Complexities and Comorbidities in Patients with Chronic Non-Cancer Pain: Suicidal Behaviours and Pharmaceutical Opioid Use

Complexities and comorbidities in patients with chronic non-cancer pain: Suicidal behaviours and pharmaceutical opioid use

Gabrielle Campbell Masters of Criminology

A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy

National Drug and Alcohol Research Centre School of Public Health and Community Medicine Faculty of Medicine University of New South Wales

January 2016

THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Campbell (Szytkiel)

First name: Gabrielle Other name/s:

Abbreviation for degree as given in the University calendar: PhD

School: Medicine Faculty: School of Public Health and Community Medicine

Title: Complexities and comorbidities in patients with chronic non-cancer pain: Suicidal behaviours and pharmaceutical opioid use

Abstract: Chronic non-cancer pain (CNCP) is a worldwide, common complaint and has a major impact on the individual and community in terms of social, financial and health costs. To date, research has been limited in a number of areas, including the association of CNCP with suicidality, the nature of pharmaceutical opioid use in people with CNCP and the prevalence of Borderline Personality Disorder (BPD) in people with CNCP, and associations with suicidality and problematic opioid use.

This thesis aimed to address the gaps in previous literature using two data sources; firstly, the 2007 Australian National SuNey of Mental Health and Well-being and secondly, data from a large, na tional community suNey, the Pain and Opioids IN Treatment (POINT) study. The studies included in this thesis aimed to: 1) examine the association of CNCP with suicidality in the Australia population; 2) describe the formation of the POINT cohort to examine prescription opioid use in CNCP; 3) examine the characteristics of CNCP patients prescribed pharmaceutical opioids; 4) examine characteristics associated with pharmaceutical opioid dose and pharmaceutical opioid dependence in the POINT cohort; 5) determine the prevalence of suicidality and contribution of general and pain-specific risk factors to suicidality in the POINT cohort and 6) explore BPD in the POINT cohort and associations with problematic pharmaceutical opioid use and suicidality.

There were a number of significant findings from th is thesis. Firstly, there was large proportion of the cohort consuming opioids in the high range (over 90mg of oral morphine equivalents per day) and this was associated with more problematic opioid use. Secondly, suicidality was common in people with CNCP, especially in people prescribed opioids for CNCP. An individual's confidence in coping with their pain was the only independent factor that was predictive of past 12 month elevation from ideation to suicide attempt. Finally, the presence of BPD was high amongst this sample and was associated with pharmaceutical opioid dependence and suicidality.

Clinical and research implications are discussed in terms of the need for a multidisciplinary approach to CNCP and the need for appropriate screening for the use of pharmaceutical opioids in the treatment of CNCP.

Declaration relating to disposition of project thesis/dissertation

1 hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media. now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the rig ht to use in future works (such as articles or books) all or part of this thesis or dissertation.

1 also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only). I Signature · · ··· · ···· · ··· · · · · ~W itness · · ·· · Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and re uire the a roval of the Dean of Graduate Research .

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS DECLARATIONS

Originality statement

I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.

Gabrielle Campbell January 2016

ii Copyright

I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the

Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only).

I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted

I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.

Gabrielle Campbell January 2016

Authenticity statement

I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.

Gabrielle Campbell January 2016

iii

ACKNOWLEDGMENTS First and foremost I would like to give my sincerest thanks to my supervisors, Louisa Degenhardt (supervisor), Shane Darke (joint-supervisor) and Raimondo Bruno (co- supervisor). I feel honoured and privileged to have been supervised by individuals who are so highly regarded in their respective fields. To Louisa, you are inspiring and your passion is what led me to appreciate research in the first place. Your generosity and support (and pushing) throughout the years have been greatly appreciated, and I am eternally grateful for all the opportunities you share with me. To Shane, thank you for the support you have given me over the years we have worked together. Your door has always been open whenever I have needed to talk. Through you; I have learnt to appreciate the English language. To Raimondo, I have immensely enjoyed working with you. You kept me grounded throughout my thesis, and your passion for statistics has inspired me. Finally, to my unofficial mentor Briony Larance. Thank you for your support and wisdom. You are always there for a chat (whether you want to or not), you make me laugh, and that has kept me going throughout the years.

Thank you to the POINT participants who were willing to share their experiences. I hope that this research may be of some assistance in the understanding of CNCP and the difficulties that you often experience.

Thanks to the Chief Investigators of the POINT study, Louisa Degenhardt, Wayne Hall, Richard Mattick, Suzanne Nielsen, Briony Larance, Raimondo Bruno, Nicholas Lintzeris, Milton Cohen, Fiona Shand and Michael Farrell. I have enjoyed being challenged and am thankful for the opportunity to work with you all on a project that is of great importance.

Thanks to the POINT study team, Jessica Belcher, Bianca Hoban, Sarah Freckleton, Anika Martin, Ranira Moodley, Teleri Moore, Kimberley Smith, and Rachel Urquhart- Secord. Your dedication to the project is greatly appreciated. Without your hard work this project would not be possible. v

Thank you to and friends and parents. Natasha Winsley, thank you for always being available to listen, you have endured this PhD with me. To my fellow PhD students, Elizabeth Whittaker and Amanda Roxburgh, I am grateful to have undertaken this journey with you. To my mum, thank you for coming and helping at home so that I was able to concentrate on my PhD. To my dad, thank you for your comments on my thesis. That you had to read my thesis, whilst in hospital, on opioids and in pain must have been insightful.

Thank you to my family for their continual support throughout this journey. To my step-son Robbie, I know you have been missing my cooking and must be grateful this journey is now coming to a close. Thank you to my husband, Geo, for your support and managing the family, house and animals whilst I have been working on my PhD many a weekend. It will be nice to be able to do things as a family again. Finally, to my baby girl, Louella. You have been my main motivator for completing this PhD. Without you, I would have been a major procrastinator, time spent working on my thesis was time away from you, therefore not a minute was wasted.

ABSTRACT Chronic non-cancer pain (CNCP) is a worldwide, common complaint and has a major impact on the individual and community in terms of social, financial and health costs. To date, research has been limited in a number of areas, including the association of CNCP with suicidality, the nature of pharmaceutical opioid use in people living with CNCP, the prevalence of Borderline Personality Disorder (BPD) in people living with CNCP, and associations between suicidality (i.e. ideation, plans and attempts) and problematic opioid use (e.g. dependence) in this population.

This thesis aimed to address these gaps using two data sources: the 2007 Australian National Survey of Mental Health and Well-being, and data from a large, national community survey of people living with CNCP, the Pain and Opioids IN Treatment (POINT) study. The studies included in this thesis aimed to: 1) examine the association of CNCP with suicidality in the Australia population; 2) describe the formation of the POINT cohort to examine prescription opioid use in CNCP; 3) examine the characteristics of CNCP patients prescribed pharmaceutical opioids; 4) examine characteristics associated with pharmaceutical opioid dose and pharmaceutical opioid dependence in the POINT cohort; 5) determine the prevalence of suicidality and independent predictors of general and pain-specific risk factors to suicidality in the POINT cohort, and 6) explore BPD in the POINT cohort and associations with problematic pharmaceutical opioid use and suicidality.

There were a number of important findings from this thesis. Firstly, in the POINT cohort, a large proportion consumed opioids in the high range (over 90mg of oral morphine equivalents mg per day) high-range opioid consumption was associated with problematic opioid use. Secondly, suicidality was common in people living with CNCP, especially in people prescribed opioids for CNCP. Multivariable regressions suggested that an individual’s confidence in coping with their pain was the only independent factor that was predictive of past 12-month progression from ideation to suicide

vii attempt. Finally, screening positively for BPD was high amongst this sample and was associated with pharmaceutical opioid dependence and suicidality.

viii

TABLE OF CONTENTS Declarations ...... ii Originality statement ...... ii Copyright ...... iii Authenticity statement ...... iii Supervisor statement ...... iv Thesis by publication statement ...... iv Acknowledgments ...... v Abstract ...... vii List of tables ...... xiii List of figures ...... xv Abbreviations ...... xvi List of publications included in this thesis ...... xviii List of other publications during candidature...... xix List of presentations ...... xxi 1 Introduction ...... 1 1.1 Chronic non-cancer pain ...... 2 1.1.1 Defining chronic non-cancer pain ...... 2 1.1.2 The biopsychosocial theory of pain ...... 3 1.1.3 Prevalence of chronic non-cancer pain ...... 8 1.1.4 Pharmaceutical opioids in the treatment of chronic non-cancer pain ...... 8 1.2 Suicidality and chronic non-cancer pain ...... 15 1.2.1 Defining suicide-related behaviours ...... 15 1.2.2 Joiner’s Interpersonal Theory of Suicide ...... 16 1.2.3 Prevalence of suicidality in chronic non-cancer pain ...... 17 1.2.4 General and pain-specific suicide risk factors...... 18 1.3 Borderline personality disorder ...... 24 1.3.1 Defining Borderline Personality Disorder ...... 24 1.3.2 Prevalence of Borderline Personality Disorder in chronic non-cancer pain ...... 25 1.3.3 Borderline Personality Disorder and chronic non-cancer pain ...... 26 1.3.4 Borderline Personality Disorder and problematic opioid use in chronic non- cancer pain ...... 26 1.3.5 Borderline Personality Disorder and suicidality in chronic non-cancer pain ...... 27 1.4 Limitations of previous research ...... 27

1.4.1 Problematic opioid use in people with chronic non-cancer pain ...... 28 1.4.2 Suicidality in people living with chronic non-cancer pain ...... 29 1.4.3 Borderline Personality Disorder and chronic non-cancer pain ...... 30 1.5 Aims of thesis ...... 30 1.6 Overview of thesis structure ...... 31 1.6.1 Thesis outline ...... 32 2 Paper one: The prevalence and correlates of chronic pain and suicidality in a nationally representative sample ...... 34 2.1 Copyright Statement ...... 35 2.2 Preamble ...... 36 2.3 Abstract ...... 37 2.4 Introduction ...... 38 2.5 Method ...... 39 2.6 Results ...... 44 2.7 Discussion ...... 53 3 Paper two: Cohort protocol paper: The Pain and Opioids IN Treatment (POINT) study 57 3.1 Copyright Statement ...... 58 3.2 Preamble ...... 59 3.3 Abstract ...... 60 3.4 Introduction ...... 62 3.5 Methods ...... 65 3.6 Discussion ...... 75 4 Paper three: The Pain and Opioids IN Treatment (POINT) study: Characteristics of a cohort using opioids to manage chronic non-cancer pain ...... 78 4.1 Copyright Statement ...... 79 4.2 Preamble ...... 80 4.3 Abstract ...... 81 4.4 Introduction ...... 82 4.5 Methods ...... 83 4.6 Results ...... 91 4.7 Discussion ...... 105 5 Paper four: Pharmaceutical opioid use and dependence among people living with chronic pain: Associations observed within the Pain and Opioids IN Treatment (POINT) cohort ...... 110 5.1 Copyright Statement ...... 111 x

5.2 Preamble ...... 112 5.3 Abstract ...... 113 5.4 Methods ...... 115 5.5 Results ...... 119 5.6 Discussion ...... 129 6 Paper five: Prevalence and correlates of suicidal thoughts and suicide attempts in people prescribed pharmaceutical opioids for chronic pain...... 133 6.1 Copyright Statement ...... 134 6.2 Preamble ...... 135 6.3 Abstract ...... 136 6.4 Introduction ...... 137 6.5 Materials and Methods...... 139 6.6 Results ...... 144 6.7 Discussion ...... 154 7 Paper six: Associations of borderline personality with pain, problems with medications and suicidality in a community sample of chronic non-cancer pain patients prescribed opioids for pain ...... 160 7.1 Copyright Statement ...... 161 7.2 Preamble ...... 162 7.3 Abstract ...... 163 7.4 Introduction ...... 164 7.5 Materials and Methods...... 165 7.6 Results ...... 172 7.7 Discussion ...... 181 8 Discussion and implications ...... 186 8.1 Socio-demographic and clinical characteristics of people living with CNCP and prescribed opioids for CNCP ...... 187 8.2 Suicidality and chronic non-cancer pain ...... 189 8.3 Pharmaceutical opioid use and problematic use of opioids in people living with chronic non-cancer pain ...... 193 8.4 Borderline Personality Disorder and chronic non-cancer pain ...... 196 8.5 Strengths and limitations ...... 198 8.5.1 Strengths ...... 198 8.5.2 Limitations ...... 200 8.6 Clinical recommendations ...... 202 xi

8.6.1 Multidisciplinary approach to CNCP treatment ...... 202 8.6.2 Education and training of general practitioners ...... 203 8.6.3 Guidelines for opioid prescribing in CNCP ...... 204 8.7 Research recommendations ...... 206 9 Conclusions ...... 208 10 References ...... 210 11 Appendices ...... 238 Appendix A: Pain conditions by age-groups ...... 238 Appendix B: Prevalence of suicidality by pain condition...... 239 Appendix C: Prevalence of chronic pain amongst people who had experienced lifetime and past 12-month suicidality, by age-groups ...... 240 Appendix D: Factors associated with past 12-month suicidal behaviours in people with chronic pain1 ...... 241 Appendix E - Glossary of conditions that may lead to chronic pain ...... 242 Appendix F: General pain categories...... 250 Appendix G: Common medications used for pain conditions ...... 252 Appendix H: Initial fax sent to pharmacies ...... 261 Appendix I: Participant invitation to research...... 263 Appendix J: Participant information ...... 265 Appendix K: POINT participant consent form ...... 268 Appendix L: POINT locator form ...... 270 Appendix M: POINT baseline interview ...... 273 Appendix N: POINT baseline self-complete questionnaire ...... 294 Appendix O: POINT 7-day medication diary ...... 307

xii

LIST OF TABLES

Table 1-1: Criteria included in ICD-10, DSM-IV and DSM-5 classification systems...... 13 Table 2-1: Prevalence of chronic pain among Australian adults, 2007 ...... 44 Table 2-2: Socio-demographic and clinical characteristics of people with and without chronic pain, NSMHWB, 2007 ...... 46 Table 2-3: Prevalence of chronic pain amongst people who had experienced lifetime and past 12-month suicidality ...... 48 Table 2-4: Logistic Regression Models#: Odds of lifetime suicidality by presence of chronic pain conditions and ‘any pain’, NSMHWB, 20071 ...... 50 Table 2-5: Logistic Regression Models*: Odds past 12-month suicidality by presence of ‘any pain’ condition, NSMHWB, 20071 ...... 51 Table 3-1: Time-points and data collection method of the POINT prospective cohort ...... 69 Table 3-2: Measures, tools, domains and time-points for data collection for the POINT study ...... 72 Table 4-1: Demographic characteristics of the POINT cohort ...... 92 Table 4-2: Pain conditions and physical functioning of the POINT cohort ...... 94 Table 4-3: Mental health and substance use history of the POINT cohort ...... 98 Table 4-4: Opioid medications, adverse events, non-adherence and aberrant behaviours of the POINT cohort ...... 101 Table 4-5: Other prescription medication and health service use of the POINT cohort ...... 104 Table 5-1: Demographics and clinical characteristics of the POINT cohort (n=1,085) by mean daily opioid consumption (measured in oral morphine equivalent (OME) mg) ...... 122 Table 5-2: Demographic and clinical characteristics of the POINT cohort, according to past year ICD-10 pharmaceutical opioid dependence ...... 127 Table 6-1: Lifetime and Past 12-month suicidal behaviours in the POINT cohort ...... 146 Table 6-2: Correlates and predictors of lifetime suicidal ideation in people that experienced ideation after the onset of their pain ...... 148 Table 6-3: Correlates and predictors of past 12-month suicidal ideation in people with chronic pain ...... 149 Table 6-4: Ideation-to-action in people that experienced a lifetime attempt after the onset of their pain ...... 151 Table 6-5: Past 12-month ideation-to-action in people with chronic pain ...... 153 Table 7-1: Socio-demographic and clinical characteristics of people with and without borderline personality disorder in a sample of people with chronic pain . 173 Table 7-2: Socio-demographic and clinical characteristics of people with and without borderline personality disorder in a sample of people with chronic pain . 176

xiii

Table 7-3: Logistic Regression Models: Independent associations of BPD to pain severity, interference, medication non-adherence, problems with medications, lifetime overdose and lifetime opioid dependence (n=747) ...... 178 Table 7-4: Logistic Regression Models#: Independent associations of BPD to past 12- month suicidal ideation and lifetime suicide attempts (n=916) ...... 180

xiv

LIST OF FIGURES Figure 1-1: A conceptual model of the biopsychosocial interactive processes involved in health and illness. From “Comorbidity of Chronic Mental and Physical Health Conditions: The Biopsychosocial Perspective,” Adapted from R. J. Gatchel (2004), American Psychologist, 59, 792–805...... 4 Figure 2-1: Odds of past 12-month suicidal behaviour amongst those with chronic pain ...... 53 Figure 4-1: Pharmacy recruitment ...... 84 Figure 4-2: Participant recruitment flowchart ...... 86 Figure 5-1: Scatterplot and LOWESS of OME by pain severity score, length of time on opioids, pain relief from medications and pain self-efficacy score ...... 125 Figure 8-1: Comparison of suicidality in the Australian general population (with and without chronic pain) and the POINT sample ...... 191

ABBREVIATIONS BMI Body Mass Index

BPD Borderline personality disorder

BPI Brief Pain Inventory

CI Confidence interval

CIDI Composite International Diagnostic Interview

CNCP Chronic non-caner pain

DSM Diagnostic and Statistical Manual of Mental Health Disorders

EPC Expected population count

GAD General Anxiety Disorder

ICD International Classification of Disease

IMMPACT Initiative on Methods, Measurement and Pain Assessment in Clinical Trials

IPDE International Personality Disorder Examination

MOS Medical Outcomes Study

NSMHWB National Survey of Mental Health and Wellbeing

OR Odds ratio

OME Oral Morphine Equivalent

ORBIT Opioid-Related Behaviours In Treatment Scale

OST Opioid substitution therapy

PC-PTSD Primary Care PTSD screen

PHQ Pfizer Health Questionnaire

PSEQ Pain Self-Efficacy Questionnaire

PODS Prescribed Opioids Difficulties Scale

POINT Pain and Opioids IN Treatment

xvi

PTSD Post Traumatic Stress Disorder

RR Relative Risk

RRR Relative risk ratio

SD Standard deviation

SF-12 Short Form 12

WHO World Health Organisation

WHOQOL-BREF World Health Organisation Quality of Life

WMH World Mental Health

xvii

LIST OF PUBLICATIONS INCLUDED IN THIS THESIS 1. Campbell, G., Darke, S., Bruno, R., & Degenhardt, L. (2015). The prevalence and correlates of chronic pain and suicidality in a nationally representative sample. Australian and New Zealand Journal of Psychiatry 49(9) 803-11.

2. Campbell, G., Mattick, R., Bruno, R., Larance, B., Nielsen, S., Cohen, M., Lintzeris, N., Shand, F., Hall, W., Hoban, B., Kehler, C., Farrell, M., & Degenhardt, L. (2014). Cohort protocol: The Pain and Opioids IN Treatment (POINT) study. BMC Pharmacology and Toxicology, 15; 17

3. Campbell, G., Nielsen, S., Bruno, R., Lintzeris, N., Cohen, M., Hall, W., Larance, B., Mattick, R.P., & Degenhardt, L. (2015).The Pain and Opioids IN Treatment (POINT) study: Characteristics of a cohort using opioids to manage chronic non-cancer pain. Pain, 156(2), 231-42

4. Campbell, G., Nielsen, S., Bruno, R., Larance, B., Smith, K., Hall, W., Cohen, M., & Lintzeris, N., & Degenhardt, L., (2015). Correlates of pharmaceutical opioid use and dependence among people living with chronic pain: Baseline findings from the Pain and Opioids IN Treatment (POINT) study. Pain Medicine 16(9) 1745-58.

5. Campbell, G., Bruno, R., Shand, F., Hall, W., Darke, S., Farrell, M., & Degenhardt, L. (2015). Prevalence and correlates of suicidal thoughts and suicide attempts in people prescribed pharmaceutical opioids for chronic pain. Clinical Journal of Pain. Doi: 10.1097/ajp.0000000000000283

6. Campbell, G., Bruno, R., Darke, S., & Degenhardt, L. (2015). Associations of borderline personality with pain, problems with medications and suicidality in a community sample of chronic non-cancer pain patients prescribed opioids for pain. General Hospital Psychiatry, 37(5), 434-440. xviii

LIST OF OTHER PUBLICATIONS DURING CANDIDATURE 1. Belcher, J., Nielsen, S., Campbell, G., Bruno, R., Larance, B., Lintzeris, N., & Degenhardt, L., (2014). Diversion of prescribed opioids by people living with chronic pain: Results from an Australian community sample. Drug and Alcohol Review, 33(1), 27-32.

2. Nielsen, S., Lintzeris, N., Bruno, R., Campbell, G., Larance, B., Hoban, B., Hall, W., Cohen, M., & Degenhardt, L. (2015). Benzodiazepine use amongst chronic pain patients prescribed opioids: associations with pain, physical and mental health and health service utilization. Pain Medicine, 16(2), 356-66

3. Degenhardt, L., Lintzeris, N., Campbell, G., Bruno, R., Cohen, M., Farrell, M., & Hall, W., (2015). Experience of adjunctive cannabis use for chronic non-cancer pain: Findings from the Pain and Opioids IN Treatment (POINT) study. Drug and Alcohol Dependence, 147, 144-50.

4. Degenhardt, L., Bruno, R., Lintzeris, N., Hall, W., Nielsen, S., Larance, B., Cohen, M., & Campbell, G., (2015). DSM and ICD definitions of pharmaceutical opioid dependence in people treated with opioids for chronic pain: Findings from the Pain and Opioids IN Treatment (POINT) study. The Lancet Psychiatry, 2(4), 314-322

5. Lintzeris, N., Moodley, R., Campbell, G., Bruno, R., Larance, B., Nielsen, S., & Degenhardt, L. (2015). Sleep quality among people living with chronic non-cancer pain: Findings from the Pain and Opioids IN Treatment (POINT) cohort. Clinical Journal of Pain. doi: 10.1097/AJP.0000000000000282

6. Smith, K., Mattick, R., Bruno, R., Nielsen, S., Cohen, M., Campbell, G., Larance, B., Farrell, M., & Degenhardt, L. (2015). Factors associated with the development of depression in chronic non-cancer pain patients following the onset of opioid treatment for pain. Journal of Affective Disorders, 184, 72-80. xix

7. Nielsen, S., Campbell, G., Peacock, A., Smith, K., Larance, B., Cohen, M., Lintzeris, N., Hall, W., & Degenhardt, L. (accepted). The use of non-opioid interventions among people living with chronic non-cancer pain. Australian Health Review

8. Peacock, A., Degenhardt, L., Campbell, G., Larance, B., Nielsen, S., Hall, W., Mattick R.P., & Bruno, R. (submitted). Predicting non-adherence utilizing a typology of risk among chronic non-cancer pain patients prescribed opioids: A cohort study. Pain Physician

9. Larance, B., Campbell, G., Peacock, A., Lintzeris, N., Bruno, R., Farrell, M., & Degenhardt, L. (submitted). Alcohol use disorders, risky drinking and adverse events among chronic non-cancer pain patients receiving opioid therapy. Drug and Alcohol Dependence.

LIST OF PRESENTATIONS Presenting author

1. Campbell, G. (2014). POINT: Pain and Opioids IN Treatment study: Preliminary findings. Invited presentation at the National Pharmacy Guild Conference; Sydney Convention Centre; Sydney, Australia, 21st June, 2014.

2. Campbell, G. and Hoban, B. (2014) POINT: Pain and Opioids IN Treatment study: Preliminary findings. Presentation at the NDARC in-house seminar series; Sydney, Australia, 26th June 2014.

3. Campbell, G., Darke, R., Bruno, R. & Degenhardt, L. Prevalence, correlates and associations of suicidality in people with chronic non-cancer pain. Poster presentation Postgraduate Research Symposium, UNSW; Sydney, Australia, 29th August 2014.

4. Campbell, G. (2014). Characteristics of a national sample of people with chronic non-cancer pain prescribed prescription opioids. Poster presented at the National Drug and Alcohol Research Centre Annual Symposium, University of New South Wales; Sydney, Australia, 8th September, 2014

5. Campbell, G. Degenhardt, L., Larance, B., Nielsen, S., Hall, W., Cohen, M., Lintzeris, N., Bruno, R., Mattick, R., Shand, F. and Farrell, M. (2014). Characteristics of a national sample of people with chronic non-cancer pain prescribed opioids. Symposium presentation at the Australian Professional Society on Alcohol and other Drugs, Adelaide Convention Centre; Adelaide, Australia, 10-12th November, 2014.

6. Campbell, G., Nielsen, S., Larance, B., Bruno, R., Mattick, R.P., Hall, W, Lintzeris, N., Cohen, M., Smith, K. and Degenhardt. L. Pharmaceutical opioid use and dependence among people living with chronic pain: Associations observed within the Pain and Opioids IN Treatment (POINT) cohort. Poster presentation at the xxi

National Institute on Drug Abuse (NIDA) International Forum; Phoenix, United States, 12-15th June, 2015.

