rheumatoid arthritis. Patient Education Counselling 20, pp.63 - 76 Ter Kuile, M.M., Spinhoven, P., Linssen, A.C.G. and Houwelingen, C. (1995) Cognitive coping and appraisal in the treatment of chronic headaches. Pain 64, pp.257 - 264 Thompson, S.G., Pitts, J.S. and Schwantovsky, L. (1992) Preferences for involvement in medical decision making: situational and demographic influence. Patient Education and Counselling 22, pp.133 - 140 Titchen, A. (1993) Action research as a research strategy: finding our way through a philosophical and methodological maze. In Titchen, A. Ed. Changing Nursing Practice Through Action Research. Report No 6. National Institute for Nursing, Oxford pp.49 - 58 Titchen, A (1994) Roles and relationships in collaborative research. Surgical Nurse 7(5) Oct., pp.15 - 19 Titchen, A. (1995) Issues of validity in action research. Nurse Researcher 2(3) Mar, pp.38 - 48 Titchen, A. (1996) A case study of a patient-centred nurse. In Fulford, K., Ersser, S., Hope, T. Eds. Essential Practice in Patient-Centred Care. Blackwell Science, Oxford. pp.182 - 197 Titchen, A. and Binnie, A. (1993a) Research partnerships: collaborative action research in nursing. Journal of Advanced Nursing 18, pp.858 - 865 Titchen, A and Binnie, A. (1993b) A “double-act”: co-action researcher roles in an acute hospital setting. In Titchen, A. (Ed.) Changing Nursing Practice Through Action Research. National Institute for Nursing. Report No 6, Oxford, pp.19 - 28 Torbert, W.R. (1981) Why educational research has been so uneducational. In Reason, P. and Rowan, J. Eds. Human Inquiry: A Sourcebook of New Paradigm Research. John Wiley & Sons, Chichester pp.141 - 151 Townsend, J. (1990) Teaching/learning strategies. Nursing Times 86(23), pp.66 - 68 Trnobranski, P.H. (1994) Nurse-patient negotiation: assumption or reality? Journal of Advanced Nursing 19, pp.733 - 737 Turk, D.C. and Okifuji, A. (1999) A cognitive-behavioral approach to pain management. In Wall, P. and Melzack, R. Eds. Textbook of Pain. 4th ed. Churchill Livingstone, Edinburgh, pp.1431 - 1443 Turner, J.A. and Keefe, F.J. (1999) Cognitive-behavioral therapy for pain. In Max, M. Ed Pain 1999 - An Updated Review. IASP Press, Seattle pp.523 - 533 United Kingdom Central Council for Nursing, Midwifery and Health Visiting (1992) Code of Professional Conduct. 3rd ed. UKCC, London United Kingdom Central Council for Nursing, Midwifery and Health Visiting (1996) Guidelines for Professional Practice. UKCC, London Wade, S. (1995) Partnership in care: a critical review. Nursing Standard 9(48), pp.29 - 32 Wagner, E.H., Barrett, P., Barry, M.J., Barlow, W. and Fowler, F.J. (1995) The effect of a shared decision-making program on rates of surgery for benign prostatic hyperplasia: pilot results. Medical Care 33, pp.765 - 770 Walding, M. (1991) Pain, anxiety and powerlessness. Journal of Advanced Nursing 16, pp.388 - 397 Walker, J. and Campbell, S. (1988) Pain assessment and the nursing process. Senior Nurse 88(5), pp.28 - 31 Wall, P. and Melzack, R. Eds. (1999) Textbook of Pain. 4th ed. Churchill Livingstone, Edinburgh Waterhouse, M. (1996) Why pain assessment must start with believing the patient. Nursing Times 92(38), pp.42 - 43 Waterman, H. (1995) Distinguishing between ‘traditional’ and action research. Nurse Researcher 2(3) Mar, pp.15 - 23 Waterworth, S. and Luker, K. (1990) Reluctant collaborators: do patients want to be involved in decisions concerning care. Journal of Advanced Nursing 15, pp.971 - 976 Webb, C. (1989) Action research: philosophy, methods and personal experiences. Journal of Advanced Nursing 14, pp.403 - 410 Webb, C. (1990) Partners in research. Nursing Times 86(32) pp. 40 - 44 Weis, O.F., Sriwatanakul, K., Alloza, J.L., Weintraub, M. and Lasagna, L. (1983) Attitudes of patients, housestaff and nurses towards postoperative analgesic care. Anaesthesia and Analgesia 62, pp.72 - 74 Weisenberg, M. (1999) Cognitive aspects of pain. In Wall, P. and Melzack, R. Eds. Textbook of Pain. 4th ed. Churchill Livingstone, Edinburgh, pp.345 - 358 Wells, N. (1987) Cognitive behavioural strategies in pain management. JRSH 3, pp.92 - 95 Wilder-Smith, C. and Schuler, L. (1992) Postoperative analgesia: pain by choice? The influence of patient attitudes and patient education. Pain 50, pp.257 - 262 Wilkinson, J. (1999) Understanding patients’ health beliefs. Professional Nurse 15(5), pp. 320 - 322 Wilkinson, R. (1996) A non-pharmacological approach to pain relief. Professional Nurse 11(4), pp.222 - 224 Williams, A. (1995) Pain measurement in chronic pain management. Pain Reviews 2, pp.39 - 63 Williams, A. CdeC., Nicholas, M. and Richardson, P. (1993) Evaluation of a cognitive behavioural programme for rehabilitating patients with chronic pain. British Journal of General Practice 43, pp.513 - 518 Williams, D.A., Robinson, M.E. and Geisser, M.E. (1994) Pain beliefs, assessment and utility. Pain 59, pp.71 - 78 Williams, S.L. and Kinney, P.J. (1991) Performance and non-performance strategies for coping with acute pain: the role of perceived self-efficacy, expected outcomes and attention. Cognitive Therapy and Research 15(1), pp.1 - 19 Winter, R. (1989) Learning From Experience: Principles and Practice in Action Research. Falmer Press, London Winter, R. (1993) Action research, practice and theory. Educational Action Research 1(2), pp.315 - 316 Wolf, A.M.D., Nasser, J.F., Wolf, A.M. and Schorling, J.B. (1996) The impact of informed consent on patient interest in prostate-specific antigen screening. Archives of Internal Medicine 156, pp.1333 - 1336 18th World Medical Association (1964) Declaration of Helsinki. World Medical Association, Helsinki 29th World Medical Association (1975) Declaration of Helsinki. World Medical Association, Tokyo 35th World Medical Association (1983) Declaration of Helsinki. World Medical Association, Venice 41st World Medical Association (1989) Declaration of Helsinki. World Medical Association, Honk Kong 48th World Medical Association (1996) Declaration of Helsinki. World Medical Association, Somerset West, Republic of South Africa Worthington, E.L. (1978) The effects of imagery content, choice of imagery content and self-verbalization of the self-control of pain. Cognitive Therapeutic Research 2, pp.225 - 240 Zautra, A. and Manne, S. (1992) Coping with rheumatoid arthritis: a review of a decade of research. Annals of Behavioural Medicine 14, pp.31 - 40 322