7. Campbell, G., Nielsen, S., Larance, B., Bruno, R., Mattick, R.P., Hall, W, Lintzeris, N., Cohen, M., Smith, K. and Degenhardt, L. Pharmaceutical opioid use and dependence among people living with chronic pain: Associations observed within the Pain and Opioids IN Treatment (POINT) cohort. Oral presentation at the College on Problems of Drug Dependence 77th Annual Scientific Meeting; Phoenix, United Sates, 11-16th June 2015.

8. Campbell, G., Darke, R., Bruno, R. & Degenhardt, L. Prevalence, correlates and associations of suicidality in people with chronic non-cancer pain. Oral presentation The National Suicide Prevention conference; Hobart, Australia, 26- 29th July, 2015.

9. Campbell, G. (2015). Pharmaceutical opioid use and dependence among people living with chronic pain: Associations observed within the Pain and Opioids IN Treatment (POINT) cohort. St. Vincent’s weekly seminar, Drug and Alcohol Services; Sydney, Australia, 7th July, 2015.

10. Campbell, G., Nielsen, S., Larance, B., Bruno, R., Mattick, R.P., Hall, W, Lintzeris, N., Cohen, M., Smith, K. and Degenhardt. L. Pharmaceutical opioid use and dependence among people living with chronic pain: Associations observed within the Pain and Opioids IN Treatment (POINT) cohort. Presentation at the NDARC Annual Symposium, Sydney Australia, 15th September, 2015.

11. Campbell, G., Darke, R., Bruno, R. & Degenhardt, L. Prevalence, correlates and associations of suicidality in people with chronic non-cancer pain. Poster presentation NDARC Annual Symposium, Sydney Australia, 15th September, 2015.

xxii

Contributing author

1. Nielsen, S., Bruno, R., Campbell, G., Cohen, M., Farrell, M., Hall, W., Hoban, B., Larance, B., Lintzeris, N., Mattick, R. & Degenhardt L. (2014) Benzodiazepine use amongst a sample of chronic pain patients prescribed opioids. College on Problems of Drug Dependence 76th Annual Meeting; San Juan, Puerto Rico, 14-19th June, 2014.

2. Larance, B., Peacock, A., Campbell, G., Freckleton, S., Smith, K., Mattick, R., Bruno, R., Nielsen, S., Cohen, M., Lintzeris, N., Shand, F., Hall, W., Farrell, M., Degenhardt, L. (2014) Alcohol use disorders, risky alcohol consumption and adverse events among chronic non-cancer pain patients receiving opioid therapy. Poster presented at the National Drug and Alcohol Research Centre Annual Symposium, University of New South Wales; Sydney, Australia, 8th September, 2014.

3. Moodley, R., Lintzeris, N., Campbell, G., Bruno, R., Larance, B., Nielsen, S., & Degenhardt, L. (2014). Sleep quality among people living with chronic non-cancer pain: Findings from the Pain and Opioids IN Treatment (POINT) cohort. Poster presented at the National Drug and Alcohol Research Centre Annual Symposium, University of New South Wales; Sydney, Australia, 8th September, 2014

4. Smith K, Mattick R.P, Campbell G, Larance B & Degenhardt L. (2014) Correlates of depressive disorders with an onset post-opioid initiation among chronic non-cancer pain patients. Poster presented at the National Drug and Alcohol Research Centre Annual Symposium, University of New South Wales; Sydney, Australia, 8th September, 2014.

5. Larance, B., Nielsen, S., Hordern, A., Lintzeris, N., Burns, L., Mattick, R., Campbell, G. Farrell, M., Cohen, M., Bruno, R., Brown, A., Johnson, J., and Degenhardt, L. (2014). Pharmaceutical opioid use and pain among people who inject drugs:

xxiii

Baseline findings from a prospective cohort. Symposium presentation at the Australian Professional Society on Alcohol and other Drugs, Adelaide Convention Centre; Adelaide, Australia, 10-12th November, 2014.

6. Nielsen S; Larance B; Lintzeris N; Holliday S; Vanderhaven M; Hordern A; Dunlop A; Haber P; Murnion B; Silsbury C; Johnson J; Demirkol A; Sadler C; Phung N; Burns L; Mattick R; Campbell G; Farrell M; Cohen M; Bruno R; Hardy M; Brown A; Houseman J; and Degenhardt L, (2014) Seeking treatment for pharmaceutical opioid dependence: baseline characteristics from an in-treatment cohort. Symposium presentation at the Australian Professional Society on Alcohol and other Drugs, Adelaide Convention Centre; Adelaide, Australia, 10-12th November, 2014.

7. Larance, B., Peacock, A., Campbell, G., Freckleton, S., Smith, K., Mattick, R., Bruno, R., Nielsen, S., Cohen, M., Lintzeris, N., Shand, F., Hall, W., Farrell, M., Degenhardt, L. (2014) Alcohol use disorders, risky alcohol consumption and adverse events among chronic non-cancer pain patients receiving opioid therapy. Presentation to the Australian Professional Society on Alcohol and other Drugs, Adelaide Convention Centre; Adelaide, Australia; 10-12th November, 2014.

8. Degenhardt, L., Lintzeris, N., Campbell, G., Bruno, R., Cohen, M., Farrell, M., & Hall, W., (2015). Experience of adjunctive cannabis use for chronic non-cancer pain: Findings from the Pain and Opioids IN Treatment (POINT) study. National Institute on Drug Abuse (NIDA) International Forum; Phoenix, United States, 12-15th June, 2015.

9. Degenhardt, L, Bruno, R., Lintzeris, N., Hall, W., Nielsen, S., Larance, B., Cohen, M. and Campbell, G. Examining DSM and ICD definitions of pharmaceutical opioid dependence in people taking opioids for chronic pain: Findings from the Pain and Opioids IN Treatment (POINT) study. College on Problems of Drug Dependence 77th Annual Scientific Meeting; Phoenix, United Sates, 11-16th June 2015. xxiv

10. Peacock, A., Degenhardt, L., Lintzeris, N., Larance, B., Campbell, G., Nielsen, S., Cohen, M., & Bruno, R. (2015). Opioid therapy for chronic non-cancer pain: Does typology of risk predict aberrant behaviour? College on Problems of Drug Dependence 77th Annual Scientific Meeting; Phoenix, United Sates, 11-16th June 2015.

11. Larance, B., Peacock, A., Campbell, G. Bruno, R., Mattick, R., Nielsen, S., Cohen, M., Smith, K., Degenhardt, L. (2015) Alcohol use disorders and risky alcohol consumption among chronic non-cancer pain patients receiving opioid therapy. College on Problems of Drug Dependence 77th Annual Scientific Meeting; Phoenix, United Sates, 11-16th June 2015.

12. Nielsen, S., Campbell, G., Larance, B., Bruno, R., Mattick, R., Hall, W., Lintzeris, N., Cohen, M., Smith, K., and Degenhardt, L. Is higher opioid dose associated with better functioning or more pain relief? Findings from the pain and opioids in treatment (POINT) study. European 9th congress of the European pain federation (EFIC); Vienna, Austria, 2-5th September 2015.

13. Larance, B., Campbell, G., Nielsen, S., Bruno, R., Lintzeris, N., Cohen, M., & Degenhardt, L. Medication control and help-seeking for opioid-related problems among a cohort of chronic non-cancer pain patients prescribed strong opioids. Presentation at the NDARC Annual Symposium, Sydney Australia, 15th September, 2015.

14. Larance, B., Campbell, G., Nielsen, S., Bruno, R., Lintzeris, N., Cohen, M. & Degenhardt L. Help-seeking for opioid medication-related problems among a cohort of chronic non-cancer pain patients prescribed strong opioids. Presentation to the Australian Professional Society on Alcohol and other Drugs, Pan Pacific Perth; Perth, Australia; 8-11th November, 2014.

xxv

Chapter one: Introduction

1 INTRODUCTION Chronic non-cancer pain (CNCP) is a common worldwide complaint, with rates in the general population estimated from 17% (Blyth et al. 2001) to 41% (Tsang et al. 2008) in the past year. CNCP is associated with poorer physical health (Eriksen et al. 2006), poorer quality of life (Hagg, Fritzell, and Nordwall 2003; Ojala et al. 2014) and a greater risk of developing depression (Currie and Wang 2004) and other mental health problems (Dersh, Polatin, and Gatchel 2002; McWilliams, Cox, and Enns 2003). Although there are behavioural, physical and medication interventions for chronic pain, it is often the case that many patients have to accept a lifetime treatment plan for the management of their pain, which may not alleviate the pain completely (Breivik et al. 2006).

Over the past decade, there has been increasing professional and public concern regarding the use of pharmaceutical opioids and related harms (Leong, Murnion, and Haber 2009b; Compton and Volkow 2006b). This has been driven by increases in prescribing of these drugs, especially in the USA and Canada. The increase in prescribing in Australia has been less than in Europe and the United States (US) (Centers for Disease -Control Prevention 2008; Dhalla et al. 2009; Fredheim et al. 2010), but nonetheless, between 1992-2012 the number of opioid prescriptions in Australia increased 15-fold (Blanch, Pearson, and Haber 2014). Examination of problematic use of opioids amongst people living with CNCP has been limited for a number of reasons. There has been a focus on clinical trials, which often exclude people with complex comorbidities, and therefore, the people most likely to encounter problems with their opioid use, and problems in consistent definitions of problematic use, a focus on clinical samples, which limit the generalisability of the findings and problems with defining problematic use, including non-adherent or aberrant behaviours (Smith et al. 2013; Smith, Paillard, et al. 2015; Vowles et al. 2015).

There has also been emerging research examining the association between CNCP and suicidality (Calati et al. 2015). Previous research has found that suicide-related Chapter one: Introduction

behaviours are two-three times greater in those with CNCP. Further, there are both general (e.g. previous suicide attempt, depression, substance use) and pain-specific factors (e.g. location of pain, pain severity, length of time in pain) that need to be considered (Hassett, Aquino, and Ilgen 2014; Hooley, Franklin, and Nock 2014; Tang and Crane 2006). Although there have been epidemiological studies examining prevalence of suicide-related behaviours and CNCP (Braden and Sullivan 2008; Ilgen et al. 2008; Ratcliffe et al. 2008), much less is known about the influence of general and pain-specific risk factors due to limits arising from small sample sizes, or focus on clinical populations which are not representative of the whole CNCP population (Hassett, Aquino, and Ilgen 2014; Ilgen, Kleinberg, Ignacio, and et al. 2013; Ilgen et al. 2010; Racine, Choiniere, and Nielson 2014; Smith et al. 2004). More research is essential to our understanding of the association between CNCP and suicidality.

The aims of this chapter are to provide a detailed overview of the research into CNCP and its association with problematic opioid use and suicidality. Further, several key clinical issues will be discussed, such as Borderline Personality Disorder (BPD), which are common in people living with CNCP. Secondly, limitations of previous research, and therefore, the context for this current thesis will be discussed. An overview of the thesis will be presented at the end of the chapter.

1.1 Chronic non-cancer pain

1.1.1 Defining chronic non-cancer pain

The International Association for the study of Pain defines pain (acute and chronic) as an “unpleasant sensory and emotional experience associated with actual or potential tissue damage or described in terms of such damage” (Merskey 1986). CNCP is defined as 'pain which has persisted beyond normal tissue healing time' (Merskey 1986). Although there is no formally accepted time for this, pain lasting for more three months is generally considered to be chronic (Merskey and Bogduk 1994).

Chapter one: Introduction

1.1.2 The biopsychosocial theory of pain

Traditional approaches to pain purport a dualistic approach of the mind and body functioning separately and independently (Gatchel et al. 2007), referred to as the biomedical model, or “unidimensional models”. This traditional model implies a direct transmission of pain from the periphery to the spine and then the brain. As a consequence, a patient’s report of pain should be directly and proportionally related to the extent of their physical pathology (Flor and Turk 2011). Challenges to this theory arose where patients with identical diseases associated with pain, such as degenerative disk disease, varied greatly in their reports of pain severity and pain behaviour (Rosenblum et al. 2008). Pain not supported by objective measures, or deemed disproportionate to the identifiable physical pathology, was thought to be the result of psychological factors, such as personality traits and underlying psychopathology (Flor and Turk 2011). Although pain related to psychological factors, exclusively or primarily, occurs, it was considered to be far less prevalent than pain associated with organic processes, that are powerfully influenced by psychosocial mediators and psychiatric comorbidities (Rosenblum et al. 2008).

The recognition of a dynamic relationship between mental and physiological components led to criticism of the biomedical model. In the 1970s the biopsychosocial model was proposed as a counterpoint to the biomedical model (Engel 1977). Rather than focusing primarily on physical signs and symptoms, the biopsychosocial model takes into consideration physical, psychological, and social components of the individual’s problems (detailed below), with the view that the mind and the body are inherently connected (see Figure 1-1), and the understanding that the experience of pain differs for each individual.

The biopsychosocial model has been especially influential in the area of chronic pain (Gatchel et al. 2007). The model suggests that biological changes, psychological status and sociocultural context all need to be considered to understand a patient’s pain perception and response to pain. A focus on only one of these aspects, for example the 3

Chapter one: Introduction

physical aspect, will mean that treatment is incomplete and the patient’s experience of CNCP may continue (Gatchel et al. 2007; Kamper, Apeldoorn, Chiarotto, Smeets, et al. 2014). Detailed below are the biological, psychological and social factors associated with CNCP. More attention will be given to the psychological and social aspects of CNCP as these factors are covered throughout the thesis.

Figure 1-1 has been removed due to Copyright restrictions. For a conceptual model of the biopsychosocial interactive processes involved in health and illness see “Comorbidity of Chronic Mental and Physical Health Conditions: The Biopsychosocial Perspective,” R. J. Gatchel (2004), American Psychologist, 59, 792–805

1.1.2.1 Biological dimensions of CNCP Under the biological component of the biopsychosocial model it is acknowledged that there are physical aspects, including neurophysiological processes, which interact with the endocrine and autonomic systems in the body, maintaining pain symptoms. This includes the process of nociception, which refers to stimulation of nerves “that convey information about potential tissue damage to the brain” (Goodacre and McArthur 2013) (pp.158). It is now accepted that genetics also play a crucial role in the

4 Chapter one: Introduction

susceptibility, maintenance and aggravation of CNCP. Further, an imbalance in the variability of neurotransmitters and types of receptors may influence chronic pain, that neurons are wired in certain patterns and are dynamic, and that chronic pain is phasic depending on the psychosocial status of the patient and can change within hours, days, or weeks (Gatchel et al. 2007).

1.1.2.2 Psychological dimensions of CNCP Psychological factors involve both emotion and cognition. Emotion is the immediate reaction to the experience of pain, while cognitions can be a process whereby the individual attaches meaning to the emotional experience (Gatchel et al. 2007). Anxiety, depression and anger all play a significant role in a person’s negative affect and, therefore, their experience of pain (Gatchel et al. 2007). As negative affect can play a significant role on treatment motivation and treatment compliance, understanding of a person’s psychological state is considered imperative to successful treatment (Gatchel et al. 2007). Factors including previous experiences with pain, and consequences of pain, mood, and stress can influence initial reactions to the pain sensation and psychological factors can lead to a vicious circle of nociception, pain and distress, and disability is maintained (Gatchel et al. 2007).

Empirically, CNCP has consistently been associated with mental health problems (Blyth et al. 2001; Breivik et al. 2006; Currie and Wang 2004; McWilliams, Cox, and Enns 2003; McWilliams, Goodwin, and Cox 2004; McWilliams and Higgins 2013). Epidemiological studies have found major depression to be elevated in people living with CNCP (Currie and Wang 2004; Scott et al. 2010). Other mental health disorders, including, dysthymia, generalised anxiety disorder, panic disorder and social phobia, have also been found to be elevated in people living with CNCP (Demyttenaere et al. 2007; McWilliams, Cox, and Enns 2003). CNCP has also been found to associated with personality disorders, such as Borderline Personality Disorder (BPD) (McWilliams and Higgins 2013).

Chapter one: Introduction

CNCP patients have been found to be two-three times more likely to suffer from anxiety and depression, either alone or combined, than the general population. This was greater in those with back/neck problems and headaches compared with those with arthritis. Further, those with comorbid anxiety and depression had 4.5 times the odds of reporting multiple pain conditions (a combination of back/neck, arthritis and headaches) compared with those with no comorbid anxiety and depression, adjusting for age and gender (Scott et al. 2010).

Not only has CNCP been found to be associated with mental health, CNCP can also result in the development of mental health problems, such as depression (Breivik et al. 2006; Fishbain et al. 1997; Smith, Mattick, et al. 2015). The presence of depression in CNCP can also result in poorer physical functioning and poorer pain treatment outcomes (Bair et al. 2003).

Cognition involves a patient’s beliefs, pain catastrophizing, fear-avoidance beliefs and perceived pain self-efficacy which can inherently trigger further emotional responses and, thereby, amplify the experience of pain as well as influence pain behaviour (Gatchel 2004). Furthermore, analgesic medications have been found to lead to cognitive impairment (Moriarty, McGuire, and Finn 2011). Research has found that patients who are better able to cope with their pain, who avoid catastrophizing, see themselves as less physically limited, than other patients, appear to function better (Jensen et al. 1991). A Cochrane review found that Cognitive Behavioural Therapy (CBT), which aims at addressing these negative cognitions, shows small positive effects for pain, disability and mood in people with CNCP (Eccleston, Williams, and Morley 2009). These findings support the importance of cognitions on the course and consequence of CNCP.

1.1.2.3 Social dimensions of CNCP The social dimension involves the context of the patient, including environmental stressors, interpersonal relationships, and social support, and involves the ways that 6

Chapter one: Introduction

these social factors interact with the bio- and psycho-dimensions (Gatchel et al. 2007). CNCP can have a major impact on a person’s physical functioning in terms of ability to engage in daily activities, sleep, exercise, household chores, social activities and quality of life (Blyth et al. 2001; Breivik et al. 2006). Further, people living with CNCP have been found to have significantly poorer health-status, poorer functioning and experience problems in their social relationships (Reid et al. 2011). A large cross- national European study found two-thirds of people were less able or unable to sleep because of their pain, and about half found walking and household chores difficult due to pain (Breivik et al. 2006). In Australia, 13.5% of female respondents and 11.0% of male respondents reported that their CNCP interfered to some degree with their daily functioning (Blyth et al. 2001). CNCP has also been found to decrease a person’s quality of life to the extent they may not consider life worth living (Ojala et al. 2014).

CNCP can also have a major impact on an individual’s employment status. In a large, random, community sample of 5,036 CNCP patients in the United Kingdom, people living with CNCP were eight times more likely to be unable to work due to illness or disability, and were more likely to be living in a council rented property, compared with those without CNCP (Elliott et al. 1999). In a large cross-national European study, of 46,394 respondents, one-quarter reported that CNCP had impacted upon their careers (Beubler et al. 2006). In an Australian study people living with CNCP were less likely to be employed (full-time or part-time), were more likely to be receiving unemployment or disability pension, and after adjusting for age, gender and comorbidity, unemployment remained significantly associated with CNCP (Blyth et al. 2001). In a large, random sample of 28, 902 working adults in the United States (Stewart et al. 2003), it was found that 13% of the total workforce experienced a loss in productive time in a two-week period due to a pain condition.

As can be seen, CNCP is complex and the varied aetiology can impact on the experience, and consequently, the treatment of CNCP (Blyth et al. 2004; Blyth et al. 2001; Blyth, March, and Cousins 2003; Savage 1999). People living with CNCP suffer

Chapter one: Introduction

from a multitude of complex problems including, physical health functioning, lowered quality of life, mental health comorbidities and employment and financial difficulties. Holistic treatments, which encompass the individual as a whole, are therefore considered the best approach. Two systematic reviews have found that multidisciplinary treatments that involve specialists from a number of fields such as pain medicine, physical therapists and mental health professionals, are the most effective treatment of CNCP in the individual (Kamper, Apeldoorn, Chiarotto, Smeets, et al. 2014; Scascighini et al. 2008).

Chapter 3 of this thesis will describe the Pain and Opioids IN Treatment (POINT) study, a large national sample of community patients prescribed opioids for CNCP, on which Chapters 4-8 are based. Chapter 4 will examine the characteristics of the cohort on a number of demographic and clinical variables to understand the complexities and comorbidities of people living with CNCP.

1.1.3 Prevalence of chronic non-cancer pain

In 2010, the Global Burden of Disease 2010 study estimated that tension-type headache affected 21% of the global population (equivalent to 1.4 billion people), 15% suffered from migraines (1.0 billion), 9% suffered from lower back pain (630 million people), and 5% currently suffered from neck pain (332 million people) (Vos et al. 2012). Low back pain, neck pain and migraine were the 1st, 4th and 8th largest contributors respectively to global non-fatal health burden (years lived with disability) (Vos et al. 2012). CNCP is a common, world-wide problem and is likely to increase with an ageing population.

1.1.4 Pharmaceutical opioids in the treatment of chronic non-cancer pain

Although there are behavioural, physical and medication interventions for chronic pain, the focus of this thesis is prescription opioids, so details of other treatments will not be discussed here. In the last two decades there has been an increase in the 8

Chapter one: Introduction

prescription of pharmaceutical opioids for CNCP in many developing countries. This increase has been based on combined efforts of clinical needs and recommendations from pain clinicians (Eriksen et al. 2006). Among the general population, between 3% (Eriksen et al. 2006), 5% (Breivik et al. 2006), and 7% (Moulin et al. 2002) of people living with CNCP are currently prescribed strong opioids (those medications associated with drugs of addiction), however the rates of prescribing opioids for CNCP are on the increase (Chou et al. 2015).

The increase in the use of opioid analgesics for CNCP would suggest there is conclusive evidence regarding their positive impact on pain reduction and quality of life. Indeed, qualitative and quantitative reviews of the evidence have concluded that chronic opioid therapy does produce clinically significant reductions in pain, albeit in the range of 2 to 3 points on a 0 to 10 visual analogue scale, or around 30% (Chou 2009; Kalso et al. 2004; Noble et al. 2010). This is a best-case scenario because Randomised Controlled Trials (RCT) often select the sub-set of patients most likely to receive a clinical benefit, without complex comorbidities, and have short follow-up periods (average trial duration of sixteen weeks, longest study duration one year), meaning they do not report on longer term outcomes (Ballantyne and Shin 2008; Moulin et al. 1996). It is important to note that no studies have yet run for over 12-months, enough to demonstrate long-term benefit of opioids for chronic non-cancer pain (Chou et al. 2015; Martell et al. 2007).

Additionally, much less is known about the impacts of opioids on a CNCP patient’s quality of life. Eriksen et al (2006) found that opioid use in CNCP patients was significantly associated with moderate/severe or very severe pain, poor self-rated health, not being engaged in employment, higher use of the health care system, and had a negative influence on quality of life, compared with those with CNCP not prescribed opioids. The authors concluded that “it is remarkable that opioid treatment of long-term/chronic non-cancer pain does not seem to fulfil any of the key outcome

Chapter one: Introduction

opioid treatment goals: pain relief, improved quality of life and improved functional capacity.” (Eriksen et al. 2006, pp.176).

Opioid treatment guidelines specify that careful risk assessments need to be undertaken and that low-risk patients are the most suitable for opioid therapy, particularly long-term opioid therapy (Chou 2009). Despite this, there have been a number of studies that have found that those prescribed opioids for CNCP, and receiving the highest doses, are the people with the most complex comorbidities (Rogers et al. 2013; Sullivan 2010; Sullivan and Howe 2013). The term “adverse selection” has been coined to describe this apparent contradiction in which the likelihood of a patient receiving opioid therapy increases as the number of risk factors for adverse outcomes increases (Sullivan and Howe 2013). Sullivan and Howe (2013) reference two studies supporting this concern. In the Trends and Risks of Opioid Use for Pain study between 2000 and 2005 the prevalence of opioid prescriptions grew fastest amongst those CNCP patients with mental health or substance use problems (Sullivan and Howe 2013). This effect has also been found in the CONsortium to Study Opioid Risks and Trends study, which identified that a history of mental health and substance use disorders greatly increased the likelihood of receiving long-term opioid therapy (Sullivan and Howe 2013).

Although there is support from peak pain organisations of the use of opioids in the treatment of CNCP (American Academy of Pain Medicine and American Pain Society 1997; Beubler et al. 2006; Cohen and Wodak 2010; Royal Australasian College of Physicians 2008), the use of opioids for CNCP has been associated with adverse outcomes, such as overdose and concerns about problematic opioid use in people with CNCP (Chou et al. 2015). Debate continues about how, when, and in what manner opioids should be prescribed for this diverse patient group (Baca and Grant 2007; Ballantyne 2006; Ballantyne 2007; Eriksen et al. 2006; Lipman 2007; Savage 1999).

Chapter one: Introduction

1.1.4.1 Problematic prescription opioid use in people living with CNCP This increase in prescribing has been accompanied by increased concern about the appropriateness of prescribing these drugs for CNCP (Noble et al. 2010). Indeed, pharmaceutical opioids now comprise the majority of fatal and non-fatal drug overdoses in the US and Australia (Bohnert et al. 2014; Hasegawa et al. 2014; Roxburgh et al. 2011), and for some patients, there is the development of problematic and/or dependent use (Chou et al. 2015). Problematic use may occur for a number of reasons, including, self-medication, reward, compulsive use due to addiction and diversion for profit (Vowles et al. 2015).

The use of opioids, within, and outside the bounds of a doctor’s prescription has been cause for concern because of the risk of iatrogenic dependence (Compton and Volkow 2006a). One review found that 3.3% developed opioid abuse/dependence and 11.5% of chronic pain patients engaged in aberrant drug-related behaviours such as diversion, prescription forgery, injecting, multiple episodes of prescription loss, escalating doses, and doctor shopping, while (Fishbain, Cole, Lewis, Rosomoff, Rosomoff, et al. 2008). It is important, however, to understand the patient samples involved in these studies. Many of the 67 studies were clinical trials, which typically find far lower rates of aberrant drug-related behaviours and abuse/dependence due to the exclusion of people with complex comorbidities (e.g. history of substance use disorder) in the screening phase (Fishbain, Cole, Lewis, Rosomoff, Rosomoff, et al. 2008).

There is also inconsistency in reporting problematic behaviours. In a recent review of 38 studies, Vowles et al (Vowles et al. 2015) acknowledged that many studies were inconsistent in their reporting of problematic use. They used predefined definitions of problematic behaviours, including, misuse, abuse and addiction and found that 76% of studies reported on misuse, 32% on addiction and only one study reported on abuse. The authors calculated prevalence rates of 2%-29% for opioid misuse, 8%-12% for addiction (Vowles et al. 2015). Many of these studies, however, were based on clinical

Chapter one: Introduction

pain samples, or were from primary care with smaller sample sizes and limit the generalisability to CNCP in the community. Further, none of the studies examined all factors associated with addiction, misuse, and abuse in the one study, and typically focused on only one area (Vowles et al. 2015).

There is little consensus regarding the diagnosis of opioid dependence and problematic use in the context of chronic opioid treatment for chronic pain and there are a number of varied approaches in measuring problematic use (Smith, Paillard, et al. 2015). While there is less controversy in diagnosing substance use disorder for illicit drug use, there is much less certainty about diagnosing iatrogenic opioid addiction (Ballantyne and LaForge 2007). When patients are maintained on opioids for the treatment of pain, there is currently no satisfactory means of distinguishing true addiction from problematic behaviours caused by a variety of factors other than addiction (Smith,

Paillard, et al. 2015). In this current thesis, problematic opioid use was measured in a number of ways, detailed below.

Pharmaceutical opioid dependence

There are two major classification systems for opioid use disorders, the World Health Organization (WHO) International Classification of Diseases (ICD) and the American Psychiatric Association Diagnostic and Statistical Manual (DSM). Whereas the DSM-IV broadly operationalised opioid abuse and opioid dependence (American Psychiatric Association 2000) in similar ways to the terms used in ICD-10, the DSM-5 shifted to classifications of opioid use disorders (mild, moderate and severe) (American Psychiatric Association 2013). A total of 11 criteria are included; a person who fulfils at least two of these meets the definition of a use disorder (see Table 1-1).

In the pain medicine field, opioid dependence is often referred to as “addiction” (Martell et al. 2007), with less emphasis upon tolerance and withdrawal, which are considered to be physiological consequences of long term opioid use and not markers

Chapter one: Introduction

of problematic use. In DSM-5, tolerance and withdrawal are among these criteria, but a person taking opioids “solely under medical supervision” is not judged to fulfil either one (American Psychiatric Association 2013), in keeping with the view of pain physicians (describe as “conditional exclusion”). For a comparison of the symptoms across the different classification systems see Table 1-1.

Table 1-1: Criteria included in ICD-10, DSM-IV and DSM-5 classification systems

DSM ICD Criterion Definition DSM-IV DSM-5 ICD-10 Draft ICD-11 diagnoses diagnoses diagnoses dependence Larger Used in larger amounts or Dependence Use disorder longer than intended Cut down Persistent desire or Dependence Use disorder Dependence+ unsuccessful efforts to reduce Dependence# use Craving Craving or a strong desire for n/a Use disorder Dependence pharmaceutical opioids Time spent Great deal of time spent using Dependence Use disorder and recovering from use Dependence++ Give up Social, occupational, other Abuse Use disorder activities reduced due to use Social Continued used despite Dependence Use disorder # Dependence recurrent social/interpersonal Harmful use+++ problems due to use Continue Continued use despite Dependence Use disorder recurrent physical or Dependence++++ psychological problems Major role Recurrent use resulting in Abuse Use disorder failure to fulfil major role Harmful use n/a obligations due to use Legal Recurrent use despite legal Abuse n/a Harmful use n/a problems Hazard Recurrent use in hazardous Abuse Use disorder Harmful use n/a situations Tolerance Tolerance Dependence Use disorder1 Dependence Dependence+++ Withdrawal Withdrawal Dependence Use disorder1 Dependence + ICD 10 combines ’used in larger amounts or longer than intended’ and ‘persistent desire or unsuccessful efforts to reduce use’. ++ ICD 10 combines ‘great deal of time spent using and recovering from use’ and ‘social, occupational, other activities reduced due to use’. +++ ICD 10 does not include the extra requirement of continued use despite social problems # ICD-11 draft criteria propose combining these. ++++ ICD-10 dependence criteria require social/interpersonal problems due to use AND continued use despite these problems 1. Tolerance and withdrawal, although perhaps reported, do not count in the criteria for DSM-5 use disorder if people are “taking opioids solely under medical supervision”. 2. The levels of tolerance and withdrawal without a conditional exclusion based upon patient adherence.

In a recent paper (Degenhardt et al. 2015), we found that the ICD-11 had excellent agreement with previous gold standard classifications and was the best model fit, followed by the ICD-10. For DSM-5, we only applied the conditional exclusion if they had not engaged in aberrant behaviours. We found the DSM-5 did not perform as well 13

Chapter one: Introduction

as the ICD. Since the ICD-11 criteria have not been released as yet, this thesis uses the next best fit, and most commonly applied diagnosis internationally, the ICD-10.

Medication non-adherence and aberrant behaviours

There is no gold standard for measuring medication misuse events in people prescribed medication (Smith, Paillard, et al. 2015; Vowles et al. 2015). Many of the scales used in previous studies were developed for other purposes, such as diagnosis of substance use disorders, and have been found to be ‘imperfect’ in identifying medication non-adherence in people prescribed medications (Smith, Paillard, et al. 2015). Further, some measures are too lengthy to conduct in primary care, others unable to effectively monitor changes overtime and others with unspecified time frames (Larance et al. 2015). The Opioid Related Behaviours in Treatment scale (ORBIT) is a 10-item measure of medication non-adherence and aberrant behaviours and other forms of unsanctioned medication use (Larance et al. 2015). The ORBIT was developed to address the shortcomings of previous scales, with the aim of better identifying medication non-adherence and aberrant behaviours in people being prescribed opioid medications. (Larance et al. 2015). The ORBIT is used in this thesis to examine medication non-adherence and aberrant behaviours.

The prescribed opioids difficulties scale

Research has shown that among chronic pain patients prescribed opioids, the traditional ICD and DSM-IV diagnostic categories of dependence and harmful use/abuse, are distinct from many of the misuse problems that patients report (Banta- Green et al. 2009). Further, aberrant drug behaviours may not encompass the full range of problems that patients experience with their opioid use (Banta-Green et al. 2010). The Prescribed Opioids Difficulties Scale (PODS) is a patient-centred scale that assesses problems and concerns of the patient prescribed opioids for CNCP (Banta- Green et al. 2010). The intent of the scale is to identify patient problems, not to identify problematic patients and are distinct from opioid misuse and addiction

Chapter one: Introduction

(Sullivan 2010). This was included in the current thesis to examine problematic use form the patient’s perspective.

Chapter 5 will describe opioid use, including dose and medications, in the POINT cohort and problematic opioid use measures including ICD-10 criteria for pharmaceutical opioid dependence, the ORBIT for medication non-adherence and aberrant behaviours, and the PODS for patient-centred concerns and problems associated with opioids.

1.2 Suicidality and chronic non-cancer pain The increase in prescription opioid overdoses has led to a focus of problematic opioid use in CNCP (Dunn et al. 2010). Research, however, also suggests that some of these overdoses may be intentional (Darke, Duflou, and Torok 2011; Madadi and Persaud 2014; Roxburgh et al. 2011). Despite emerging evidence of a link between suicide and CNCP, this area of research has received far less attention (Cheatle 2014b). This next section will detail the current literature surrounding CNCP and associations with suicidality.

1.2.1 Defining suicide-related behaviours

Suicide and suicide-related behaviours have been inconsistently defined across research (Bridge, Goldstein, and Brent 2006; Nock et al. 2008; Silverman et al. 2007) and it is therefore important to define suicide, suicidal thoughts and behaviours used in this thesis. These are detailed below.

Suicide ideation refers to thoughts of engaging in behaviour intended to end one’s life. Although these can range from vague non-specific thoughts, this thesis is based on the definition of ‘serious thoughts about committing suicide’, and therefore, does not encompass vague thoughts of suicide (Silverman et al. 2007). Suicide plan refers to the specific formulation of a specific method through which one intends to die (Silverman 15

Chapter one: Introduction

et al. 2007). Suicide attempt refers to engagement in potentially self-injurious behaviour in which there is at least some intent to die, and is different to non-suicidal self-harm, i.e. where there is no intent to die (Silverman et al. 2007).

Suicide-related behaviours encompasses the terms defined above: suicidal ideation, suicide plans and suicide attempts. Although it has been suggested that the term suicide-related behaviours be used over the term suicidality (Silverman et al. 2007), for ease of reading, the current thesis uses the term suicidality.

1.2.2 Joiner’s Interpersonal Theory of Suicide

Although theory is not the focus of this thesis, it is important to our understanding of the elevated risk of suicide in CNCP patients. The prevailing model of suicide is Joiner’s Interpersonal Theory (IPT) of suicide (Joiner 2009; Van Orden et al. 2010). This theory posits that suicidal thoughts and desires arise from two unique and independent factors; thwarted belongingness and perceived burdensomeness. Thwarted belongingness refers to the unfulfilled need for social interaction and connectedness and is reflected in problems such as loneliness and withdrawal from others. Perceived burdensomeness is the perception of oneself as a burden or liability to others. These concepts, as related to CNCP have been supported in recent research. Self-perceived burden has been reported by over 70% of patients in a tertiary pain setting (Kanzler et al. 2012) and many CNCP report the experience of social withdrawal after the onset of their pain (Osborn and Smith 2015). Distress in interpersonal relationships (thwarted belongingness) and self-perceived burden on others (burdensomeness) have also been found to be significant predictors of suicidal ideation in people living with CNCP (Wilson et al. 2013).

Another aspect of Joiner’s IPT is acquired capability. A person must not have only the desire to suicide, but they must have the ability to do so. Hooley and colleagues (Hooley, Franklin, and Nock 2014) suggest that to acquire this capability a person must

Chapter one: Introduction

overcome their natural fear of death, which is acquired over time, through habituation to exposure to painful and provocative events. One aspect of this is developing a greater pain tolerance. Studies have found that those who had greater fearlessness of death also had greater pain tolerance during a pain pressure test and report that they are less fearful of severe pain, minor pain and medical pain (Hooley, Franklin, and Nock 2014).

Therefore, the ideas surrounding thwarted belongingness and perceived burdensomeness as a facilitation of suicidal desire, and the experience of physical pain as a factor in acquiring the capability to carry out the act of suicide, have been deemed plausible explanations in of suicidal behaviours in CNCP patients (Hooley, Franklin, and Nock 2014).

1.2.3 Prevalence of suicidality in chronic non-cancer pain

The association between suicidality and CNCP is, now, well established (Fishbain, Lewis, and Gao 2014). This association has been documented in recent epidemiological studies in countries, including, Canada (Ratcliffe et al. 2008), the United States (Braden and Sullivan 2008; Ilgen et al. 2008) and a large cross-national study of developed and developing countries (Scott et al. 2010). A systematic review found that people suffering from chronic pain had two-three times the risk of engaging in suicidal thoughts or behaviours, and almost double the risk of death by suicide, compared with the general population (Tang and Crane 2006). Rates of lifetime ideation range from 21% to 50%, and current ideation ranged from 0% to 24% amongst people with chronic pain and rates of lifetime suicidal attempts range from 5% to 14% (Tang and Crane 2006). Furthermore, chronic medical illness has been identified as a motivating factor in approximately 25% of all suicides (Moscicki 1997).

Chapter one: Introduction

The paper presented in Chapter 2 is the first to examine CNCP and associations with suicidality in the Australian population and is the first study to estimate the contribution of CNCP to overall suicide statistics.

1.2.4 General and pain-specific suicide risk factors

Although the link between suicide and CNCP has been generally recognised (Calati et al. 2015), less is understood about the specific risk factors. It is unclear if the same risk factors for suicide in the general population are independently associated with suicide in people living with chronic pain (Tang and Crane 2006). Researchers Fishbain (Fishbain 1999), Tang and Crane (Tang and Crane 2006) and Hassatt (Hassett, Aquino, and Ilgen 2014) have attempted to document the general and pain-specific factors associated with suicide in CNCP patients. General suicide risk factors, identified in people living with CNCP, include: gender, unemployment, social isolation, history of childhood abuse, previous history of suicide attempt, mental health comorbidities, history of substance abuse, hopelessness and access to weapons/substances. Pain- specific risk factors include: pain severity, pain location/type, pain duration, pain- related helplessness, desire to escape from pain, pain-related cognitions, burdensomeness, access to analgesics and poor quality sleep. Those relevant and examined in the current thesis are detailed below.

1.3.4.1 General suicide risk factors in CNCP patients

Socio-demographics

In the general population, females have been found to have higher rates of suicidal ideation and suicide attempts compared with males (Nock et al. 2008; Weissman et al. 1999), although males are overrepresented in suicide mortality figures (Mościcki 1994). Interestingly, this has not been found consistently amongst people living with CNCP. In CNCP research, some studies have found that females were more likely to suicide than males (Timonen et al. 2003), or there were no gender differences in suicidal ideation, plans or attempts (Edwards et al. 2006; Smith et al. 2004). Although

Chapter one: Introduction

these results need to be interpreted with caution, due to the over-representation of females in CNCP and a lack of power in the studies, it is also probable that gender as a suicide risk factor is less important in CNCP patients.

As detailed previously in Section 1.2.3, CNCP can have a major impact on an individual’s employment status. Losing a job, home foreclosure and financial uncertainty can lead to an increased risk of suicide through comorbidity with other risk factors such as depression, anxiety and the harmful use of alcohol (Nordt et al. 2015). Amongst people living with CNCP, being unable to return to work or previous levels of work, the impact of unemployment can be significant on the individual (Breivik et al. 2006) and can be a factors in the engagement in suicide-related behaviours.

Childhood abuse

Previous literature has found childhood abuse and neglect to be associated with chronic pain (Davis, Luecken, and Zautra 2005) and engagement in suicidal behaviours (Dube et al. 2001). Amongst people living with CNCP, childhood abuse has been found to be significantly associated with suicidal ideation, though not a suicide attempt (Okifuji and Benham 2011). On the other hand, a childhood abuse history in CNCP patients prescribed opioids was associated with a lifetime suicidal attempt (Balousek, Plane, and Fleming 2007), and a risk factor for suicide (Cheatle et al. 2014). Again, due to the limited research in this area, it is difficult to understand the extent of this relationship and more research is necessary to understand this association.

Psychiatric comorbidity

Mental disorders have been found to be present in approximately 90% of people who die by suicide (Cavanagh et al. 2003). The presence of depression has consistently been demonstrated to be associated with an elevated risk of suicide (Harris and Barraclough 1997; Nock et al. 2008). In CNCP patients, however, although some research has shown a link between depression and suicide in CNCP patients (Breslau 1992; Fisher et

Chapter one: Introduction

al. 2001; Timonen et al. 2003), other research has been less clear on the link (Smith et al. 2004). Epidemiological studies have found physical conditions (Scott et al. 2010) and chronic pain conditions (Braden and Sullivan 2008; Ilgen et al. 2008; Ratcliffe et al. 2008) to be a risk factor for suicidal behaviours, over and above mental health disorders. The presence of depression has consistently been associated with suicide in the general literature, the findings that CNCP is associated with suicidality over mental health has major clinical implications in identifying and treating those at risk. It is important that research is able to delineate this association.

Substance use disorders have been found to be associated with suicide in the general suicide literature. Alcohol and other substance use disorders are found in 25−50% of all suicides (Schneider 2009), and suicide risk is further increased if alcohol or substance use is comorbid with other psychiatric disorders (World Health Organization 2014). Amongst people living with CNCP, emerging research suggests that people with a history of substance use problems are at a heightened risk for suicidal ideation (Racine, Choiniere, and Nielson 2014). It has been suggested that this has implications for the potential misuse of prescription opioids (Hassett, Aquino, and Ilgen 2014), and warrants further investigation.

1.3.4.2 Pain-specific suicide risk factors in CNCP patients

Fishbain (Fishbain 1999), Tang and Crane (Tang and Crane 2006) and Hassatt (Hassett, Aquino, and Ilgen 2014) have identified pain-specific factors associated with suicidality in CNCP patients and are detailed below. As burdensomeness was discussed above under Joiner’s IPT, it will not be repeated.

Pain location and type of pain

Specific locations and types of pain have been found to be associated with suicidality. Presence of abdominal pain has been associated with an increased risk of suicidality (Spiegel, Schoenfeld, and Naliboff 2007). Neuropathic pain was associated with a

Chapter one: Introduction

reduced risk of active and passive suicidal thoughts (Smith et al. 2004), compared with those without neuropathic pain.

CNCP has also been found to be associated with suicidality, over and above mental health factors. In three large national studies, headache and/or migraine pain was found to be independently associated with suicidal ideation and suicide attempts after controlling for demographics, mental health and other physical problems (Braden and Sullivan 2008; Ilgen, Kleinberg, et al. 2013b; Ratcliffe et al. 2008). Back pain has also been found to be independently associated with suicidality (Ilgen, Kleinberg, et al. 2013b). Further, although some studies find an independent association between fibromyalgia and suicide (Wolfe et al. 2011) other studies have found no independent association (Jimenez-Rodríguez et al.). Arthritis is the one pain condition that has consistently been found to be not associated with suicidality (Braden and Sullivan 2008; Ilgen, Kleinberg, et al. 2013b; Ratcliffe et al. 2008). These findings suggest a link between pain conditions and suicidality that needs to be better understood, importantly the independent contribution of certain pain conditions to suicide, over and above other risk factors such as demographics and mental health.

Pain severity

The association of pain severity and suicidality is inconsistent in the literature. In a recent large study of Veteran’s affairs medical records (Ilgen et al. 2010), after controlling for demographic and psychiatric characteristics, severe pain was associated with suicide compared with those experiencing none, mild or moderate pain. In a study of tertiary care pain patients, however, pain severity was found not to be associated with suicidality in adjusted models (Smith et al. 2004). Due to limited research in this area, more research is needed to determine the association between pain severity and suicidality.

Chapter one: Introduction

Pain duration

Again, there have been conflicting studies relating to pain duration and its association with suicidality. Although a positive relationship between pain duration and suicidality was found in one study of outpatient pain patients (Hinkley and Jaremko 1994), another study of patients in a tertiary pain setting found no link between suicidality and pain duration (Smith et al. 2004).

Pain-related helplessness and desire to escape from pain

Hopelessness (the expectation that desirable outcomes will never occur) and a desire to escape from pain, although not specifically researched in CNCP patients to date, have been suggested by Tang and Crane (Tang and Crane 2006) as possible motivating factors for developing suicidal ideation and behaviours.

Pain-related cognitions

Pain catastrophizing has been conceptualized as an exaggerated, negative focus on pain that can contribute to depression, pain intensity, and disability (Tang and Crane 2006). Pain self-efficacy refers to a person’s confidence in their ability to control the pain (Somers et al. 2012). Although Tang and Crane (Tang and Crane 2006) acknowledge that pain-related cognitions are possible important factors in understanding the relationship between CNCP and suicidality, relatively few studies had examined this at the time of their review (Racine, Choiniere, and Nielson 2014). In recent years the concepts of pain catastrophizing and pain self-efficacy have been found to be important factors in understanding this relationship. In one study pain catastrophizing was associated with past overdose and lifetime suicide attempt (Sansone, Watts, and Wiederman 2014). Another found pain-related catastrophizing, along with depressive symptoms, were the two most consistent predictors of the presence and degree of suicidal ideation among patients seeking treatment for CNCP (Edwards et al. 2006). Further, although pain self-efficacy has not been studied in relation to CNCP and suicidality, a study has found that both pain catastrophizing and

Chapter one: Introduction

pain self-efficacy were significantly related to physical symptoms and psychological distress (Somers et al. 2012). Although there has been little research on the association of pain-cognition factors and suicidality, since pain coping factors can easily be modified using psychological intervention much more research is necessary in this area (Somers et al. 2012).

Access to analgesics

As mentioned previously, there has been an increase in prescriptions of opioids specifically related to CNCP patients. A number of recent studies have characterised opioid overdose (intentional and unintentional) decedents (Darke, Duflou, and Torok 2011; Madadi and Persaud 2014; Roxburgh et al. 2011) and found that there are a significant proportion of decedents with a history of CNCP, prescribed opioids. It has been suggested that in patients at risk of suicide, alternative pain treatments and medications should be utilised (Madadi and Persaud 2014).

Poor quality sleep

A final pain-related risk factor that has been identified in the literature is poor quality sleep. CNCP patients often suffer from disturbed sleep which impacts on their daily functioning (McCracken and Iverson 2001; Menefee et al. 2000; Morin, Gibson, and Wade 1998). Sleep problems, such as sleep onset insomnia, have been found to be associated with suicidal ideation, independently of depressive symptoms in a sample of outpatient CNCP patients (Smith and Haythornthwaite 2004). Further, one study of CNCP patients found that sleep quality was an independent correlate of suicidal ideation over and above other physical health problems such as disability, pain duration and severity (Racine, Choiniere, and Nielson 2014).

As can be seen, our understanding of the risk factors associated with suicidality in CNCP is less understood. As mentioned, Chapter 2 will examine the association of CNCP and suicidality at the population level, and will also determine the contribution

Chapter one: Introduction

of CNCP to suicide statistics. Chapter 7 will examine the influence of previously identified general and pain-specific risk factors in a large, national community based sample of people prescribed opioids for CNCP.

1.3 Borderline personality disorder As seen in Section 1.1.2.1 mental health problems are highly prevalent in people living with CNCP. Although there has been much attention in the pain literature focused upon affective and anxiety disorders in people living with CNCP (Asmundson and Katz 2009; Fishbain et al. 1997; Gatchel 2004), there has been much less research on personality disorders, such as anti-social personality disorder (ASPD) and BPD. In a epidemiological study, the presence of ASPD and BPD was found to be elevated in people living with CNCP (Braden and Sullivan 2008). A recent epidemiologic study found that people with CNCP were more likely than the general population to have greater symptoms of BPD (McWilliams and Higgins 2013). Further, both chronic pain and BPD are frequently encountered in primary care (Gross et al. 2002; Gureje et al. 1998; Blyth et al. 2004). Although a strong association between BPD and substance misuse (Darke et al. 2004), suicidality (Black et al. 2004) and chronic pain (Sansone and Sansone 2007, 2012; Sansone et al. 2001) has been well documented individually, much less is known about the inter-relationships of these factors. The last chapter of the thesis will examine the prevalence of BPD among CNCP patients and associations with problematic opioid use and suicidality.

1.3.1 Defining Borderline Personality Disorder

Borderline Personality Disorder (BPD) is characterised by a longstanding, pervasive pattern of affective dysregulation, identity disturbance, and interpersonal dysfunction, and is associated with high levels of distress and suffering in the individual, as well as difficulties in functioning (Gunderson 2009). The main diagnostic themes for BPD include (American Psychiatric Association 2001);

• Unstable and intense interpersonal relationships

Chapter one: Introduction

• Recurrent suicidal threat, gestures or self-mutilation • Tendency to act impulsively without consideration of consequences • Fears of abandonment • Chronic feelings of emptiness • Unstable moods • Inappropriate or difficulty in controlling anger • Disturbances in identity.

1.3.2 Prevalence of Borderline Personality Disorder in chronic non-cancer pain

The estimated prevalence of BPD in the general population is 2% (Jackson and Burgess 2000). People with BPD are often high users of health services and access general practitioners and specialists more frequently than those without BPD. In primary care settings the prevalence of BPD is estimated at around 6% (Gross et al. 2002). Between 1994 and 2011 eight studies examined the relationship between BPD and chronic pain (Sansone and Sansone 2012). Averaging the prevalence rates across these studies, the prevalence of BPD in patients with chronic pain has been estimated at 30% (Sansone and Sansone 2012). People living with CNCP, who also have BPD, have been found to present more frequently to pain clinics or present with more pain problems in primary care, than those without BPD (Kinney et al. 1993; Sansone et al. 2001). In general population samples, after adjusting for sociodemographic, and other mental health disorders, the presence of BPD was significantly associated with a number of health conditions, including arthritis (El-Gabalawy, Katz, and Sareen 2010). More specifically, studies based on the US National Comorbidity Survey Replication found people living with CNCP were more likely to screen positive for BPD (Braden and Sullivan 2008) and BPD symptoms were positively associated with chronic back/neck problems, frequent headaches and other chronic pain conditions (McWilliams and Higgins 2013).

Chapter one: Introduction

1.3.3 Borderline Personality Disorder and chronic non-cancer pain

Although self-injurious behaviour is common amongst people with BPD, 50-80% report no pain associated with this behaviour (Bohus et al. 2000). On the other hand, BPD is over-represented in patients with chronic pain. Sansone refers to this as the ‘pain paradox’ (Sansone and Sansone 2007), and suggests that chronic pain, as opposed to self-injurious behaviours, is not under the individual’s control, and thus, poorly tolerated in people with BPD. It has been suggested that BPD and pain conditions may be associated because of a shared vulnerability involving difficulties in self-regulation (Sansone and Sansone 2007). Personality disorders may also reflect non-specific deficits in coping skills and ability to cope with pain (Gatchel 2004). It has also been suggested that chronic pain symptoms may consciously or unconsciously be pronounced, in an attempt to elicit a caring response from others (Sansone and Sansone 2012). Sansone suggests that “from a psychiatric perspective, chronic pain syndromes meld well with the dynamics of BPD including, for example, the need for frequent appointments (i.e., the attachment dynamics), role of ongoing debilitation (i.e., victim dynamics), extensive use of medications (i.e., oral self-regulation difficulties), and dependency issues” (Sansone et al. 2011, pp. 196). Further, a history of trauma is has been found to be elevated in both CNCP and BPD (Sansone, Whitecar, and Wiederman 2009).

1.3.4 Borderline Personality Disorder and problematic opioid use in chronic non-

cancer pain

Substance use disorders are common in patients with borderline personality disorder (American Psychiatric Association 2001; Dulit et al. 1990). Persons with borderline personality disorder often use substances in an impulsive fashion (Maloney et al. 2009), as is suggested by BPD definitions.

Although there have been very few studies, the limited research suggests that CNCP patients with BPD may over-utilise their opioid medication in an effort to cope with

Chapter one: Introduction

chronic pain (Sansone and Sansone 2007). Frakenburg and Zanarini (Frankenburg and Zanarini 2004) found those with current BPD were more likely to have sustained use of pain medicines and daily use of sleeping medications. Further, in a 10-year follow-up of the study, the authors found that BPD was significantly associated with prescription medication use, compared with those diagnosed with other mental health disorders (Frankenburg, Fitzmaurice, and Zanarini 2014). More specifically, in a study of opioid abuse in people living with CNCP, personality disorders, along with panic attacks and trait anxiety were found to be associated with prescription opioid abuse and accounted for 38% of the variance (Wilsey et al. 2008). These limited findings suggest the need to further understand this relationship of CNCP, BPD and problematic opioid use.

1.3.5 Borderline Personality Disorder and suicidality in chronic non-cancer pain

Suicidal threats, gestures, and attempts are very common among patients with borderline personality disorder. At least three-quarters of BPD patients attempt suicide, and approximately 10% eventually complete suicide (Black et al. 2004). This rate is approximately 400 times greater than the general population (Oumaya et al. 2008). As mentioned previously, there are high rates of suicidality in CNCP patients. To date, however, there have been studies investigating the associations between CNCP, BPD and suicidality. With such high rates of suicidality in patients with chronic pain and patients with BPD, the association between BPD, chronic pain and suicidality warrants investigation.

1.4 Limitations of previous research Although previous research clearly demonstrates the importance of examining problematic opioid use, suicidality and BPD in people living with CNCP, it has been limited due to a number of factors. These are detailed below.

Chapter one: Introduction

1.4.1 Problematic opioid use in people with chronic non-cancer pain

Chou, et al. (Chou, Fanciullo, Fine, Miaskowski, et al. 2009) suggest there is a need for high-quality evidence to help guide decisions regarding patient selection and monitoring during opioid therapy. Previous studies, that have examined problematic opioid use in people prescribed opioids for CNCP, have generally been based on clinical samples, RCT’s, or from pain clinics. Clinical trials typically find far lower rates of aberrant drug-related behaviours and opioid dependence than have been reported in some observational studies (Fishbain, Cole, Lewis, Rosomoff, and Rosomoff 2008). This is a consequence of many of these studies often excluding more complex patients, and therefore those at most risk of developing problems. There is a need for studies that include patients who are potentially at higher risk of adverse outcomes to determine the actual prevalence of problematic prescription opioid use in people living with CNCP. There is also a need for information on the benefits and harms of higher doses of opioids (Chou, Ballantyne, et al. 2009; Chou et al. 2015).

Additionally, there are a number of measurement problems in identifying problematic opioid use in people prescribed opioids for CNCP, as mentioned previously, the controversy around defining dependence, and determining medication misuse and non-adherence (Banta-Green et al. 2009; Sees and Clark 1993; Smith, Paillard, et al. 2015; Sullivan 2010). It is difficult to determine actual rates of problematic use due to these inconsistencies in measurement. Furthermore, many studies only focus on one aspect of problematic use. Problematic opioid use can be seen in problems associated with dependence, medication non-adherence, and patient-centred problems and concerns. In order to understand the extent of problematic use, it is important to measure these different aspects.

In this current thesis a large, national community sample of people prescribed opioids for CNCP will be utilised to examine problematic prescription opioid use. Further, problematic opioid use will be measured in a number of ways: using the gold standard for determining substance dependence, the ICD-10, using the ORBIT to examine 28

Chapter one: Introduction

medication non-adherence and aberrant behaviours and the PODS to determine patients-centred problems and concerns.

1.4.2 Suicidality in people living with chronic non-cancer pain

The association between suicidality and CNCP is now widely accepted (World Health Organization 2014), and although general and pain specific risk factors have been identified, research testing and examining these factors has been limited (Hassett, Aquino, and Ilgen 2014; Tang and Crane 2006). There have been a number of nationally representative studies examining the suicide risk in people living with CNCP (Braden and Sullivan 2008; Ilgen et al. 2008; Ratcliffe et al. 2008; Scott et al. 2010), and although these has been essential in understanding the association of CNCP and suicidality at the population level, many of the surveys are based on national mental health data collection systems, that do not collect pain-specific information. Examination of pain-specific risk factors at the population level, therefore, have been limited. Other more detailed studies have involved small clinical populations where there is insufficient power to examine the range of general and pain-specific suicide risk factors (Edwards et al. 2006; Okifuji and Benham 2011; Racine, Choiniere, and Nielson 2014; Smith et al. 2004; Smith and Haythornthwaite 2004). Further, although there have been other larger studies, they have generally focused on a specific group of people, for example veterans (Ilgen, Kleinberg, et al. 2013a; Ilgen et al. 2010).

This current thesis will make use of a large nationally representative survey to examine the prevalence and associations of suicidality and CNCP at the population level to address the lack of research on suicidality and CNCP in Australia. Further, using a large community sample of people prescribed opioids for CNCP, this thesis will examine the association and importance of previously identified general and pain-specific risk factors for suicidality in CNCP. Finally, an ideation-to-action framework will be utilised, for the first time in a CNCP sample, to examine factors associated with progression from suicide ideation to suicide attempt.

Chapter one: Introduction

1.4.3 Borderline Personality Disorder and chronic non-cancer pain

Although there appears to be associations between BPD, chronic pain, problematic opioid use and suicidality, there have been very few studies that have examined these areas concurrently. Further, although there have been large nationally representative surveys examining associations between CNCP and BPD at the population-level, research examining BPD and problematic opioid use have been few and based on small, clinical samples (Sansone and Sansone 2012). Additionally, the association between CNCP, BPD and suicidality has not previously been examined. With emerging research suggesting a higher prevalence of BPD in people living with CNCP, it is important that the associations between BPD and suicidality and problematic opioid use be examined in people living with CNCP. The current thesis will use a large community sample of people prescribed opioids for CNCP to examine these associations.

1.5 Aims of thesis The specific aims of the thesis are;

1. Estimate the prevalence and associations of chronic non-cancer pain and suicidality in the Australia population (paper one); 2. Describe the Pain IN Opioids Treatment (POINT) study, a large national community sample of people prescribed opioids for CNCP (paper two); 3. Examine the sociodemographic and clinical characteristics of the POINT cohort (paper three); 4. Examine sociodemographic and clinical characteristics associated with prescription opioid dose and problematic opioid use in a large community sample of people prescribed opioids for CNCP (paper four); 5. Determine the prevalence of suicidality in people prescribed opioids for CNCP and the contribution of general and pain-specific risk factors (paper five); and 6. Explore the association of screening positive for BPD in CNCP prescribed pharmaceutical opioids with problematic pharmaceutical opioid use and suicidality (paper six). 30

1.6 Overview of thesis structure

Chronic non-cancer pain and suicidality in the Australian population 2. Paper one: The prevalence and correlates of chronic pain and suicidality in a nationally representative sample

Overview of the Pain and Opioids IN Treatment Study 3. Paper two: Cohort Protocol paper: The Pain and Opioids IN Treatment study

Baseline characteristics of the POINT cohort 4. Paper three: The Pain and Opioids IN Treatment (POINT) study: Characteristics of a cohort using opioids to manage chronic non-cancer pain

Pharmaceutical opioid use in the POINT cohort Suicidality in the POINT cohort 5. Paper four: Pharmaceutical opioid use and dependence among 6. Paper five: Prevalence and correlates of suicidal thoughts and people living with chronic pain: Associations observed within the suicide attempts in people prescribed pharmaceutical opioids Pain and Opioids IN Treatment (POINT) cohort for chronic pain

Borderline personality and problematic opioid use and suicidality in the POINT cohort 7. Paper six: Associations of borderline personality with pain, problems with medications and suicidality in a community sample of chronic non-cancer pain patients prescribed opioids for pain

Discussion and conclusions 8. Summary of finding and recommendations for future research and practise 9. Conclusions Chapter one: Introduction

1.6.1 Thesis outline

This thesis comprises nine chapters. Each of the chapters are detailed below.

Chapter 2 (paper one) will provide prevalence estimates of CNCP and suicidality in the Australian population using the National Survey of Mental Health and Wellbeing (NSMHWB). Associations of CNCP will be examined, as will risk-factors of suicidality in people living with CNCP. The study will also estimate the contribution of CNCP factors to suicide-related behaviour statistics.

Chapter 3 (paper two) describes the Pain and Opioids IN Treatment (POINT) study methodology, a national study of community pain patients prescribed opioids, on which the remainder of the thesis will be based.

Chapter 4 (paper three) provides an overview of the baseline characteristics of the POINT cohort. Due to the strong association between employment and CNCP, examination of these characteristics will based on age-groups in terms of their employment status, i.e. prime working age, nearing retirement and retirement.

Chapter 5 (paper four) describes pharmaceutical opioid use in the POINT cohort. Oral morphine equivalents are examined in terms of their association to socio- demographics, pain related factors, physical health, mental health and problematic opioid use. Pharmaceutical opioid dependence in the POINT sample will be examined using the International Classification Diagnostic system Version 10 (ICD-10), including determining independent associations between patient characteristics and dependence.

Chapter 6 (paper five) examines the prevalence and characteristics associated with suicidality in the POINT sample. The importance of general and pain-specific risk

Chapter one: Introduction

factors will examined and for the first time in a CNCP sample a novel approach of examining factors that led to progression of suicide ideation to suicide attempt will be employed.

Chapter 7 (paper six) will examine the prevalence of screening positive to BPD in the POINT sample and independent association with pain-related factors, problematic opioid use and suicidality.

Chapter 8 will provide a summary of the main key findings of the thesis, these include; socio-demographics and clinical characteristics of people living with CNCP, suicidality and CNCP, problematic opioid use and problematic use in people with CNCP and the association between CNCP, BPD, problematic opioid use and suicidality in people living with CNCP. Recommendations will be made including; need for multidisciplinary treatment, the use of screening tools to reduce adverse events such as problematic opioid use and suicidality, and careful monitoring of medications and treatment progress. Strengths and limitations of the thesis will be discussed.

Chapter 9 will provide a conclusion for the thesis.

Chapter two: Prevalence of CNCP and suicidality in Australia

2 PAPER ONE: THE PREVALENCE AND CORRELATES OF CHRONIC PAIN AND SUICIDALITY

IN A NATIONALLY REPRESENTATIVE SAMPLE

Gabrielle Campbell1, Shane Darke1, Raimondo Bruno2 and Louisa Degenhardt1

1. National Drug and Alcohol Research Centre, UNSW, Australia 2. School of Medicine, University of Tasmania, Australia

Paper one has been published in the Australian New Zealand Journal of Psychiatry (Campbell, Darke, et al. 2015)

Chapter two: Prevalence of CNCP and suicidality in Australia

2.1 Copyright Statement I certify that this publication was a direct result of my research towards this PhD, and that reproduction in this thesis does not breach copyright regulations.

Campbell, G., Darke, S., Bruno, R., & Degenhardt, L. (2015). The prevalence and correlates of chronic pain and suicidality in a nationally representative sample. Australian and New Zealand Journal of Psychiatry 49(9) 803-11.

Gabrielle Campbell November, 2015

Chapter two: Prevalence of CNCP and suicidality in Australia

2.2 Preamble Although there have been a number of international studies examining CNCP and suicidality at the population level (Braden and Sullivan 2008; Ilgen et al. 2008; Ratcliffe et al. 2008), there have been no studies in the Australian population. Further, the last prevalence estimates of CNCP in Australia were from 2001 (Blyth et al. 2001) and with an ageing population these would have likely increased. The paper presented in this Chapter uses the National Survey of Mental Health and Wellbeing (NSMHWB) from 2007, to estimate the population prevalence of chronic pain problems in the Australian population, and the association between suicidality and chronic pain in the Australian population. This study is one of the first to examine the prevalence of CNCP in people who engage in suicide-related behaviours. The paper presented in this Chapter provides the overall context on which the rest of this thesis is based.

Chapter two: Prevalence of CNCP and suicidality in Australia

2.3 Abstract Background: Research suggests that people suffering from chronic pain have elevated rates of suicidality. With an ageing population, more research is essential to gain a better understanding of this association.

Aims: To document the prevalence and correlates of chronic pain and suicide, and estimate the contribution of chronic pain to suicidality.

Method: Data from the 2007 Australian National Survey of Mental Health and Wellbeing, a nationally representative household survey on 8,841 people, aged 16–85 years, was analysed.

Results: The odds of lifetime and past 12-month suicidality were two to three times greater in people with chronic pain. Sixty-five percent of people who attempted suicide in the past 12-months had a history of chronic pain. Chronic pain was independently associated with lifetime suicidality after controlling for demographic, mental health and substance use disorders.

Conclusions: Health care professionals need to be aware of the risk of suicidality in patients with chronic pain, even in the absence of mental health problems.

Chapter two: Prevalence of CNCP and suicidality in Australia

2.4 Introduction Chronic pain, defined as pain that persists for greater than three months (Merskey and Bogduk 1994), is a common worldwide complaint, with rates in the general population estimated from 17% (Blyth et al. 2001) to 41% (Tsang et al. 2008). Chronic pain is associated with poorer physical health (Eriksen et al. 2006), poorer quality of life (Hagg, Fritzell, and Nordwall 2003; Ojala et al. 2014) and a greater risk of developing depression (Currie and Wang 2004) and other mental health problems (Dersh, Polatin, and Gatchel 2002; McWilliams, Cox, and Enns 2003). Although there are behavioural, physical and medication interventions for chronic pain, it is often the case that many patients have to accept a lifetime treatment plan for the management of their pain, which may not alleviate the pain completely (Breivik et al. 2006).

It is not unexpected then, that rates of suicidality (defined as ideation, plans and attempts) are elevated amongst this group (Meilman 1984). Emerging research suggests that people suffering from chronic pain have almost double the risk of death by suicide (Tang and Crane 2006). Amongst people with chronic pain, rates of lifetime ideation range from 21% to 50%, current ideation from 0% to 24%, and rates of lifetime suicidal attempts from 5% to 14% (Tang and Crane 2006). These studies, however, have generally been based on small, clinical samples, so it is unknown if the same pattern exists in the general pain population. Further, recent epidemiological research suggests chronic pain is independently associated with suicidality after controlling for demographic and mental health factors (Braden and Sullivan 2008; Ilgen et al. 2008; Ratcliffe et al. 2008).

With an ageing population, and the likely increase in chronic pain, the issue of chronic pain and its relationship to suicidality is of timely importance. To our knowledge, there has been no study that has estimated the contribution that chronic pain may make to suicidality in the general population, and there are few existing data on the relationship between chronic pain and suicide in the Australian population. The aims of the current study were to:

Chapter two: Prevalence of CNCP and suicidality in Australia

1. document the prevalence of chronic pain in Australia; 2. examine lifetime and past 12-month suicidal behaviours in people with chronic pain, and estimate the contribution that a history of chronic pain may make to suicidality in the general population; 3. determine whether chronic pain is associated with suicide when controlling for socio-demographic variables and mental health diagnoses; and, 4. determine the risk factors for the past 12-month suicidal behaviours amongst people with chronic pain.

2.5 Method 2.5.1 Sample

Data from the Confidentialised Unit Record File (CURF) for the Australian 2007 National Survey of Mental Health and Wellbeing (NSMHWB) (Slade et al. 2009) were obtained from the Australian Bureau of Statistics (ABS). The 2007 NSMHWB was conducted by trained interviewers from the ABS between August and December of 2007. The aim of the survey was to provide representative estimates on a national level. The 2007 NSMHWB sample was based on a stratified, multistage area probability sample of people aged 16–85 years in rural and urban private dwellings across all Australian States and Territories. The area-based selection ensured that all sections of the population living in private dwelling within the geographical scope of the survey were represented, representing a total of 17,352 dwellings. Special dwellings (hospitals, nursing homes, gaols, hotels) and dwellings in remote and sparsely settled regions of the country were not sampled, which reduced the effective sample size to 14,805 dwellings. All interviews were conducted in English, meaning that non-English speakers were excluded from the sample. Each interview was conducted in person and took approximately 90 minutes to complete

There were 8,841 fully responding participants, which represented a rate of 60%. Owing to the lower than expected response rate (it was expected that a 75% response rate would be achieved), extensive non-response analyses were undertaken to assess

Chapter two: Prevalence of CNCP and suicidality in Australia

the reliability of the survey estimates. A purposive sample of fully non-responding households in two metropolitan areas, Sydney and Perth, was followed up with a short-form interview, to gain a qualitative assessment of non-response bias. The study indicated the impact of non-response to be small at the aggregate level, but possible underestimation of the prevalence of mental disorders may exist for men, young persons and the Perth population. The final sample represented an estimated population count (EPC) of 16,015,000 Australian adults. For further information about the procedure and instrument see Slade, et al. (Slade et al. 2009).

2.5.2 Survey

The 2007 NSMHWB was based on the World Mental Health Survey Initiative version of the Composite International Diagnostic Interview (WMH-CIDI) (Kessler and Üstün 2004). This interview has undergone extensive methodological development and testing and has been used in household surveys in at least 28 countries around the world. This version of the instrument collects both diagnostic information on a broad range of mental disorders and information on risk factors, impacts, patterns and treatment of mental disorders. There were some modifications made to the WMH-CIDI to tailor the survey to the Australian context. The following questions and measures from the 2007 NSMHWB were used in the current study.

2.5.2.1 Chronic pain

The chronic conditions section was based on the WMH-CIDI chronic conditions module. The information provided is self-reported by the respondent and is not verified by a medical practitioner. The current study used questions that related to the presence of arthritis, migraines and back/neck pain that had persisted for a period of 6 months. A combined category of ‘any pain’ was created on the basis of endorsement of any one chronic pain condition (arthritis/migraines, back/neck problems).

Chapter two: Prevalence of CNCP and suicidality in Australia

2.5.2.2 Socio-demographics

Consistent with Ratcliff et al. (Ratcliffe et al. 2008), age was split into categories of <25 years, 25–44 years, 45–64 years and 65 years, and was dichotomised into above and below 60 years of age when examining⩾ bivariate and multivariate relationships. Marital status was dichotomised to married/de facto, with widowed/separated/single/ divorced as the reference category. A dichotomous variable was created whereby those who had completed high school were deemed to have completed school. Not completing school was the reference category. Employment was dichotomised: not employed or not in the labour force was the reference category

2.5.2.3 Mental health and substance use

The WMH-CIDI produces diagnostic information according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) and the International Classification of Diseases, 10th Revision (ICD-10). Only DSM-IV diagnoses were reported here. The WMH-CIDI uses a lifetime time frame, with questions on experiences of symptoms in the previous 12-months in combination with lifetime diagnoses used to establish 12-month diagnoses. Diagnoses were derived using standard WMH-CIDI diagnostic algorithms. DSMIV lifetime and past 12-month diagnoses were assessed. The following diagnoses were used in the current study based on previous studies (Braden and Sullivan 2008): major depression (MD), general anxiety disorder (GAD), post-traumatic stress disorder (PTSD), alcohol use disorders (AUD), drug use disorders (DUD). AUD and DUD were those that met criteria for abuse and/or dependence.

Suicidal behaviour

The 2007 NSMHWB suicide questions were adopted from the WMH-CIDI without modification. Participants were asked about suicidal behaviour in their lifetime and in the 12-months prior to interview. Suicidal behaviour includes: ideation (serious thoughts about committing suicide), plans and attempts. Lifetime and past 12-month

Chapter two: Prevalence of CNCP and suicidality in Australia

suicidality were dichotomously coded with a code of ‘1’ representing a positive response to the relevant experience.

For the past 12-month risk factors, due to small numbers, correlates of past 12-month suicidal behaviours in people with chronic pain were again dichotomously coded, so that endorsement of one suicidal behaviour in the past 12-months was coded as ‘1’, representing at least one suicidal behaviour (thought, plan, or attempt) in the preceding 12-months. Although there may be concerns grouping these behaviours together, due to small numbers of the past 12-month suicide attempt, and that 60% of suicide attempts occur within a year of the onset of suicidal ideation (Nock et al. 2010), we believe that for the purposes of this exploratory study the grouping is deemed sufficient.

2.5.2.4 Physical health

A dichotomous variable was created from the global question ‘In general, would you say your health is excellent, very good, good, fair or poor?’, as those who endorsed poor or fair health, with excellent/very good/good as the reference category.

2.5.3 Data

2.5.3.1 Weights

Person weights and replicate weights were generated by the ABS and obtained from the CURF (Australian Bureau of Statistics 2009b). Person weights were generated to reflect the characteristics of the total population in 2007 NSMHWB. Person weights were based on the following: first, to reflect the inverse probability of being selected for survey participation. Second, weights were calibrated to benchmarks for state, part of state, age and sex. In this case, benchmarks were based on the ABS Estimated Resident Population (ERP) for June 2007, which were projected from the 2006 Census of Population and Housing (Australian Bureau of Statistics 2008). Third, calibration for

Chapter two: Prevalence of CNCP and suicidality in Australia

household composition, educational attainment and labour force status was performed (Australian Bureau of Statistics 2009a).

Replicate weights were produced using the delete-a-group jack-knife method (Rust and Rao 1996; Australian Bureau of Statistics 2009a). This involved division of the sample into 60 non-overlapping subsamples. Each was produced following the selection procedures used to create the full sample. One at a time each subsample is dropped and the procedures applied to produce the person weights are applied to the remaining sample. When applied in statistical analysis procedures, the replication approach produces estimates for the parameter of interest using each replicate weight. The variability among these estimates is used to estimate the sampling error for the whole sample estimate.

2.5.3.2 Data analysis

All analyses were based on weighted data, controlling for stratification and clustering, and were conducted using STATA, version 12.0 (StataCorp. 2011; College Station, TX: StataCorp LP). The application of the weights ensured the data were representative of the national population. In the descriptive analyses, statistical differences were assessed using odds ratios (ORs) and 95% confidence intervals (CIs). Owing to the overlap of pain conditions, a general ‘any pain’ category was used in some of the analyses. Logistic regressions were used to estimate the ORs for suicidal behaviours (i.e. thoughts, plans and attempts) while controlling sequentially for socio- demographic, mental health and chronic pain covariates. Covariates in the first set of regressions were: age, sex, married/de facto, completed school and employed (demographics). Covariates in the second model were: demographics and major depression, GAD, PTSD, AUD and DUD (mental health). Covariates in the third model were: demographics, mental health and the presence of one or more of the other pain conditions. Although the survey data have been weighted by age and sex, they have been included in the regressions due to their documented association with suicidality. The covariates and order in which they were added were determined by previous

Chapter two: Prevalence of CNCP and suicidality in Australia

research on known factors for suicide. It was expected that mental health would account for the biggest risk.

A logistic regression was conducted to determine protective and risk factors of past 12- month suicidal behaviour in people with any chronic pain. Factors, such as social connectedness, have been found to be protective suicide risk factors and have been less focused on than suicide risk factors (Donald et al. 2006). All suicidal behaviour was grouped together due to the sequencing of the questions (i.e. had to experience thoughts to be asked about plans and attempts) and small numbers.

2.6 Results 2.6.1 Prevalence of chronic pain in the Australian population

Lifetime chronic (greater than 6 months) arthritis/rheumatism was reported by 15.6% of the sample, equating to 2.5 million people aged 16–85 years (Table 2-1). Lifetime chronic migraines were reported by 8.3%, which amounted to 1.3 million people in Australia. Lifetime chronic back/ neck problems were reported by 28.0% of the sample, which equates to around 4.5 million people. Overall, more than 6 million people in Australia reported suffering from any of these chronic pain conditions, almost one- third of the Australian population

Table 2-1: Prevalence of chronic pain among Australian adults, 2007

N, Weighted% Total sample N=8,841 Chronic pain n % 95% CI EPC# (‘000) 95% CI of EPC

Arthritis 1577 15.6 14.5-16.6 250.0 232.5-265.0

Migraines 783 8.3 7.5-9.0 130.0 120.0-145.0

Back/neck problems 2487 28.0 26.6-29.4 450.0 425.0-475.0

Any pain 3585 38.9 37.4-40.4 625.0 595.0-650.0

*rounded to the nearest 50,000 #expected population Count (EPC)

Chapter two: Prevalence of CNCP and suicidality in Australia

There were high rates of comorbidity of chronic pain conditions. Just under two-fifths of those with back pain reported another chronic pain condition (arthritis and/or migraines). Approximately 50% of those with arthritis and three-fifths of those with migraines also reported suffering from another chronic pain condition.

Across the age groups, the 65 years and over age group had the highest levels of chronic pain (59%) – the most common pain condition being arthritis (see Appendix A). Half the 45–64-year-old age group reported experiencing pain, with the most common reported problem being back/ neck problems. Just under one-third of the 25–44-year- old age group and one-sixth of the under 25 years age group reported the experience of ‘any pain’.

2.6.2 Socio-demographic and clinical characteristics of the sample

Owing to the comorbidity of pain conditions, the following bivariate analyses were conducted on the ‘any pain’ category (Table 2-2). People with chronic pain were more likely to be female and over 60 years of age, and less likely to have completed school, be employed or be married or in a de facto relationship. People with pain were also more likely to suffer from any lifetime or past 12-month mental health disorders (major depression, GAD and PTSD). Although those with chronic pain were more likely to meet criteria for a lifetime AUD compared with people without chronic pain, there were no differences for past 12-month AUDs. Conversely, people with chronic pain were more likely to have 12-month diagnoses of substance use disorder, but not lifetime, compared to those without chronic pain (Table 2-2).

Table 2-2: Socio-demographic and clinical characteristics of people with and without chronic pain, NSMHWB, 2007

N, Weighted% Total Chronic pain No chronic pain Comparison of chronic pain vs no N=8841 n=3585 n=5256 chronic pain n (%) n (%) n (%) OR 95% CI p value Socio-demographics Gender (female) 4814 50.3 2138 56.6 2676 46.4 1.51 (1.33-1.71) P<0.001 Age over 60 years 2515 28.5 1465 32.7 1050 14.6 2.83 (2.50-3.20) P<0.001 Married /de-facto 4002 53.0 1747 60.0 2255 48.6 1.58 (1.40-1.79) P<0.001 Completed School 4061 46.4 1330 37.7 2731 52.0 0.56 (0.49-0.63) P<0.001 Employed (%) 5499 65.2 1870 56.5 3629 70.8 0.53 (0.47-0.60) P<0.001. Mental Health Major depression Lifetime 1341 14.8 760 21.2 581 11.1 2.35 (1.98-2.78) P<0.001 Past 12-months 663 7.2 390 10.6 273 5.1 2.20 (1.75-2.75) P<0.001 General Anxiety Disorder Lifetime 728 7.9 451 12.1 277 5.2 2.53 (2.02-3.17) P<0.001 Past 12-month 505 5.6 314 8.9 191 3.5 2.74 (2.09-3.60) P<0.001 Post-Traumatic Stress Disorder Lifetime 664 7.2 377 10.2 287 5.3 2.02 (1.60-2.54) P<0.001 Past 12m 400 4.4 232 6.2 168 3.2 2.02 (1.51-2.71) P<0.001 Substance-use disorders Alcohol Use disorder Lifetime 1847 22.2 812 24.2 1035 20.8 1.21 (1.04-1.42) 0.013 Past 12m 379 4.7 139 4.4 240 4.8 0.90 (0.68-1.20) 0.482 Drug use disorder Lifetime 576 7.0 241 6.5 335 7.4 0.88 (0.69-1.12) 0.293 Past 12m 136 1.6 65 2.0 71 1.3 1.55 (1.02-2.36) 0.041

Table 2-2 (cont): Socio-demographic and clinical characteristics of people with and without chronic pain, NSMHWB, 2007

N, Weighted% Total Chronic pain No chronic pain Comparison of chronic pain vs no N=8841 n=3585 n=5256 chronic pain n (%) n (%) n (%) OR 95% CI p value Suicidality Thoughts Lifetime 1252 13.3 700 19.5 552 9.4 2.3 (1.94-2.69) P<0.001 Past 12m 239 2.3 141 3.4 98 1.6 2.0 (1.56-3.12) P<0.001 Plan Lifetime 386 4.0 239 6.1 147 2.56 2.5 (1.87-3.30) P<0.001 Past 12m 65 0.6 42 1.0 23 0.32 3.1 (1.71-5.50) P<0.001 Attempt Lifetime 328 3.2 204 5.2 124 2.0 2.7 (1.99-3.61) P<0.001 Past 12m 42 0.4 24 0.7 18 0.2 2.9 (1.40-6.05) 0.004 Bold is significant

Chapter two: Prevalence of CNCP and suicidality in Australia

2.6.3 Chronic pain and suicidality in the Australian population

People with chronic pain had 2.3 times the odds of lifetime suicidal thoughts (Table 2- 2) and 2.0 times the odds of reporting past 12-month suicidal thoughts, compared to people without chronic pain. People with chronic pain had 2.5 greater odds of having made a suicide plan at some point, and 3.1 greater odds of a past 12-month suicide plan, compared to those without pain. People with chronic pain had 2.7 times greater odds of making a lifetime suicide attempt, and a 2.9 greater odds of having made a suicide attempt in the past year, compared to people without pain.

Among those who engaged in suicidal thoughts or behaviours, people with a history of chronic pain made up the majority of the total (Table 2-3). Just over 50% of people with lifetime suicidal thoughts had a history of chronic pain; two-thirds (64.8%) of those who had attempted suicide in the previous 12-months had chronic pain, equating to approximately 42,000 people out of the estimated 64,000 who had attempted suicide nationally. The contribution of pain to overall suicidality by age groups can be seen in Appendix B.

Table 2-3: Prevalence of chronic pain amongst people who had experienced lifetime and past 12-month suicidality

Total EPC (‘000) % 95%CI Lifetime… suicidal thoughts 1200.0 56.4 52.7-60.1 suicide plan 400.0 60.4 53.6-66.8 suicide attempt 300.0 62.3 55.2-68.9 Past 12-month… suicidal thoughts 200.0 58.0 49.5-66.0 suicide plan 50.0 66.0 51.7-77.9 suicide attempt 50.0 64.8 46.2-79.7 *rounded to the nearest 50,000 #expected population Count (EPC)

Chapter two: Prevalence of CNCP and suicidality in Australia

There was a difference between pain conditions and their contribution to suicidal thoughts or behaviours. Between 15.2% and 20.6% of people who had experienced lifetime and/or past 12-month suicidality had a history of migraines (Appendix C). Between 13.1% and 23.2% of people who had experienced lifetime and/or past 12- month suicidality had a history of arthritis. Between 45.3% and 54.3% of people who had experienced lifetime and/or past 12-month suicidality had a history of chronic back pain.

2.6.4 Chronic pain as an independent predictor of suicidality

In the unadjusted models and after adjusting for sociodemographic characteristics, all pain conditions and the ‘any pain’ condition were significantly associated with lifetime suicidality (Table 2-4).

Table 2-4: Logistic Regression Models#: Odds of lifetime suicidality by presence of chronic pain conditions and ‘any pain’, NSMHWB, 20071

Arthritis Migraines Back/neck problems Any pain n=1,577 n=783 n=2,487 n=3585 Suicidal thoughts OR 95% CI p value OR 95% CI p value OR 95% CI p value OR 95% CI p value Unadjusted 1.41 1.15-1.72 0.001 2.55 2.02-3.19 <0.001 2.43 2.03-2.89 <0.001 2.28 1.94-2.69 <0.001 Model 1 1.91 1.50-2.42 <0.001 2.43 1.93-3.05 <0.001 2.71 2.27-3.23 <0.001 2.73 2.30-3.26 <0.001 Model 2 1.58 1.21-2.05 0.001 1.74 1.31-2.32 <0.001 1.82 1.47-2.22 <0.001 1.86 1.53-2.27 <0.001 Model 3 1.37 1.04-1.80 0.023 1.49 1.12-2.00 0.006 1.66 1.34-2.06 <0.001 na na na Suicide plan Unadjusted 1.69 1.22-2.32 0.001 2.35 1.70-3.25 <0.001 2.50 1.89-3.32 <0.001 2.48 1.87-3.30 <0.001 Model 1 2.37 1.64-3.43 <0.001 2.10 1.52-2.93 <0.001 2.71 2.05-3.58 <0.001 2.89 2.18-3.82 <0.001 Model 2 1.83 1.26-2.67 0.002 1.25 0.85-1.85 0.257 1.52* 1.11-2.09 0.010 1.64 1.21-2.25 0.002 Model 3 1.68 1.15-2.48 0.008 1.08 0.73-1.60 0.690 1.39 0.99-1.96 0.059 na na na Suicide attempt Unadjusted 1.59 1.07-2.36 0.022 2.65 1.84-3.79 <0.001 2.56 1.91-3.43 <0.001 2.69 1.99-3.61 <0.001 Model 1 2.14 1.34-3.92 0.001 2.30 1.56-3.38 <0.001 2.77 2.06-3.73 <0.001 3.14 2.30-4.29 <0.001 Model 2 1.65 0.94-2.74 0.052 1.50 0.96-2.34 0.075 1.61 1.13-2.30 0.008 1.87 1.30-2.67 0.001 Model 3 1.46 0.86-2.48 0.158 1.31 0.82-2.08 0.258 1.47* 1.02-2.11 0.037 na na na #Model 1: socio-demographic factors (age over 60, sex, married, completed school, employed); Model 2: socio-demographic factors, mental health (lifetime major depression, lifetime GAD, lifetime PTSD, lifetime alcohol use disorder, lifetime drug use disorders); Model 3: socio-demographic factors, mental health and number of other pain conditions (i.e. arthritis, chronic migraines, chronic back or neck problems). Bold is significant Regressions were run with the removal of age and gender to determine if there was an effect for controlling for age and gender in the regressions and in the survey weightings. There were no differences in any of the models with the exception of 1) arthritis- Model 3 was no longer significant for suicidal thoughts (OR 1.24, 95%ci 0.96-1.61)and 2). migraines became significant in Model 2 for lifetime attempts (OR 1.60, 95%CI 1.04- 2.48)

Chapter two: Prevalence of CNCP and suicidality in Australia

Table 2-5: Logistic Regression Models*: Odds past 12-month suicidality by presence of ‘any pain’ condition, NSMHWB, 20071

Past 12m N=239 OR 95% CI P value Suicidal thoughts (n=239) Unadjusted 2.21 1.56-3.12 <0.001 Model 1 2.74 1.95-3.86 <0.001 Model 2 1.66 1.11-2.47 0.012 Suicide plan (n=65) Unadjusted 3.07 1.71-5.50 <0.001 Model 1 3.58 1.91-6.71 <0.001 Model 2 1.71 0.85-3.46 0.130 Suicide attempt (n=42) Unadjusted 2.91 1.40-6.05 0.004 Model 1 3.52 1.57-7.91 0.002 Model 2 1.95 0.74-5.1 0.173 Model 1: socio-demographic factors (age, sex, married, income, completed school, unemployed); Model 2: socio- demographic factors, mental health (lifetime major depression, lifetime GAD, lifetime PTSD, lifetime alcohol use disorder, lifetime drug use disorders); Bold is significant 1. Regressions were run with the removal of age and gender to determine if there was an effect for controlling for age and gender in the regressions and in the survey weightings. There were no differences.

After adjusting further for mental health disorders, the model, although reduced, remained significant for: suicidal thoughts for all the pain conditions and the ‘any pain’ category, suicide plan for arthritis, back/neck problems, and the ‘any pain’ category. After adjusting for socio-demographic and mental health, only back/neck problems and ‘any pain’ remained significant for lifetime suicide attempts.

After further adjusting for the number of other pain conditions, arthritis remained significantly associated with lifetime suicidal thoughts and plans. Migraines remained significantly associated with lifetime thoughts, but not lifetime suicidal plans or attempts. Back/neck problems remained significant after further adjusting for the other pain conditions for lifetime suicidal thoughts and attempts, but not lifetime suicidal plans.

Chapter two: Prevalence of CNCP and suicidality in Australia

Owing to the small numbers, only the ‘any pain’ category was used in analyses of past 12-month suicidality (Table 2-5). In the unadjusted model and after controlling for socio-demographic factors, ‘any pain’ was significantly associated with past 12-month suicidal thoughts, plans and attempts. After further adjusting for mental health disorders, chronic pain remained significantly associated with suicidal thoughts (adjusted OR 1.66, 95% CI=1.11–2.47), although it was no longer associated with suicide plans or attempts.

2.6.5 Risk factors associated with past 12-month suicidal behaviours in people with chronic pain

Figure 2-1 presents a logistic regression of the odds of suicidal behaviours amongst people with pain (data from the regression can be found in Appendix D). People who had chronic pain were more likely to experience past 12-month suicidal behaviours if they suffered from past 12-month major depression (OR 3.51, 95% CI=1.91–6.45), past 12-month GAD (OR 2.11, 95% CI=1.07–4.17), past 12-month PTSD (OR 2.52, 95% CI=1.36–4.67), past 12-month AUD (OR 2.53, 95% CI=1.13–5.63), or reported their physical health to be fair-to-poor (as opposed to good or very good; OR 2.57, 95% CI=1.54–4.20).

Chapter two: Prevalence of CNCP and suicidality in Australia

Figure 2-1: Odds of past 12-month suicidal behaviour amongst those with chronic pain

People with pain were less likely to experience past 12-month suicidal behaviours if they were over 60 years old (OR 0.14, 95% CI=0.20–0.83) or they were married/ defacto (OR 0.39, 95% CI=0.12–0.70).

2.7 Discussion 2.7.1 Key findings

To our knowledge, this is the first population-based study that has estimated the proportion of people with suicide attempts who had chronic pain specifically: we estimate two out of three people who attempt suicide in Australia in a given year (65%) have a co-occurring chronic pain condition. People with a history of chronic pain contributed a substantial proportion of the total number of people who had engaged in lifetime and past 12-month suicidality. Approximately 58% of the people with suicidal thoughts in the previous 12-months had a history of chronic pain. While high- risk groups such as younger people, males, people with mental disorders, and gay, lesbian, bisexual and transgender people have generally been recognised as at-risk groups, there has been little discussion of the contribution that people with chronic pain may make to the population of those who experience suicidality (Suicide Prevention Australia 2012; Senate Community Affairs Committee Secretaria 2010). Our finding suggests that those with chronic pain are at high risk for suicide and a

Chapter two: Prevalence of CNCP and suicidality in Australia

substantial proportion of people who engage in suicidal behaviours have a history of chronic pain.

To our knowledge, this is also the first Australian study that has compared the socio- demographics and clinical characteristics of those with chronic pain in a nationally representative survey. Consistent with previous literature (Blyth et al. 2001; Braden and Sullivan 2008; Tang and Crane 2006), chronic pain was found to be associated with a number of socio-demographics and clinical characteristics. People with chronic pain have a number of complex issues surrounding employment, physical health and mental health (Rogers et al. 2013), and with a heightened risk of suicide it is clear that a number of these issues need to be addressed when receiving treatment for their chronic pain. These results provide support to the complexity of chronic pain patients and the need for a multi-disciplinary approach to their care and treatment.

Consistent with the literature (Tang and Crane 2006), lifetime and past 12-month suicidality were more common in those with chronic pain, with substantial variations across different pain conditions, whereby those with neck or back problems had by far the highest levels of suicidality. The current results, however, differed from the other epidemiological studies from Braden et al. (Braden and Sullivan 2008), Ratcliff et al (Ratcliffe et al. 2008) and Ilgen et al. (Ilgen et al. 2008) with regards to specific pain conditions. Specifically, migraines in the current study were not associated with lifetime suicidality, after controlling for mental health and substance use. Pain conditions are highly comorbid, and it may be a limiting factor of general national surveys to determine the effect that specific pain conditions have on suicidality. More focused research in this area is essential to understand the relationship between specific pain conditions and suicidality.

The finding that chronic pain was independently associated with lifetime suicidality and past 12-month suicidal ideation, after controlling for mental health, raises important clinical implications. Although previous suicide research has identified the

Chapter two: Prevalence of CNCP and suicidality in Australia

importance of mental health factors and their contribution to suicide, the finding of chronic pain as independently associated with suicide suggests the need for clinicians to be aware of this heightened risk in people presenting with chronic pain, irrespective of their mental health status (Cheatle 2011).. Further, patients that present with pain and mental health issues may be at a much heightened risk for active (plans and attempts) suicidality and should be assessed with appropriate referrals to mental health professionals made if necessary. Future research will be important in examining what specific pain-related factors, such as pain severity, pain coping, length of time in pain and pain conditions, are associated with suicide.

The prevalence of chronic pain in the current study was higher than previous estimates by Blyth et al (Blyth et al. 2001) in the Australian population. Over one-third of the sample (38%) reported the experience of chronic pain compared to the previous estimates of 17% of males and 20% of females. An ageing population may explain some of this difference (the first estimates were made on data from 1997, compared to 2007 in the current study), though it is difficult to determine if this is an increase or differences in methodology. Despite these concerns, it is clear that chronic pain is common in Australia and is an area that requires future research and consideration.

2.7.2 Strengths and limitations

The strengths of this study are that it is population-based and provides detailed information on DSM-IV mental health problems and chronic conditions, but there are a number of limitations that need to be considered. First, it is not possible to determine whether the chronic pain and suicidality were directly related to each other. Although we are able to examine associations, the cross-sectional design of the survey does not allow us to determine causality. Furthermore, pain factors, such as length and severity, have been found to be factors associated with suicidality in chronic pain patients (Hassett, Aquino, and Ilgen 2014; Tang and Crane 2006), the current survey did not examine these factors. The numbers for past 12-month suicidal plans and attempts amongst people with chronic pain were relatively low (n=65 and n=42, respectively) and the current findings were not consistent with previous research that suggests an

Chapter two: Prevalence of CNCP and suicidality in Australia

association between chronic pain and past 12-month plans and attempts after controlling for mental health (Ratcliffe et al. 2008). This finding needs to be interpreted with caution: it may be the lack of statistical power that led to a non-significant result. Despite these limitations, the findings of this study are important and suggest that more focused research on chronic pain patients may aid in the understanding of this relationship. Prospective longitudinal studies of pain, suicide and mental health will help our understanding of these associations and will be important clinically.

2.7.3 Conclusions

The current study found that chronic pain was a common complaint in the Australian population and a substantial proportion of people who engaged in suicidal behaviours had a history of chronic pain. Although previous research and treatments have focused on mental health and suicide, with an ageing population and the finding of chronic pain being independently associated with suicidality, it suggests clinicians need to be aware of this heightened risk in patients with chronic pain and, if necessary, provide referrals to appropriate mental health professionals.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3 PAPER TWO: COHORT PROTOCOL PAPER: THE PAIN AND OPIOIDS IN TREATMENT

(POINT) STUDY

Gabrielle Campbell1, Richard Mattick1, Raimondo Bruno1,2, Briony Larance1, Suzanne Nielsen3,4, Milton Cohen5, Nicholas Lintzeris3,4, Fiona Shand6, Wayne D. Hall7,8, Bianca Hoban1, Chyanne Kehler1, Michael Farrell1 and Louisa Degenhardt1,9,10,11

1. National Drug and Alcohol Research Centre, UNSW, Australia 2. School of Medicine, University of Tasmania, Australia 3. Sydney Medical School, Sydney University, Australia 4. The Langton Centre, South East Sydney Local Health District (SESLHD) Drug and Alcohol Services, Australia 5. St Vincent’s Clinical School, UNSW Medicine, UNSW Australia 6. Black Dog Institute, UNSW Australia 7. Centre for Youth Substance Abuse Research, University of Queensland, AUSTRALIA 8. National Addiction Centre, Kings College, London ENGLAND 9. School of Population and Global Health, University of Melbourne, Australia 10. Murdoch Children’s Research Institute, AUSTRALIA 11. Department of Global Health, School of Public Health, University of Washington, USA

Paper two has been published in BMC Pharmacology and Toxicology (Campbell et al. 2014)

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3.1 Copyright Statement I certify that this publication was a direct result of my research towards this PhD, and that reproduction in this thesis does not breach copyright regulations.

Campbell, G., Mattick, R., Bruno, R., Larance, B., Nielsen, S., Cohen, M., Lintzeris, N., Shand, F., Hall, W., Hoban, B., Kehler, C., Farrell, M., & Degenhardt, L. (2014). Cohort protocol: The Pain and Opioids IN Treatment (POINT) study. BMC Pharmacology and Toxicology, 15; 17

Gabrielle Campbell November, 2015

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3.2 Preamble The paper presented in Chapter 2 presented the prevalence of CNCP and the association of CNCP and suicidality at the population level. Approximately two-fifths of the Australian sample reported the experience of pain which had lasted longer than six months. Consistent with the international literature (Braden and Sullivan 2008; Ilgen et al. 2008; Ratcliffe et al. 2008), people living with CNCP had a greater number of comorbidities than those without CNCP and suicide-related behaviours were two-three times greater in those with CNCP. Further, it was estimated that between 50-60% of people who had engaged in suicidal-related behaviours had a history of CNCP. This paper was the first to demonstrate the significant contribution of CNCP to suicide- related behaviours.

Although the paper presented in Chapter 2 was important in providing the nature and extent of CNCP in Australia, the survey was not specifically designed to examine chronic pain and therefore important pain-related factors were not included. Chapter 3 describes the Pain and Opioids IN Treatment (POINT) study, which was designed to examine the trajectories of opioid use in CNCP and the benefits and outcomes of opioids for CNCP over a two-year period. This thesis only presents data from the baseline data collected from the POINT study.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3.3 Abstract Background: Internationally, there is concern about the increased prescribing of pharmaceutical opioids for chronic non-cancer pain (CNCP). In part, this is related to limited knowledge about the long-term benefits and outcomes of opioid use for CNCP. There has also been increased injection of some pharmaceutical opioids by people who inject drugs, and for some patients, the development of problematic and/or dependent use. To date, much of the research on the use of pharmaceutical opioids among people with CNCP, have been clinical trials that have excluded patients with complex needs, and have been of limited duration (i.e. fewer than 12 weeks). The Pain and Opioids In Treatment (POINT) study is unique study that aims to: 1) examine patterns of opioid use in a cohort of patients prescribed opioids for CNCP; 2) examine demographic and clinical predictors of adverse events, including opioid abuse or dependence, medication diversion, other drug use, and overdose; and 3) identify factors predicting poor pain relief and other outcomes.

Methods/Design: The POINT cohort comprises around 1,500 people across Australia prescribed pharmaceutical opioids for CNCP. Participants will be followed-up at four time points over a two year period. POINT will collect information on demographics, physical and medication use history, pain, mental health, drug and alcohol use, non- adherence, medication diversion, sleep, and quality of life. Data linkage will provide information on medications and services from Medicare (Australia’s national health care scheme). Data on those who receive opioid substitution therapy, and on mortality, will be linked.

Discussion: This study will rigorously examine prescription opioid use among CNCP patients, and examine its relationship to important health outcomes. The extent to which opioids for chronic pain is associated with pain reduction, quality of life, mental and physical health, aberrant medication behaviour and substance use disorders will be extensively examined. Improved understanding of the longer-term outcomes of chronic opioid therapy will direct community-based interventions and health policy in Australia and internationally. The results of this study will assist clinicians to better

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

identify those patients who are at risk of adverse outcomes and who therefore require alternative treatment strategies.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3.4 Introduction Chronic non-cancer pain (CNCP) is a worldwide, common complaint. The prevalence of chronic pain (defined as pain present daily for three months or more) in the Australian population is 17% for males and 20% for females (Blyth et al. 2001). In one survey of 16 European countries, between 10% and 30% of participants reported chronic pain, 16% of whom said that some days the pain made them “want to die” (Beubler et al. 2006). Chronic pain can have a major impact on an individual and the community, with social, financial, employment and health costs (Beubler et al. 2006).

CNCP is caused by many factors, including trauma. The varied aetiology probably impacts upon the effectiveness of treatment (Blyth et al. 2004; Blyth et al. 2001; Blyth, March, and Cousins 2003; Savage 1999). Physical and psychological factors such as depression and anxiety, a history of psychological trauma, and sleep problems moderate the pain experience Context is also important: relationships, occupational setting and culture all affect the experience and expression of pain (Savage 1999). Between 30-50% of those with chronic pain in the above study reported that their pain was not “adequately controlled” (Beubler et al. 2006).

Effective behavioural treatments and non-opioid pharmacotherapy exist for CNCP (Savage 1999), but even when a combination of interventions is used, some patients continue to suffer. The increase in the use of opioid analgesics for CNCP would suggest that there is conclusive evidence regarding their positive impact on pain reduction and quality of life. Indeed, qualitative and quantitative reviews of the evidence have concluded that chronic opioid therapy does produce clinically significant reductions in pain, albeit in the range of 2 to 3 points on a 0 to 10 visual analogue scale, or around 30% (Noble et al. 2010; Chou 2009; Kalso et al. 2004). Nevertheless, reviewers have cautioned that patients need to be carefully selected and monitored, given that opioids are also associated with potentially serious harms and significant treatment drop out due to adverse effects (Chou 2009; Noble et al. 2010). Further, the degree to which pain reduction is achieved varies, and evidence on changes to quality of life and functional status is inconclusive (Noble et al. 2010). Some of this variability in pain

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

reduction may be related to the type of opioid used, with one meta-analysis finding that strong but not weak opioids reduced pain more effectively than other analgesics (Furlan et al. 2006).

Controlled trials have evaluated pharmaceutical opioids in the treatment of a range of CNCP conditions and have demonstrated modest attenuation of pain (Bloodworth 2005). Evidence indicates a modest short-term benefit; no studies have yet run for long enough to demonstrate long-term benefit of opioids for chronic non-cancer pain. The three systematic reviews of opioids for chronic non-cancer pain published to date provide evidence for a modest, short-term analgesic benefit (Eisenberg, McNicol, and Carr 2006; Furlan et al. 2006; Manchikanti, Abdi, Atluri, Balog, Benyamin, Boswell, Brown, Bruel, Bryce, and Burks 2012). This is a best-case scenario because RCTs select the sub-set of patients most likely to receive a clinical benefit and have short follow-up periods (average trial duration of five weeks), meaning they do not report on longer term outcomes (Ballantyne and Shin 2008; Moulin et al. 1996). An ongoing systematic review shows that the only evidence of long-term analgesic benefit (improved physical function and quality of life) is weak, because it is based on non-blinded studies with significant potential for reporting bias (Noble et al. 2010).

There is support from peak pain organisations of the use of opioids in the treatment of CNCP (American Academy of Pain Medicine and American Pain Society 1997; Beubler et al. 2006; Cohen and Wodak 2010; Royal Australasian College of Physicians 2008). Debate continues about how, when, and in what manner opioids should be prescribed for this diverse patient group (Baca and Grant 2007; Ballantyne 2006; Ballantyne 2007; Eriksen et al. 2006; Lipman 2007; Savage 1999). Consensus statements have recommended that prescription of opioids for chronic pain is considered only after following: a thorough assessment of the patient’s pain problem and history; development of a treatment plan; consultation with a pain specialist, if necessary; and regular reviews of patient progress (American Academy of Pain Medicine and American Pain Society 1997; Beubler et al. 2006).

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

Over the past decade, there has been increasing professional and public concern in a number of countries about pharmaceutical opioid use and related harms(Leong, Murnion, and Haber 2009b; Compton and Volkow 2006b). This has been driven by increases in prescribing of these drugs, especially in the USA and Canada. The increase in prescribing in Australia has been less than in Europe and the United States (US) (Centers for Disease -Control Prevention 2008; Dhalla et al. 2009; Fredheim et al. 2010), but nonetheless, between 1992-2007 the number of opioid prescriptions in Australia increased by around 300% (Leong, Murnion, and Haber 2009b). This increase in prescribing has been accompanied by increased injection of some opioids by people who inject drugs (Degenhardt et al. 2006); increased concern about the appropriateness of prescribing these drugs for chronic non-cancer pain (CNCP)(Noble et al. 2010); and for some patients, the development of problematic and/or dependent use.

The use of opioids, within, and outside the bounds of a doctor’s prescription has been cause for concern because of the risk of iatrogenic dependence (Compton and Volkow 2006a), and opioid overdoses, with pharmaceutical opioids now comprising the majority of fatal and non-fatal drug overdoses in the US (Compton and Volkow 2006b; Paulozzi, Budnitz, and Xi 2006). A review of 67 studies found that 11.5% of chronic pain patients engaged in aberrant drug-related behaviours such as diversion, prescription forgery, injecting, multiple episodes of prescription loss, escalating doses, and doctor shopping, 3.3% developed opioid abuse/dependence (Fishbain, Cole, Lewis, Rosomoff, Rosomoff, et al. 2008).

Despite this increasing concern, little is known about the magnitude of risk of such adverse events in patients prescribed opioids. Clinical trials, because they often exclude more complex patients, including those with comorbid conditions that may increase risks for developing opioid related problems (e.g. history of substance use disorder), typically find far lower rates of aberrant drug-related behaviours and abuse/dependence (Fishbain, Cole, Lewis, Rosomoff, Rosomoff, et al. 2008). Many studies have also been of limited duration (~12 weeks) and few have examined

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

aberrant drug use behaviours. Those of longer duration have had a small number of participants and therefore lack statistical power. Other studies have examined treatment of localised conditions, e.g. low back pain, or have evaluated a single drug or formulation of a drug. Only a few of the studies reviewed were prospective or longitudinal, thus limiting the conclusions that can be drawn. Further, there is little consensus regarding the diagnosis of opioid dependence in the context of chronic opioid treatment for chronic pain (Lintzeris 2013).

Little is known about the patterns of opioid prescribing for individual patients, and the long term outcomes for these patients. Small retrospective cohort studies conducted elsewhere have examined treatment duration, side effects, pain reduction, and adverse events (Reid et al. 2010) and aberrant behaviours (Hariharan et al. 2007). Larger retrospective cohort studies have examined the risk of overdose (Dunn et al. 2010), the impact on disability (Franklin et al. 2008), non-medical use (Pletcher et al. 2006),conditions treated in older adults (Reid et al. 2002), and rates of adverse events (Hartung et al. 2007).

The Pain and Opioid IN Treatment (POINT) study is an Australian-first study that aims to document patterns of pharmaceutical opioid prescribing, and risk of adverse events, in a prospective cohort of patients prescribed opioids for chronic non-cancer pain.

3.5 Methods 3.5.1 Study aims

1. To examine patterns and outcomes of opioid analgesic use in a cohort of patients prescribed opioids for chronic non-cancer pain (CNCP).

2. To examine the demographic and clinical predictors of adverse events among a cohort of CNCP patients, including opioid abuse or dependence, medication diversion, other drug use, and overdose.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3. To identify factors which predict poor self-reported pain relief and other clinical outcomes.

3.5.2 Study design and setting

The POINT study is a prospective cohort study of 1,500 persons who have been prescribed opioids for chronic non-cancer pain that follows their progress over two years and examine the predictors of clinical outcomes over this period. Participants provided consent to link data on health service utilisation and mortality data; the study will utilise data linkage to examine other clinical outcomes in the longer-term, such as opioid substitution therapy utilisation and hospital admissions. Prospective cohort studies can provide highly reliable and detailed data about a range of outcomes with the advantage of collecting data at the time or close to the time an event occurs, reducing the effects of recall bias (Euser et al. 2009) and making it easier to draw causal inferences about associations with later outcomes.

3.5.3 Ethics

The study was approved by the Human Research Ethics Committee of the University of New South Wales (HREC reference: # HC12149). The study also received A1 National Pharmacy Guild Approval to approach pharmacists to assist with recruitment of participants (Approval n. 815). Approval was obtained by the Strategic Information Design and Governance Branch of the Department of Human Services to access Medicare data of participants that consent to access of their records (reference number: 2012/C011091).

3.5.4 Sample size calculations

Power analyses were conducted to estimate minimum sample size required to assess the potential effects of opioid use on key clinical outcomes. We examine, dropout due to adverse events as an indicative example.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

Power analyses were conducted using GLIMMPSE (Kreidler et al. 2013) to determine the minimum sample size to examine the potential effect of opioid use on key outcomes. As a conservative indicative example, we consider changes in pain score. Data on Australian prescriptions suggests that approximately 10% of opioid analgesic scripts can be defined as high dose (National Drug and Alcohol Research Centre 2012); we therefore assumed that at least 10% of the POINT cohort will be on a high dose of opioid analgesics. The variability of changes in pain score from baseline over a six month period was estimated from a recent Cochrane review (Noble et al. 2010).Even with a drop-out rate as high as 25%, with an initial sample of 1500 we will be able to detect differences between low and high dose groups over time as small in magnitude as 0.30 at above 80% power and a Type I error rate of 5%. These calculations consider the effect of repeated observations over time (base correlation of 0.5 between adjacent repeated measures, with a linear exponential autoregressive decay rate of 0.05) and controlling for a normally distributed baseline covariate explaining up to 20% of the variation in pain scores. Given that other research questions relate to more equally balanced subgroups (e.g. comparisons of chronic neck/back pain vs. other types of pain which are distributed in a ratio of 4:1) or outcomes with less variability (e.g. changes in quality of life:) then we can be confident that the POINT cohort is sufficiently powered to address the primary outcomes.

3.5.6 Eligibility criteria

POINT participants were: 18 years or older; competent in English; and mentally and physically able to complete telephone and self-complete interviews; without memory or comprehension difficulties; living with chronic non-cancer pain (defined as pain present daily for a period of three months or more); prescribed a Schedule 8 opioid (an Australian classification of drugs of dependence that are subject to additional regulatory controls regarding their manufacture, supply, distribution, possession and use (Therapeutic Goods Administration 2013)); and having taken such opioids for CNCP for more than 6 weeks.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

A history of injecting drug use (IDU) was not an exclusion criterion, but those currently prescribed pharmaceutical opioids for opioid substitution therapy (OST) for heroin dependence were not eligible. Nor were those taking opioids for cancer pain.

3.5.7 Recruitment

A database of pharmacies and chemists across Australia and their contact details were purchased in May 2012 (Maven Marketing 2012). The list included 7,136 pharmacies. After removing duplicates, those that had closed down, or were not suitable for the study (i.e. located in a hospital or were a compounding pharmacy), we had a final list of 5,994 pharmacies.

Each fortnight, approximately, 100-150 pharmacies were randomly allocated into a Wave (with the exception of Tasmania which was sampled first). A flyer inviting to participate in research was faxed to all pharmacies in the Wave that had a fax number (Appendix H). Any pharmacies who expressed a lack of interest were not contacted further. Those who indicated they were interested in more information, or who did not respond to the fax, were called and the study was explained to a pharmacist who was asked if they were willing to participate.

Interested pharmacists were enrolled in the study for a six week period because it was expected that most people on prescription opioids would renew their script within this time. Pharmacists were asked to approach any customers that were prescribed a Schedule 8 opioid for CNCP for a period of greater than six weeks. We were confident that pharmacists were able to identify customers who had CNCP, rather than chronic cancer pain, by examining the customer’s record for other prescribed medications. Pharmacists were also able to use the medication history to confirm that customers had been on a prescription opioid for more than 6 weeks.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

Customers who fit the above criteria were given a flyer about the study by the pharmacist (Appendix I). Interested customers were asked whether they would like the pharmacists to send their details to researchers, or if they wanted to contact researchers themselves. All flyers had a unique pharmacy number and pharmacists were reimbursed $20 for each eligible participant they referred into the study. POINT staff made reminder calls every fortnight for the 6 week period to pharmacists.

POINT staff determined the eligibility of those who were referred to the study, or who contacted the POINT team. Eligible participants went through a voluntary informed consent process (Appendix J-L). Those who were willing to participate, after being given details of the study, were booked in for their initial interview which was conducted over the phone and took approximately 1-1.5 hours. Detailed locator information (Appendix M) was collected at Baseline to prevent sample attrition and this was updated at all follow-up assessments.

3.5.8 Interview procedure

This study had four assessment waves: Baseline, T2 follow up (3-months), T3 follow up (12-months), and T4 follow up (24-months) (see Table 3-1).

Table 3-1: Time-points and data collection method of the POINT prospective cohort

Baseline T2 3-months T3 12-months T4 24-months • Phone interview • Self-complete • Self-complete • Phone interview • Self-complete questionnaire (paper questionnaire (paper • Self-complete questionnaire (paper and pencil and and pencil and online questionnaire and pencil and online online versions) versions) (paper and pencil versions) (Appendix and online versions) N-P)

Phone interviews were conducted by trained interviewers at the baseline and 24- month time-point. Interviewers had suicide assistance training, a minimum 3-year health or psychology degree, and were provided glossaries of general and chronic pain medications and conditions (see Appendices E-G). A self-completion survey was sent to

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

all participants at each time-point. Participants were able to nominate to complete the self-complete survey either, online, by pen and paper, or on the telephone with a POINT team member. Participants were reimbursed $40 for the baseline interview, $25 at each 3-month and the 12-month time point and $60 at the final time point.

3.5.9 Cohort maintenance strategies

There were a number of methods used to prevent attrition in the current study. Firstly, interested participants (pharmacists or participants) were called within a week from their expression of interest in the study. Secondly, participants were offered a variety of study completion methods, i.e. telephone, pen and paper and online. In addition to this, we offered to record verbal consent to participants who found it difficult to return the consent forms via mail. A variety of methods of study completion that are flexible and accommodating have been found to increase retention (Booker, Harding, and Benzeval 2011). Thirdly, a detailed locator form was completed at the initial contact with participants. This form gathered detailed information on contact details of the participants, as well as the contact details of two secondary contacts (i.e. family, friends, medical professionals etc.). Finally, participants and pharmacists were reimbursed within a timely manner for their time and contribution to the research. Other methods to improve retention were letters thanking the participant for involvement in the study; this was sent after a payment had been made. A newsletter was sent annually to participants to provide study updates and progress.

3.5.10 Project governance

A reference group was established for the study which included general practitioners, consumers, pain specialists, addiction medicine specialists, pharmacists and researchers. This reference group was consulted at the beginning of the study to assist in study and questionnaire design. They were also consulted throughout the study.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3.5.11 Measures

Table 3-2 shows the measures, tools, domains and time-points at which data are collected. These measures were based on recommendations made under the auspices of the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT). This Initiative involved 27 specialists from academia, governmental agencies, and the pharmaceutical industry who participated in a consensus meeting and identified core outcome domains and measures that should be considered in clinical trials of treatments for chronic pain (Dworkin et al. 2005; Turk et al. 2003). The draft content of our interview was also reviewed and discussed by the POINT advisory committee.

Table 3-2: Measures, tools, domains and time-points for data collection for the POINT study

Domain Measure T1 T2 T3 T4 Those who discontinue opioids Demographics Age and sex  × × × × Marital status, employment, income      Educational attainment  × × ×  Pain Pain Brief pain Inventory (BPI)(Cleeland and Ryan 1994)      Current chronic pain diagnosis, incident pain     × Illness and disability history Chronic Conditions section of the CIDI 3.0(World Health     × Organization 2001; Tan et al. 2004) Current physical disabilities 6-items from the Washington Group (Centre for Disease Control      and Prevention 2006) Physical functioning Physical functioning Brief Pain Inventory (BPI) (Tan et al. 2004), Short-Form (SF)-12      (Bjorner and Turner-Bowker 2009) Exercise Exercise routine and how pain effects this      Falls Un-standardised questions examining falls  ×    Sleep Medical Outcome Sleep scale (MOS) (Rejas et al. 2007)      Coping and pain Pain: Self-Efficacy Questionnaire (PSEQ) (Nicholas 1989; Nicholas      2007) Treatment Past week use of prescription and OTC Self-complete 7 day medication diary      medications and dose Current prescribed medications and days of use      in last month Other treatments for chronic pain      Perceived effect of treatment Patients’ Global Impression of Change scale (PGIC) (Hurst and     × Bolton 2004) Beliefs about medicines Beliefs about Medicines Questionnaires (Horne, Weinman, and  ×   

Domain Measure T1 T2 T3 T4 Those who discontinue opioids Hankins 1999) Convenience of accessing medications  × × × × Side-effects of opioid medication Pain Assessment and Documentation Tool (PADT) (Passik et al.     × 2004) Barriers to treatment      Reasons for discontinuance of opioids      Aberrant opioid medication-related behaviours Opioid Related Behaviours In Treatment ORBIT scale (Larance et      al. 2015) Opioid Difficulties Prescribed Opioid Difficulties Scale (PODS) (Banta-Green et al.  ×    2010) Quality of life WHO-QoL-BRIEF (Murphy B et al. 2000)      Mental health Mental health history      Depression Patient Health Questionnaire –9 (PHQ-9) (Kroenke, Spitzer, and Williams      2001; Martin et al. 2006) General Anxiety disorder Patient Health Questionnaire (GAD) (Spitzer et al. 2006)      Social Anxiety Social Interaction Anxiety Scale (SIAS) (Mattick and Clarke 1998)      Social Phobia Social Phobia Scale (SPS) (Mattick and Clarke 1998)     

Agoraphobia From the MINI International Neuropsychiatric Interview (Sheehan et al.      1998) Post-Traumatic Stress Disorder PC-PTSD (Prins et al. 2004)      Borderline personality disorder screener Screener from the CIDI 3.0 (World Health Organization 2001)      Child abuse Childhood Trauma Questionnaire (Bernstein et al. 2003)      Suicidality and self-harm National Survey of Mental Health and Well-Being      Social support (SPQ), MOS Social Support (Sherbourne and Stewart 1993)      Locus of control Multidimensional Health Locus of Control (B, C) (Wallston, Wallston, and  ×    DeVellis 1978) Substance use Lifetime and current drug use, illicit drug market      Drug and alcohol abuse and dependence CIDI 3.0 (World Health Organization 2001)  × × × ×

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

3.5.12 Data linkage

Consent was also obtained from POINT participants to link their data with the following datasets: Medicare: Medicare (the Australian national health care scheme) collects data on all patient services with medical practitioners in Australia. The following data fields are collected: item number; Medicare benefit; date of service, processing or referral; indication of whether or not the service was provided in hospital; number of services, rendered or referred; and State of patient. Data on prescription medications will also be collected from the Pharmaceutical Benefits Scheme (PBS).

National Death Index (NDI): The NDI is a fully identified database held by the Australian Institute of Health and Welfare (AIHW) and contains mortality data collected from each of the State and Territory Births, Deaths and Marriage Registers. It collects information on deaths and includes date, state, and causes (primary causes for all records, secondary causes for deaths in 1997 and later).

Opioid substitution treatment: The Pharmaceutical Drugs of Addiction System (PHDAS) database is a comprehensive record of all individuals in NSW who received addictive drugs dispensed by clinicians with the authorisation of the NSW Director-General of Health since 1985. The PHDAS is a fully identified database of all methadone and buprenorphine recipients (i.e. first name, last name, date of birth, sex, postcode of residence). As proof of identity must be shown to the prescribing doctor, the name and date of birth variables are of high quality in this dataset. The database also records patient admissions and exits from the treatment program, and the type of pharmacotherapy dispensed.

The data linkage with PHDAS will be performed by probabilistic record linkage software by staff at Medicare (PBS and Medicare data), AIHW (NDI data) and NSW Health (PHDAS, NSW hospital data). Variables used for matching purposes will include:

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

full name, date of birth, sex, date and state of last known contact where available. Matching will be performed in several stages (“passes”), commencing with the strictest of criteria (e.g.: an exact match of full name and date of birth) to more relaxed criteria (e.g.: slight spelling and date variations permitted). The records matched in each pass will be weighted according to the quality of the match.

3.5.13 Data analysis

Descriptive statistics will be used to determine the quantity and type of opioid and other medication use, and the prevalence of other key variables such as psychiatric disorders, demographics and physical health problems. A range of standard statistical techniques will be used depending upon the research question, such as logistic and linear regressions and survival analysis. Latent class analysis may be used to examine whether there are differing latent classes of CNCP participants, and whether such classes predict clinical outcomes. Generalised estimating equations (GEE), (statistical techniques for analysing correlated data) will be used to examine whether the quantity and type of opioid analgesic use predicts later health outcomes. Latent growth curve modeling (LGCM) within a structural equation modeling framework will examine whether the quantity and type of opioid use predicts outcomes such as pain relief, quality of life, functional status, side effects, and psychiatric disorders. All GEE and LGCM analyses will control for a range of confounding factors.

3.6 Discussion This project is the first large-scale Australian cohort study that will rigorously examine opioid analgesic use among chronic pain patients, and examine the relationship of opioid use to important health outcomes including mortality. This study will be the first to comprehensively examine the extent to which opioid therapy for chronic pain is associated with pain reduction, changed quality of life, mental and physical health outcomes, aberrant medication behaviour and substance use disorders.

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

The study will shed light on the extent to which patients experience problematic opioid use, some of the precursors and protective factors to problematic use, and the consequences of problematic opioid use related to chronic opioid therapy. It will lead to improved knowledge of dose changes over time, and the positive and negative outcomes associated with changes in patterns of opioid use over time.

Improved understanding of the longer-term outcomes of chronic opioid therapy will direct community-based interventions and health policy in Australia. Regulators across jurisdictions currently use different criteria for authorising long-term opioid therapy, and different criteria for identifying at-risk patients. The results of this study will assist doctors and regulators in Australia to better identify those patients who are at risk of adverse outcomes and who therefore require alternative treatment strategies.

Finally, the project will achieve the establishment of a cohort of Australians with chronic pain. The project will provide the groundwork for further follow-up of the sample to determine the longer-term outcomes for chronic pain patients..

Chapter three: The Pain and Opioids IN Treatment (POINT) study protocol

Additional file 1: Appendix E - Glossary of conditions that may lead to chronic pain

Additional file 2: Appendix F – General pain categories

Additional file 3: Appendix G – medications that may be used for pain conditions

Chapter four: Baseline characteristics of the POINT sample

4 PAPER THREE: THE PAIN AND OPIOIDS IN TREATMENT (POINT) STUDY:

CHARACTERISTICS OF A COHORT USING OPIOIDS TO MANAGE CHRONIC NON-CANCER

PAIN

Campbell, Ga., Nielsen, Sa,b., Bruno, Rd,a., Lintzeris, Nb,c., Cohen, Mf., Hall, We., Larance, Ba., Mattick, Ra. & Degenhardt, La,g-i. a. National Drug and Alcohol Research Centre, UNSW, Sydney, NSW 2052 Australia b. The Langton Centre, South East Sydney Local Health District (SESLHD) Drug and Alcohol Services, 591 South Dowling St, Surry Hills, NSW 2010, Australia c. University of Sydney, Camperdown, NSW 2050 Australia d. School of Medicine, University of Tasmania, Sandy Bay Campus, Hobart, Tasmania 7001, Australia e. Centre for Youth Substance Abuse Research, University of Queensland, Royal Brisbane and Women’s Hospital, Herston, Queensland 4029 AUSTRALIA f. St Vincent’s Clinical School, UNSW, Darlinghurst NSW 2010, Australia g. School of Population and Global Health, The University of Melbourne, Parkville, VIC 3010 Australia h. Murdoch Children’s Research Institute, The Royal Children's Hospital, Flemington Road Parkville, VIC 3052 Australia i. Department of Global Health, School of Public Health, University of Washington, 325 9th Avenue Seattle, WA 98104 USA

Paper three has been published in Pain (Campbell, Nielsen, Bruno, et al. 2015)

Chapter four: Baseline characteristics of the POINT sample

4.1 Copyright Statement I certify that this publication was a direct result of my research towards this PhD, and that reproduction in this thesis does not breach copyright regulations.

Campbell, G., Nielsen, S., Bruno, R., Lintzeris, N., Cohen, M., Hall, W., Larance, B., Mattick, R.P., & Degenhardt, L. (2015).The Pain and Opioids IN Treatment (POINT) study: Characteristics of a cohort using opioids to manage chronic non-cancer pain. Pain, 156(2), 231-42

Gabrielle Campbell November, 2015

Chapter four: Baseline characteristics of the POINT sample

4.2 Preamble The paper presented in Chapter 3 provided an overview of the methodology of the POINT study. POINT includes a wide variety of measures to examine factors associated with CNCP treatment outcomes as recommended by the IMMPACT group (Dworkin et al. 2005; Turk et al. 2003; Turk et al. 2008). The paper presented in the current chapter describes the baseline characteristics of the POINT sample on a variety of these domains. Due to the impact of CNCP on employment, the paper divides the POINT sample into three age-groups informed by the Australian Government’s definitions of prime working age, namely 25-54 years(Vandenbroek 2013), and Australia’s current retirement age, namely 65 years (Australian Bureau of Statistics 2013). We included the “nearing retirement age” (55-64 years) due to the large number of participants in this group.

Chapter four: Baseline characteristics of the POINT sample

4.3 Abstract There has been a recent increase in public and professional concern about the prescription of strong prescription opioids for pain. Despite this concern, research to date has been limited because of a number of factors such as small sample sizes, exclusion of people with complex comorbidities, and studies of short duration. The Pain and Opioids IN Treatment is a two-year prospective cohort study of 1,500 people prescribed with pharmaceutical opioids for their chronic pain. This article provides an overview of the demographic and clinical characteristics of the cohort using the baseline data of 1,514 community-based people across Australia. Participants had been in pain for a period of 10 years and had been on prescription opioids for approximately four years. One in 10 was on a daily morphine equivalent dose of 200 mg. Employment and income levels were low, and two-thirds of the sample reported that their pain had impacted on their employment status. Approximately 50% screened positive for current moderate-to-severe depression, and one in five had made a lifetime suicide attempt. There were a number of age-related differences. The younger groups experienced higher levels of pain and pain interference, more mental health and substance use issues, and barriers to treatment, compared with the older group. This study found that the people who have been prescribed strong opioids for chronic pain have very complex demographic and clinical profiles. Major age-related differences in the experiences of pain, coping, mental health, and substance use suggest the necessity of differential approaches to treatment.

Chapter four: Baseline characteristics of the POINT sample

4.4 Introduction Chronic non-cancer pain (CNCP) is a common complaint that makes a major contribution to disease burden. In 2010, the Global Burden of Disease study estimated that tension-type headache affected 21% of the global population (equivalent to 1.4 billion people) and that 15% suffered from migraines (1.0 billion), 9% suffered from lower back pain (630 million people), and 5% from neck pain (332 million people)(Vos et al. 2012). Low back pain, neck pain, and migraine were the first, fourth and eighth largest contributors, respectively, to global nonfatal health burden (years lived with disability) (Vos et al. 2012).

Chronic non-cancer pain has a major impact on the individual in terms of quality of life, mental health, health status, relationships, and employment (Blyth et al. 2001; Breivik et al. 2006; Breivik, Eisenberg, and O'Brien 2013). Additionally, the economic costs associated with CNCP are very substantial. For example, in the United States, it is estimated that chronic pain affects 100 million adults at a cost of between USD 560 and 635 billion annually (Gaskin and Richard 2012). In 2007, the cost of chronic pain was estimated to be AUD 34.3 billion in Australia (Access Economics 2007). In Europe, the economic burden of chronic pain has been estimated to be 3% to 10% of gross domestic product (Breivik, Eisenberg, and O'Brien 2013). The costs associated with chronic pain and associated conditions have been estimated to be greater than those of established health care priorities, such as cardiovascular disease and cancer (Breivik, Eisenberg, and O'Brien 2013). With the aging of the population in many developed high-income countries, the burden of chronic pain is likely to increase. It is important to identify these patients in terms of socioeconomic and clinical characteristics to tailor treatment to address the severity and prevalence of CNCP, if not also preventative programs. As research has found age-related differences in terms of the experience of pain and pain interference (Corran et al. 1997; Coughlan et al. 1995; Wittink et al. 2006), it is important that these differences are dissected further.

Chapter four: Baseline characteristics of the POINT sample

To date, research into chronic pain has been limited mainly to clinical trials, which are often of limited duration and have excluded patients with complex comorbidities such as substance use problems and other complications. Smaller studies have focused on specific pain conditions or specific groups of people and have focused on people in treatment settings. Larger studies have not provided details of the clinical characteristics of people with chronic pain. The Pain and Opioid IN Treatment (POINT) study aims to document patterns of prescribed pharmaceutical opioid use, clinical outcomes, and risk of adverse events prospectively in a cohort of patients with CNCP. It is the first Australian study of a large non-clinic community sample to comprehensively examine the extent to which opioid therapy for chronic pain may be associated with pain reduction, improved quality of life, and favourable mental and physical health outcomes.

This article describes the cohort at the baseline interview. The aims of this article are to:

1. Describe the sociodemographic, physical health, mental health, and substance use profile of the cohort

2. Describe the experience of current pain in the cohort

3. Provide a brief overview of health service and treatment use by the cohort

4. Examine potential differences across these various domains according to age.

4.5 Methods Full details of the study design have been published elsewhere (Campbell et al. 2014). The study was approved by the Human Research Ethics Committee of the University of New South Wales (HREC reference: # HC12149). The study also received A1 Australian National Pharmacy Guild Approval to approach pharmacists to assist with recruitment of participants (Approval n. 815).

Chapter four: Baseline characteristics of the POINT sample

4.5.1 Recruitment

Pain and Opioid IN Treatment participants were 18 years or older, competent in English, mentally and physically able to partake in telephone and self-complete interviews, without serious cognitive impairments, living with CNCP, and taking prescribed schedule 8 opioids for CNCP for more than six weeks. Schedule 8 is an Australian classification of drugs of dependence that are subject to additional regulatory controls regarding their manufacture, supply, distribution, possession, and use (Therapeutic Goods Administration 2013). Schedule 8 opioids include morphine, oxycodone, buprenorphine, methadone, and hydromorphone. A history of injecting drug use was not an exclusion criterion, but people currently prescribed with pharmaceutical opioids for opioid substitution therapy for heroin dependence or for cancer were not eligible.

Total number of pharmacies n=5745

Pharmacies contacted n=5332 (93% of pharmacies)

Pharmacies agreed to recruitment n=1868 (33% of pharmacies)

Figure 4-1: Pharmacy recruitment

From a database of 5,745 community pharmacies, 1,868 pharmacies were willing to refer potentially eligible participants to the study (see (Campbell et al. 2014) for further detail). In total, 35% of pharmacies across Australia agreed to participate

Chapter four: Baseline characteristics of the POINT sample

(Figure 4-1). Of those pharmacies that agreed to participate, 36% of pharmacists sent through details of eligible participants to the POINT team.

Patients who fit the above criteria were given a flyer about the study by the pharmacist. Potential participants could opt to have the pharmacist send their details to researchers, or they could contact the research team directly. All flyers had a unique pharmacy number, and pharmacists were reimbursed AUD 20 for each eligible participant they referred into the study. Pain and Opioid IN Treatment study staff made reminder calls to pharmacists every fortnight for a 6-week period.

Approximately 2,700 participants were referred into the study (Figure 4-2). Pain and Opioid IN Treatment study staff determined the eligibility of potential participants. Eligible participants (n= 2091) then underwent a voluntary informed consent process. Those who were willing to participate (n= 1873) underwent an initial telephone interview that took approximately 1 to 1.5 hours. Eighty-one percent of eligible participants (n= 1514) completed the baseline interview

Chapter four: Baseline characteristics of the POINT sample

Participants referred into study n=2725

Excluded (n=634) Could not contact n =407 Too unwell to participate n=149 Refused n=78

Participants assessed for eligibility n= 2091 (78% of participants referred to the study)

Ineligible (n=218) Did not speak English n=9 Prescribed opioids < 6 weeks n=137 Cancer n=46 On OST for dependence n=7 No chronic pain n=19

Eligible participants n= 1873 (90% of those assessed for eligibility)

Loss after eligibility Refused n=359 Booked for baseline but did not complete n =74

Completed baseline interview n=1514 (81% of eligible participants)

Figure 4-2: Participant recruitment flowchart

Telephone interviews were conducted by trained interviewers, who had a minimum 3- year health or psychology degree, had undergone suicide response training, and were provided glossaries of chronic pain medications and conditions. A self-completion survey was also sent to participants who were reimbursed AUD40 for the baseline interview.

4.5.2 Measures

The measures, tools, and domains collected were based on recommendations made under the auspices of the Initiative on Methods, Measurement, and Pain Assessment

Chapter four: Baseline characteristics of the POINT sample

in Clinical Trials (Dworkin et al. 2005; Turk et al. 2003). Full details of the measures used in the study have been reported elsewhere (Campbell et al. 2014).

4.5.2.1 Demographic characteristics

Because of the impact of chronic pain on employment and ability to work and previous findings of age-related differences in chronic pain (Blyth et al. 2001) (Corran et al. 1997; Coughlan et al. 1995; Turk, Okifuji, and Scharff 1995; Wittink et al. 2006), this study used age groups informed by the Australian government’s definitions of prime working age, namely 25 to 54 years, (Vandenbroek 2013), and Australia’s current retirement age, namely 65 years (Australian Bureau of Statistics 2013). We defined three groups: “working age” (in our sample, we included ages of 18-54 years), “nearing retirement age” (55-64 years), and “retirement age” (65 years and older). Net income was reported as more or less than AUD 400 per week, with below this cut-off being comparable with income from unemployment or disability benefits. Participants were asked whether their employment status had changed because of their pain condition, ie, became unemployed, change of role, reduced hours, which was recorded as a dichotomous outcome.

4.5.2.2 Pain and pain-related measures

Current pain severity and pain interference were measured by the Brief Pain Inventory (Cleeland 1991), as a continuous score out of 10. The Pain Self-efficacy Questionnaire (Nicholas 2007; Nicholas, Asghari, and Blyth 2008) is a 10-item questionnaire that results in a score out of 60, with higher score indicating better pain-self efficacy. A cut- off of 30 was used in the current study, as scores below 30 have been shown to be indicative of less sustainable gains and predictive of lower rate of return to work and/or maintenance of treatment gains (Coughlan et al. 1995).

Chapter four: Baseline characteristics of the POINT sample

4.5.2.3 Physical health

The Short Form-12 is based on population norms, with a mean score of 50 and a SD of 10. A cut-off of two SD below the population mean was used to indicate severity in this study.

The Washington Group Scale (Centre for Disease Control and Prevention 2006) was used to measure disability. This scale provides a total score from 0 to 18. Each question has a response from 0, “no difficulty,” to 3, “cannot do at all.” In this article, we analysed these disabilities as continuous data with a higher score indicating greater disability.

Two items from the World Health Organization Quality of Life Questionnaire (WHOQOL-Bref) (Skevington, Lotfy, and O'Connell 2004) were used as measures of quality of life: “How would you rate your quality of life,” dichotomized to very poor or poor and neither good nor poor or good/very good, and “how satisfied are you with your health?,” dichotomized to very dissatisfied/dissatisfied and neither satisfied nor dissatisfied/satisfied/very satisfied.

Sleep patterns were measured using the Medical Outcomes Study Sleep scale. The Sleep Problems Index (SLP-9) subscale (Spritzer and Hays 2003) uses nine items from the scale to gauge overall sleep quality with a higher score reflecting poorer self- reported sleep outcomes (Spritzer and Hays 2003).

Social support was measured by the Medical Outcomes Study; the Social Support Scale (Sherbourne and Stewart 1993) score is based on a converted average score out of 12 questions. The higher the average score, the higher degree of social support.

Chapter four: Baseline characteristics of the POINT sample

A body mass index above 30 was considered to define obesity that was likely to have an impact on a person’s health (World Health Organization 1998).

4.5.2.4 Mental health and substance use

Depression and generalized anxiety disorder were measured by the Patient Health Questionnaire-9 (PHQ) and Generalized Anxiety Disorder-7 (GAD-7) modules of the Patient Health Questionnaire (Kroenke et al. 2010). Previously validated cut-offs were used for screening tools as follows: symptoms indicating moderate-to-severe depression were defined as a score of ≥ 10 on the PHQ-9 (Kroenke, Spitzer, and Williams 2001), and symptoms of moderate-to-severe anxiety were defined as a score ≥ 10 on the GAD-7 (Spitzer et al. 2006).

A score of ≥ 3 on the Primary Care Posttraumatic Stress Disorder (PTSD) screen was used to indicate the presence of PTSD (Prins et al. 2004). Features of agoraphobia were identified by the Mini- International Neuropsychiatric Interview (Sheehan et al. 1998), and were recoded to a binary variable of current agoraphobia if both questions were answered in the affirmative. Panic attacks were also measured by the Mini- International Neuropsychiatric Interview (Sheehan et al. 1998). A binary variable was created based on first screener questions “Have you had an anxiety attack in past 4 weeks?” Participants were screened for potential International Classification of Diseases (ICD-10) diagnoses of borderline personality disorder using the National Survey of Mental Health and Wellbeing version of the Composite International Diagnostic Interview (Andrews et al. 1999).

Participants were asked about lifetime and past 12-month alcohol and illicit drug use. Lifetime drug and alcohol use disorders (using ICD-10 dependence criteria) were assessed through the Composite International Diagnostic Interview (World Health Organization 2001) alcohol and illicit drug use module.

Chapter four: Baseline characteristics of the POINT sample

4.5.2.5 Medications

Oral morphine equivalent daily doses were estimated (Nielsen et al. 2014) using available references (Therapeutic Guidelines Limited 2013; Australian Medicines Handbook 2013; Faculaty of Pain Medicine ANZCA 2014) Self-reported opioid use was obtained from a medication diary completed over a one-week period as part of the self-complete questionnaire mailed to participants. A “high-risk” opioid dose variable was defined as more than 200 mg/d oral morphine equivalents (Chou, Fanciullo, Fine, Adler, et al. 2009; Nuckols et al. 2014). Participants were asked about current prescribed medications, with examples being given for each class of medications examined.

The Opioid Related Behaviours in Treatment scale (ORBIT) (Larance et al. 2015) scale is a 10-item scale designed to assist in the identification of behaviours relating to pharmaceutical opioids that may reflect problems with treatment, including diversion and nonadherence. ORBIT scores were converted into a dichotomous variable based on endorsement of at least one item in the past three months.

Adverse events were based on the Pain Assessment and Documentation Tool (PADT) (Passik et al. 2004). Participants were asked whether they had experienced the following side effects from their opioid medication: nausea, vomiting constipation, itching, mental cloudiness, sweating, fatigue, drowsiness, sexual dysfunction, other. An endorsement of one adverse event was used in this study.

4.5.2.6 Health-service utilisation

Participants were asked which services they had accessed for their pain in their lifetime and past month, including general practitioner, physiotherapist, chiropractor, psychologist, psychiatrist, counsellor, pain specialist, and acupuncturist. Participants were also asked the total number of health professionals they had ever seen about their pain. Participants were asked about any barriers to treatment they had

Chapter four: Baseline characteristics of the POINT sample

experienced, these questions being based on work by Glajchen (Glajchen 2001), and included other additions deemed relevant by the POINT team

4.5.3 Data analysis

All analyses were conducted using STATA, version 12.0. Based on definitions of the Australian Bureau of Statistics and the Australian Government, we defined 3 groups: (1) Working age (in this sample, ages from 19 to 54 years) (Vandenbroek 2013), (2) Nearing retirement age (55-64 years), categorized as those not encompassed in the “prime working” or the retirement definitions (3) Retirement age (65 years and older)(Australian Bureau of Statistics 2013).

Multinomial regression was used to compare the three age groups. Odds ratios and 95% confidence intervals were used for proportions. Means, SDs, and t tests were used where data were normal; medians, interquartile ranges, and nonparametric statistics (Mann-Whitney U) were used to compare groups where the distribution was non- normal.

4.6 Results 4.6.1 Socio-demographic characteristics

The POINT study included 1,514 people, 44% men with a mean age of 58 years (SD 5 13.7). Approximately one in 14 was employed full time, with approximately two-thirds of those in the working and nearing retirement age groups unemployed. Nearly two- fifths of the participants in the working age and nearing retirement group reported that their employment status had changed because of their pain condition. Almost half the sample reported a weekly income of AUD399 or less (equivalent to the single aged and disability pension). This proportion was highest in the retirement age group (64.2%) (Table 4-1).

Table 4-1: Demographic characteristics of the POINT cohort

Total A. Retirement B. Nearing C. Working B vs. A (ref) C vs. A (ref) age (65 years retirement age and older) (55-64 years) (19-54 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) Median age (IQR)# 58 (48-67) 71 (68-77) 59 (57-62) 45 (39-51) na na Male (%) 44.4 38.9 47.8 46.55 1.44 (1.10-1.88)** 1.37 (1.08-1.74)* % born in Australia 79.7 75.7 79.4 83.0 1.24 (0.90-1.69) 1.57 (1.17-2.10)** % Aboriginal and/or Torres Strait Islander 2.7 0.4 2.7 4.5 6.85 (1.51-31.09)* 11.45 (2.71-48.34)** % highest education level - completed high 45.4 45.4 42.8 47.0 0.90 (0.69-1.17) 1.07(0.84-1.35) school % highest education level - tertiary qualifications 35.3 27.4 34.6 41.9 1.40 (1.05-1.86)* 1.92 (1.49-2.48)*** % married or de facto 53.6 58.5 53.5 49.9 0.82 (0.63-1.06) 0.71 (0.56-0.89)** Employment status % employed full-time 7.4 1.4 6.5 12.8 4.76 (2.04-11.09)*** 10.0 (4.64-22.18)*** % unemployed 48.8 10.0 68.9 66.3 19.67 (13.68-28.29)*** 17.53 (12.50-24.60)*** % retired 31.0 83.6 13.7 0.8 0.03 (0.02-0.04)*** 0.00 (0.006-0.003)*** Self-reported change in employment due to pain 63.9 36.9 74.6 78.0 5.03 (3.76-6.73)*** 6.07 (4.66-7.90)*** % % weekly income $800 or more 11.8 5.3 11.2 17.3 2.24 (1.36-3.70)** 3.73 (2.39-5.83)*** $400 - 799 29.4 29.0 28.4 30.3 0.97 (0.72-1.30) 1.06 (0.82-1.38) $399 or less 54.0 64.2 53.7 46.1 0.64 (0.49-0.85)** 0.47 (0.37-0.61)*** Housing Own 53.9 70.6 53.0 40.3 0.59 (0.38-0.67)*** 0.28 (0.22-0.36)*** Rent (private) 23.1 12.1 20.7 33.2 1.93 (1.34-2.77)** 3.62 (2.64-4.99)*** Rent (public housing) 13.5 9.4 15.9 15.0 1.82 (1.21-2.73)** 1.71 (1.18-2.49)** % Live: alone 24.6 31.9 24.9 18.8 0.71 (0.53-0.95)* 0.49 (0.37-0.65)*** spouse 52.5 55.2 53.2 49.9 0.92 (0.71-1.20) 0.79 (0.64-1.02) children 23.1 8.0 13.4 41.1 1.79 (1.16-2.76)** 8.04 (5.59-11.58)*** *** P< 0.001; ** P<0.01; * P<0.05 1. # Mann Whitney U 92

Chapter four: Baseline characteristics of the POINT sample

4.6.2 Pain and physical functioning

The median duration of living in pain was 10 years (interquartile range 4.5-20; Table 4- 2) and the main pain condition reported by the majority of the sample was back/neck problems. Arthritis was more likely to be reported by the retirement age group than in the other two groups. The nearing retirement and the working age groups were more likely to report frequent or severe headaches than the retirement age group. Approximately 85% of the sample reported more than one lifetime pain condition, with fewer pain conditions reported by the working age group than the retirement age group.

The working and the nearing retirement age groups reported greater pain severity than the retirement age group. Similarly, mean pain interference scores for the sample were higher in the working and the nearing retirement age groups than in the retirement age group. A pain self-efficacy score below 30 was reported by a greater proportion in the working and the nearing retirement age groups than in the retirement age group. Almost four-fifths of the sample reported experiencing incident pain (a relative increase in pain due to overuse of parts of the body); this prevalence was higher in the working and the nearing retirement age groups compared with the retirement age group.

Table 4-2: Pain conditions and physical functioning of the POINT cohort

Total A. Retirement aged B. Nearing C. Working age B vs. A (ref) C vs. A (ref) (65 years and older) retirement age (19-54 years) (55-64 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) Median years living with pain# (IQR) 10 (4.5-20) 13 (4.4-25) 10 (6-23) 10 (4-17) z=0.79 z=4.59*** % pain conditions past 12-months Back or neck problems 76.4 76.3 77.1 76.1 1.06 (0.77-1.45) 0.99 (0.75-1.30) Arthritis or rheumatism 61.6 77.1 68.2 45.3 0.64 (0.47-0.86)** 0.25 (0.19-0.32)*** Frequent/severe headaches 29.4 16.6 26.4 41.3 1.80 (1.30-2.50)*** 3.53 (2.65-4.71)*** Visceral pain 23.7 21.3 22.1 26.7 1.05 (0.76-1.45) 1.34 (1.01-1.78)* Fibromyalgia 5.9 3.7 7.0 6.9 1.95 (1.07-3.60)* 1.94 (1.07-3.41)* Complex regional pain syndrome 3.2 1.2 3.0 4.8 2.48 (0.92-6.65) 4.07 (1.68-9.86)** Shingles-related pain 1.5 2.0 2.2 0.5 1.09 (0.44-2.72) 0.23 (0.06-0.84)* Other 31.3 28.0 31.6 33.7 1.18 (0.89-1.58) 1.30 (1.01-1.69)* % More than 1 lifetime pain condition 84.7 89.0 86.8 79.9 0.82 (0.54-1.22) 0.49 (0.35-0.70)*** Current pain Pain Severity (BPI)## (SD) 5.1 (1.8) 4.6 (1.9) 5.2 (1.8) 5.3 (1.6) 1.21 (1.13-1.32)*** 1.26 (1.19-1.36)*** Pain Interference (BPI)## (SD) 5.7 (2.3) 4.7 (2.3) 6.1 (2.1) 6.2 (2.1) 1.31 (1.23-1.40)*** 1.35 (1.28-1.43)*** Pain Coping and Self Efficacy 48.4 30.3 54.0 59.1 2.70 (2.05-3.56)*** 3.32 (2.59-4.27)*** under 30 % reporting incident pain 79.9 69.3 82.7 86.17 2.12 (1.47-3.08)*** 2.76 (1.96-3.90)*** Median number of other chronic 0.5 (0-1) 1 (0-2) 1 (0-1) 0 (0-1) z=0.29 z=4.87*** health problems 12-months (IQR) Median other health conditions- 1 (1-2) 2 (1-3) 1 (1-2) 1 (0-2) z=3.36*** z=11.93*** lifetime (IQR) % BMI obese category 39.8 35.2 42.5 41.7 1.36 (1.04-1.79)* 1.32 (1.03-1.69)* Median Washington disability scale 4 (3-6) 4 (2-5) 4 (3-6) 4 (3-6) z=-3.20** z= - 1.93 score (IQR) % SF 12 physical health scale below 30 55.4 59.3 59.0 49.9 0.98 (0.75-1.28) 0.68 (0.54-0.87)* % WHO QOL poor/very poor health? 23.8 15.3 27.1 28.4 2.05 (1.48-2.86)*** 2.19 (1.62-2.96)***

Table 4-2 (cont): Pain conditions and physical functioning of the POINT cohort

Total A. Retirement aged B. Nearing C. Working age B vs. A (ref) C vs. A (ref) (65 years and older) retirement age (19-54 years) (55-64 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) % WHO QOL dissatisfied/very 32.8 23.5 37.6 36.9 1.96(1.46-2.48)*** 1.90 (1.46-2.48)*** dissatisfied with health% Sleep score (SLP) 46.7 (31-62) 36.1 (24-51) 47.8 (33-63) 56.1 (40.6-68.9) Z=-11.88*** Z=-6.81*** MOS social support score 3.3 (2.3-4) 3.6 (3-4.3) 3.6 (2.3-4.7) 3.1 (2.1-4) Z=6.23*** Z=5.0*** *** P< 0.001, ** P<0.01, * P<0.05 # Mann Whitney U ## Mean (SD)

95 Chapter four: Baseline characteristics of the POINT sample

The retirement age group had higher median current and lifetime other health problems than the working age group and more median lifetime other health problems than the nearing retirement age group. Participants reported a median of 180 days over the past year of inability to perform work or general daily activities. This trend was more pronounced in the working and nearing retirement age groups, compared with the retirement age group.

The retirement age group reported more disability compared with the nearing retirement group and lower physical functioning scores compared with the working age group. The working and nearing retirement age groups reported higher rates of self-reported poor health and were more dissatisfied with their health compared with the retirement age group. The working and nearing retirement age groups scored higher on the sleep problems scale compared with the retirement age group. The working and nearing retirement age groups also reported lower levels of social support.

4.6.3 Mental health and substance use history

There were high rates of current mental health problems in the sample. Almost half the sample met criteria for moderate/severe depression (in the preceding 2 weeks) and 1 in 4 met criteria for current moderate/severe anxiety and agoraphobia. Both the working and nearing retirement age groups had higher rates of moderate/severe depression/anxiety, social phobia, agoraphobia, PTSD, panic attacks, and borderline personality disorder than the retirement age group (see Table 4-3). Furthermore, over half the sample reported a history of childhood abuse and/or neglect, the most common example being emotional abuse. The working and nearing retirement age groups had higher rates of each abuse/ neglect type compared with the retirement age group. Just under half the sample reported lifetime suicidal ideation, with one in five reporting a lifetime attempt; again this trend was more pronounced in the working and nearing retirement age groups compared with the retirement age group. Over 30% of the sample had a lifetime alcohol use disorder, and just over 10% had a

Chapter four: Baseline characteristics of the POINT sample

cannabis use disorder; the working and nearing retirement age groups had higher rates of these disorders than the retirement age group.

4.6.4 Opioid medication, adverse events, non-adherence and aberrant behaviours

The most commonly prescribed strong opioid was oxycodone, followed by buprenorphine. The working and nearing retirement age groups were more likely to be prescribed oxycodone compared with the retirement age group, whereas the retirement age group was more likely to be prescribed buprenorphine, compared with the younger groups. One quarter of the sample reported that they were also prescribed codeine; this trend was more prevalent in the working and nearing retirement age groups, compared with the retirement age group. Two-thirds of the sample reported also taking over-the-counter pain medication, and two fifths reported that they were also currently prescribed nonsteroidal anti-inflammatory drugs. The working age group was more likely to report also taking over-the-counter pain medication.

Table 4-3: Mental health and substance use history of the POINT cohort

Total A. B. Nearing C. Working B vs. A (ref) C vs. A (ref) Retirement retirement age age (65 years age (19-54 years) and older) (55-64 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) % moderate/severe depression 46.6 25.7 51.7 59.7 3.08 (2.33-4.09)*** 4.27 (3.29-5.52)*** % moderate/severe general anxiety 22.8 9.3 25.4 31.5 3.32 (2.25-4.89)*** 4.48 (3.13-6.41)*** disorder % social phobia 12.9 4.3 11.7 18.4 2.92 (1.71-4.98)*** 5.00 (3.09-8.10)*** % agoraphobia 22.2 8.6 25.1 31.0 3.57 (2.42-5.27)*** 4.77 (3.34-6.84)*** % post-traumatic stress disorder 15.3 5.3 13.2 24.6 2.70 (1.67-4.41)*** 5.79 (3.75-8.95)*** % panic attacks 17.5 6.1 21.4 23.9 4.16 (2.68-6.46)*** 4.81 (3.18-7.26)*** % borderline personality disorder 19.9 8.4 23.4 26.5 3.32 (2.04-5.42)*** 3.90 (2.46-6.17)*** screen+ SF12 mental health score below 30 14.3 6.5 18.9 17.3 3.33 (2.15-5.15)*** 2.99 (1.97-4.53)*** Reporting a history of childhood… Emotional abuse 37.5 23.1 39.6 47.5 2.18 (1.63-2.91)*** 3.01 (2.31-3.91)*** Sexual abuse 22.9 16.0 22.1 28.7 1.49 (1.07-2.10)* 2.12 (1.58-2.86)*** Physical abuse 29.3 18.8 29.6 37.4 1.81 (1.32-2.48)*** 2.57 (1.95-3.41)*** Physical neglect 12.5 7.2 12.4 16.7 1.84 (1.17-2.90)** 2.59 (1.73-3.89)*** Witness of physical violence 35.2 22.9 34.6 45.3 1.78 (1.32-2.39)*** 2.78 (2.14-3.62)*** % reporting lifetime suicidal… Ideation 45.9 24.0 50.0 60.4 3.17 (2.38-4.29)*** 4.83 (3.71-6.29)*** Plan 21.2 9.8 22.6 29.2 2.69 (1.84-3.95)*** 3.79 (2.69-5.35)*** Attempt 20.3 9.00 23.9 26.8 3.18 (2.16-4.66***) 3.70 (2.59-5.29)***

Table 4-3 (cont): Mental health and substance use history of the POINT cohort

Total A. B. Nearing C. Working B vs. A (ref) C vs. A (ref) Retirement retirement age age (65 years age (19-54 years) and older) (55-64 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) % Tobacco… Lifetime use 75.9 66.5 75.0 83.9 1.50 (1.12-2.03)** 2.61 (1.96-3.48)*** current daily use 31.1 11.7 29.6 47.4 3.18 (2.24-4.52)*** 6.81 (4.96-9.37)*** % Alcohol… use 92.8 88.0 93.0 96.3 1.81 (1.13-2.91)* 3.57 (2.18-5.88)*** use disorder 30.7 13.7 37.3 39.7 3.75 (2.70-5.20)*** 4.13 (3.05-5.61)*** used past 12m 60.5 60.1 57.0 63.1 0.88 (0.67-1.14) 1.13 (0.89-1.45) % cannabis… use 43.2 10.9 43.3 68.4 6.25 (4.42-8.86)*** 17.71 (12.72-24.68)*** use disorder 11.7 0.8 11.0 20.7 14.90 (5.30-51.85)*** 31.66 (11.61-86.32)*** used past 12m 12.9 0.6 10.5 24.1 18.90 (5.8-61.45)*** 51.37 (16.26-162.21)*** % amphetamine… use 13.9 1.6 11.1 25.5 7.46 (3.47-16.05)*** 20.47 (9.95-42.14)*** use disorder 5.6 0.6 3.5 10.8 5.84 (1.67-20.48)** 19.52 (6.10-62.46)*** used past 12m 1.6 0 1.0 3.2 na na % heroin… use 6.6 0.2 8.3 10.3 44.06 (6.00-323.64)*** 55.94 (7.73-404.76)*** use disorder 3.6 0 4.5 5.8 na na used past 12m 0.5 0 0.3 1.0 na na *** P< 0.001, ** P<0.01, * P<0.05 + of those that completed the screen n=978

Chapter four: Baseline characteristics of the POINT sample

Participants had been prescribed a strong opioid continuously for a median of 48 months (range, 19-120), with just fewer than 10% taking a morphine equivalent dose of 200 mg/d. More participants in the working (12.4%) and nearing retirement (9.2%) age groups were taking a morphine equivalent dose of 200 mg/d than in the retirement age group (4.3%).

Nearly the entire sample reported at least one side effect from their opioid medication at some point in their life. The most commonly reported side effects were constipation, drowsiness, and fatigue. Side effects from opioid medication were more likely to be reported in the working and nearing retirement age groups, compared with the retirement age group.

There were low rates of diversion to others (0.9%), use of another person’s opioid medication (2.3%), “doctor shopping” (0.9%), and injection (1.0%). The working and nearing retirement age groups were more likely to engage in non-adherent behaviours, compared with the retirement age group.

4.6.5 Treatment and health service utilisation

Participants reported a median of 12-months from the onset of pain to first prescription of opioids. The working and nearing retirement age groups had a shorter time to prescription than the retirement age group. Over half of all participants were currently prescribed an antidepressant, and approximately two-thirds were concurrently prescribed a benzodiazepine. Both the working and the nearing retirement age groups were more likely to be concurrently prescribed benzodiazepines, antidepressants, and antipsychotics than the retirement age group (Table 4-5).

100

Table 4-4: Opioid medications, adverse events, non-adherence and aberrant behaviours of the POINT cohort Total A. Retirement B. Nearing C. Working B vs. A (ref) C vs. A (ref) aged (65 years retirement age age and older) (55-64 years) (19-54 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) % Opioids currently prescribed Oxycodone 61.3 52.8 60.0 68.9 1.34 (1.02-1.75)* 1.97 (1.54-2.53)*** Morphine 15.1 11.5 19.4 15.3 1.86 (1.28-2.70)** 1.39 (0.97-1.98) Buprenorphine 21.3 34.0 19.4 12.5 0.47 (0.34-0.63)*** 0.28 (0.21-0.38)*** Fentanyl 14.5 15.3 15.4 13.3 1.01 (0.70-1.45) 0.85 (0.61-1.19) Methadone 4.1 1.2 2.7 7.2 2.26 (0.83-6.18) 6.26 (2.65-14.8)*** Hydromorphone 3.7 2.0 4.7 4.3 2.37 (1.09-5.17)* 2.17 (1.04-4.52)* Tramadol 9.8 7.2 10.7 11.2 1.55 (0.97-2.48) 1.64 (1.07-2.51)* Codeine + paracetamol 24.4 16.2 30.1 27.3 2.23 (1.62-3.08)*** 1.95 (1.45-2.62)*** % taking over the counter pain medication 62.8 59.9 60.0 66.8 1.00 (0.76-1.31) 1.34 (1.05-1.71)* % currently prescribed NSAIDS 39.0 25.6 42.2 39.1 1.34 (0.97-1.86) 1.16 (0.87-1.56) Median months before prescribed opioids 12 (1-96) 33 (2-167) 24 (1-84) 6 (0.5-48) Z=2.33* Z=7.08*** (median-iqr) % Morphine equivalent over 200mg 8.9 4.3 9.2 12.4 2.25 (1.29-3.93)** 3.14 (1.90-5.17)*** Median OME (IQR)#^ 75 (37-145) 50 (25-90) 77 (42-154) 90 (45-180) Z=-5.35*** Z=-7.29*** Median months of opioid prescribing# 48 (19-120) 48 (18-120) 60 (24-132) 48 (24-120) z=-2.87** z=-1.56 % reporting lifetime side effects from 86.9 78.9 90.6 90.9 2.55 (1.72-3.80)*** 2.64 (1.87-3.75)*** medication#

101

Table 4-4 (cont): Opioid medications, adverse events, non-adherence and aberrant behaviours of the POINT cohort Total A. Retirement B. Nearing C. Working B vs. A (ref) C vs. A (ref) aged (65 years retirement age age and older) (55-64 years) (19-54 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) % past 3mth to prescribed opioid related behaviours (ORBIT scale) asked for increase in dose 21.1 16.4 22.1 24.2 1.45 (1.04-2.03)* 1.64 (1.21-2.21)** asked for early renewal of prescription 12.3 3.1 11.4 20.1 4.10 (2.24-7.43)*** 7.93 (4.57-13.74)*** used another person’s opioids 2.3 0.4 2.0 3.9 4.94 (1.04-23.4)* 9.75 (2.29-41.5)** stockpiled opioids 7.9 2.3 9.2 11.6 4.40 (2.22-8.75)*** 5.68 (2.98-10.84)*** went to an additional doctor for opioids 0.9 0.2 1.5 1.1 7.39 (0.89-61.7) 5.54 (0.68-45.2) had prescription lost or stolen 4.2 0.6 3.7 7.2 6.28 (1.80-21.84)** 12.61 (3.90-40.84)*** diverted opioids to others 0.9 0.8 0.0 1.4 na 1.78 (0.54-5.81) altered dose 6.3 3.7 4.2 9.8 1.16 (0.59-2.72 2.84 (1.66-4.87)*** took via a route other than prescribed 1.0 0.2 0.5 1.9 2.44 (0.22-27.0) 9.58 (1.24-73.96)* used for opioids for non-pain reasons 4.0 1.0 4.0 6.4 4.01 (1.46-11.05)** 6.64 (2.60-16.96)*** % score of 1 or more ORBIT 37.5 22.1 38.1 49.1 2.17 (1.61-2.91)*** 3.40 (2.61-4.44)*** % Past month found opioid very/extremely 49.2 51.9 49.8 46.7 0.91 (0.70-1.19) 0.81 (0.64-1.02) helpful ^ based on n=922 from returned medication diaries # Mann-Whitney U *** P< 0.001, ** P<0.01, * P<0.05 # these included; nausea, vomiting constipation, itching, mental cloudiness, sweating, fatigue, drowsiness, sexual dysfunction, other

102

Chapter four: Baseline characteristics of the POINT sample

A median of one health service had been accessed in the month before interview, most commonly a general medical practitioner. The working age group had accessed more health services in the month before interview when compared with the retirement age group (see Table 4-5). The working and nearing retirement age groups reported a higher median of number of health professionals ever accessed and lifetime enrolment in a pain management program, compared with the retirement age group. One-fifth reported that they had received surgery before they were prescribed opioids. Specifically, 22% of those who reported back or neck problems reported surgery before opioid commencement.

One in six participants reported experiencing barriers to treatment, with just under one-third of the sample reporting they “feared becoming addicted or dependent on their medication,” with no differences by age group. One-third of participants reported that they “were unable to afford other pain treatments, such as physiotherapy, massage, acupuncture, etc.” The working age group reported more barriers to treatment than the retirement age group. The nearing retirement age group reported more problems with access to services and affordability than the retirement age group.

103

Table 4-5: Other prescription medication and health service use of the POINT cohort Total A. Retirement B. Nearing C. Working age B vs. A (ref) C vs. A (ref) aged (65 years retirement age (19-54 years) and older) (55-64 years) (1514) (n=489) % (n=402) % (n=623) % OR (95%CI) OR (95%CI) Benzodiazepines 34.0 21.1 35.6 43.2 2.07 (1.54-2.79)*** 2.84 (2.18-3.72)*** Antidepressants 51.8 41.3 56.5 57.0 1.84 (1.41-2.41)*** 1.88 (1.48-2.39)*** Antipsychotics 6.7 1.4 6.2 11.2 4.56 (1.95-10.67)*** 8.71 (3.70-19.13)*** Median number services ever accessed 4 (3-6) 4 (2-5) 5 (3-6) 5 (3-7) Z=-3.88** Z=-5.32*** about pain (median-iqr)# Median no. health professionals ever 6 (1-96) 4 (1-88) 6 (1-96) 8 (1-96) z=-6.32)*** z=-10.14*** seen about pain# (iqr) Median number services accessed before 1 (0-3) 1 (0-2) 1 (0-3) 2 (1-3) Z=-4.49*** Z=-9.50*** opioid treatment (iqr)# % Ever in pain management program (%) 20.8 13.1 25.9 23.6 2.31 (1.64-2.27)*** 2.05 (1.48-2.83)*** % Ever had surgery for pain condition 22.1 19.2 20.7 25.4 1.09 (0.79-1.52) 1.42 (1.07-1.91)* Median number health services accessed past month (out of 9#) month#…. 1 (1-2) 1 (1-2) 1 (1-2) 1 (1-2) Z=-1.06 Z= -3.94*** % reported more than one prescriber 14.7 14.3 13.2 16.1 0.91 (0.62-1.33) 1.14 (0.82-1.59) % Barriers to treatment Felt doc not confident prescribing 15.1 9.4 12.2 21.5 1.33 (0.87-2.05) 2.63 (1.84-3.78)*** opioids Communication problems with doc 15.0 9.0 12.7 21.2 1.47 (0.96-2.25) 2.71 (1.88-3.91)*** Felt doctor knew little about pain 12.4 7.4 10.5 17.5 1.46 (0.92-2.33) 2.67 (1.79-3.97)*** Fear you may become dependant 30.1 26.6 31.6 31.9 1.27 (0.95-1.71) 1.30 (1.00-1.68) Unable to get to doctor/pharmacy 12.0 3.3 10.5 19.7 3.44 (1.91-6.23)*** 7.27 (4.26-12.43)*** Unable to access specialist advice 15.0 7.4 15.7 20.6 2.34 (1.52-3.61)*** 3.25 (2.20-4.81)*** Unable to afford opioid medication 10.1 3.3 12.2 14.1 4.10 (2.30-7.33)*** 4.87 (2.81-8.40)*** Unable to afford other types of 17.0 7.2 21.1 22.2 3.47 (2.29-5.29)*** 3.69 (2.49-5.46)*** medication Unable to afford other pain treatments 32.8 21.7 34.6 40.5 1.90 (1.41-2.57)*** 2.45 (1.88-3.21)*** (i.e. counselling, physio, massage etc…) ^ based on n=922 from returned medication diaries # Mann-Whitney U *** P< 0.001, ** P<0.01, * P<0.05 ## included GP, physiotherapist, chiropractor, acupuncturist, psychiatrist, psychologist, counsellor, support groups, medical specialist 104

Chapter four: Baseline characteristics of the POINT sample

4.7 Discussion 4.7.1 The complexity of CNCP patient profiles

This study found that the people who have been prescribed strong opioids for CNCP have very complex demographic and clinical profiles. Consistent with previous research (Saastamoinen et al. 2005; Blyth et al. 2001), there were low rates of employment in the working and nearing retirement age groups, and the majority of the sample were on low weekly incomes, similar to rates of the aged or disability pension. In this study, approximately two-thirds of the sample reported that their pain had impacted on their employment status. Additionally, one in six reported some barriers to pain treatment, and one-third reported that they were unable to afford non-opioid prescription pain treatments. Our findings are consistent with other research that suggests that chronic pain has a major impact on working ability and that patients with CNCP experience significant socioeconomic disadvantage (Blyth et al. 2001).

Demonstrating the complexity of the cohort, most of the sample had experienced multiple pain conditions, reported poor physical health, and had experienced childhood abuse or neglect. Just under half met criteria for current moderate/severe depression, with a substantial minority meeting criteria for current moderate/severe anxiety or agoraphobia, reporting attempted suicide or alcohol or cannabis use disorder.

Although it is difficult in prospective studies to compare the cohort with other persons, this study supports recent Australian research that found that people dispensed opioid analgesics are in poorer health and have higher levels of psychological distress and poorer functioning, compared with those not receiving opioid analgesics (Rogers et al. 2013). This complex picture highlights the need for greater recognition of the social and psychological contributions to the experience of CNCP as well as indicating the

105

Chapter four: Baseline characteristics of the POINT sample

need for multifaceted (and multidisciplinary) health care approaches that address the numerous comorbidities seen in this group.

4.7.2 Differences across age-groups

This study also found large differences between the characteristics of the three age groups. The odds of frequent or severe headaches and complex regional pain syndrome were four times greater in the working age group compared with the retirement age group. It is interesting to note that effectiveness for the use of opioids in migraines has been found to be minimal (Tepper 2012) and there has been limited research into the use of prescription opioids for complex regional pain syndrome (Rowbotham 2006).

The retirement age groups seemed to have more economic, social, and physical disadvantages than the working and nearing retirement groups. Nonetheless, fewer participants in this group reported poor sleep, poor health, and health dissatisfaction compared with those in the other age groups; they also had better social support than the younger age groups. Furthermore, the retirement age group reported lower pain severity and pain interference scores and better pain and self-efficacy scores compared with the working and nearing retirement age groups. This finding is consistent with previous research (Rustoen et al. 2005). A possible explanation may be that older people perceive pain as part of being older and are therefore more accepting of it (Rustoen et al. 2005) (Molton and Terrill 2014; Terrill et al. 2014). Others have suggested that older people cope better with chronic pain because of higher levels of resilience, perceptions, and coping (Molton and Terrill 2014; Terrill et al. 2014).

The younger age groups were more likely to be prescribed oxycodone compared with the retirement age group, and the retirement age group was more likely to be prescribed buprenorphine patches. This possibility may reflect, consistent with the

106

Chapter four: Baseline characteristics of the POINT sample

clinical guidelines (Chou 2009) the safety profile of buprenorphine patches, where recent reviews have highlighted their favourable side effect profile with respect to respiratory depression (Plosker and Lyseng-Williamson 2012), which is particularly for useful in the elderly (Wolff et al. 2012) where polypharmacy and adverse greater effects of sedative medications are often reported.

Complex clinical profiles were more prevalent among the younger age groups (working and nearing retirement age groups). These groups reported more mental health problems, more experience of childhood abuse/neglect and lifetime suicidality, and more substance use than the retirement age group. These two groups were also prescribed higher doses of opioids, were more likely to also be prescribed codeine, and were likely to be taking concurrently prescribed benzodiazepines, antidepressants, and antipsychotics. Taken together, these characteristics suggest a very high-risk group, with multiple concomitant risk factors for overdose due to polypharmacy (Jones, Mack, and Paulozzi 2013; Nielsen et al. 2015). Although diversion and injection levels were low in the sample, the younger age groups were more likely to engage in non-adherent behaviours compared with the retirement age group. For a more detailed examination of diversion in the POINT cohort, please see Belcher, et al (2014) (Belcher et al. 2014).

Interestingly, although the younger age groups were more likely to have been enrolled in a pain management course and to have accessed more services, they also reported more barriers to treatment in terms of affordability and accessibility. Over two-fifths of the working age group and one-third of the nearing retirement age group were unable to afford non-opioid pain treatments. One of the challenges in providing effective treatment is accessibility to multidisciplinary pain management, including a range of treatments from allied health professionals which, unlike opioid medication, are largely not subsidized by the Australian health care system.

It is interesting that the mental health profile of the nearing retirement age group was more similar to the working age group, than to the retirement age group. This 107

Chapter four: Baseline characteristics of the POINT sample

similarity suggests that the impact of pain on an individual’s ability to work and generate income may increase the distress associated with chronic pain. While we are not able to establish cause and effect with this study design, further research is necessary to understand the relationships between chronic pain and employment.

There were also striking age differences in the initiation to prescription opioids, with the working age group prescribed with opioids six months after pain onset compared with approximately three years for the retirement age group. The younger age groups were also currently prescribed a higher median daily morphine equivalent dose and had greater proportions prescribed daily morphine equivalents of ≥ 200mg, compared with the retirement age group. These age differences may reflect the large expansion in opioid prescribing over the past two decades (Leong, Murnion, and Haber 2009b). Additionally, although the older people in our sample may have been much less likely to have been prescribed a strong opioid at the same age as the younger groups, they may also be prescribed lower doses based on their age, as recommended in the clinical guidelines (Chou 2009).

4.7.3 Limitations

Some limitations of this study need to be considered. A clear strength of the study is that all Australian community pharmacies were approached, and a significant proportion assisted with recruitment. We only have limited data, however, on those pharmacists and patients who did not participate. Another limitation is the potential biases that may be introduced by the reliance on self-report data. Information on chronic medical illnesses or other health problems was not verified through patient records. The rates of pain conditions and findings were similar to those in previous research. All participants were informed that their responses would be de-identified and confidential, which enhances the validity of self-reported substance use (Darke 1998). Furthermore, while there are differences in the way opioids are used across the different age groups, with the younger group being initiated on what appear to be more rapidly escalating doses of opioids after a median of six months with a pain

108

Chapter four: Baseline characteristics of the POINT sample

condition compared with more modest doses of opioids being initiated after almost three years for the older group, because of the cross-sectional nature of this study, we are unable to determine causality; however, through the longitudinal data collected from the POINT study, it is hoped that some of these questions may be answered.

4.7.4 Conclusions

The POINT study aims to examine the outcomes, beneficial and otherwise, of opioid prescribing for CNCP. This POINT cohort at baseline illustrates the complexity of patients with CNCP, with multiple social, mental health, and physical well-being issues. Furthermore, the major age-related differences in the experiences of pain, coping, mental health, and substance use suggest that more attention needs to be given to these factors in understanding the genesis of CNCP and differential approaches to assessment and treatment.

109

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

5 PAPER FOUR: PHARMACEUTICAL OPIOID USE AND DEPENDENCE AMONG PEOPLE

LIVING WITH CHRONIC PAIN: ASSOCIATIONS OBSERVED WITHIN THE PAIN AND

OPIOIDS IN TREATMENT (POINT) COHORT

Campbell, Ga., Nielsen, Sa., Larance, Ba, Bruno, Rb., Mattick, Ra, Hall, Wf., Lintzeris, Nc,d., Cohen, Me., Smith, K. a & Degenhardt, La,g-i. a. National Drug and Alcohol Research Centre, UNSW, Australia b. School of Medicine, University of Tasmania, Australia c. Sydney Medical School, Sydney University, Australia d. The Langton Centre, South East Sydney Local Health District (SESLHD) Drug and Alcohol Services, Australia e. St Vincent’s Clinical School, UNSW Medicine, UNSW Australia f. Centre for Youth Substance Abuse Research, University of Queensland, AUSTRALIA g. National Addiction Centre, Kings College, London ENGLAND h. School of Population and Global Health, University of Melbourne, Australia i. Murdoch Children’s Research Institute, AUSTRALIA j. Department of Global Health, School of Public Health, University of Washington, USA

Paper four has been published in Pain Medicine (Campbell, Nielsen, Larance, et al. 2015)

110

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

5.1 Copyright Statement I certify that this publication was a direct result of my research towards this PhD, and that reproduction in this thesis does not breach copyright regulations.

Campbell, G., Nielsen, S., Bruno, R., Larance, B., Smith, K., Hall, W., Cohen, M., & Lintzeris, N., & Degenhardt, L., (2015). Correlates of pharmaceutical opioid use and dependence among people living with chronic pain: Baseline findings from the Pain and Opioids IN Treatment (POINT) study. Pain Medicine 16(9) 1745-58.

Gabrielle Campbell November, 2015

111

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

5.2 Preamble The paper presented in Chapter 4 clearly demonstrated the complexity of people living with CNCP, and showed strong support for the biopsychosocial model. There were a number of socio-demographic problems, mental health comorbidities and physical health comorbidities identified in the cohort. The study also supports previous Australian research that found that people dispensed opioid analgesics are in poorer health, have higher levels of psychological distress and poorer functioning, compared to those not receiving opioid analgesics (Rogers et al. 2013). The term for this has been coined ‘adverse selection’ (Sullivan 2010). The paper presented in the current chapter examines the use of opioids and specifically problematic opioid use in the POINT cohort. Research suggests that higher opioid doses are associated with greater adverse outcomes such as overdose and problematic use. The current paper examines characteristics associated with dose using the following groups, less than 20mg OME, 21-90mgOME, 91-199mg OME and 200mg or greater OME. As stated earlier, a recent paper on the POINT study (Degenhardt et al. 2015), found that ICD-10 had the best fit in examining dependence in the sample and is therefore used a measure of dependence in the current paper.

112

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

5.3 Abstract Objective: There is increasing concern about the appropriateness of prescribing pharmaceutical opioids for chronic non-cancer pain (CNCP), given the risks of problematic use and dependence. This article examines pharmaceutical opioid dose and dependence and examines the correlates of each.

Design: Baseline data were obtained from a national sample of 1,424 people across Australia (median 58 years, 55% female and experiencing pain for a median of 10 years), who had been prescribed opioids for CNCP. Current opioid consumption was estimated in oral morphine equivalent (OME; mg per day), and ICD-10 pharmaceutical opioid dependence was assessed using the Composite International Diagnostic Interview.

Results: Current opioid consumption varied widely: 8.8% were taking <20mg OME per day, 52.1% were taking 21-90mg OME, 24.3% were taking 91-199mg OME and 14.8% were taking >=200mg OME. Greater daily OME consumption was associated with higher odds of multiple physical and mental health issues, aberrant opioid use, problems associated with opioid medication and opioid dependence. A significant minority, 8.5%, met criteria for lifetime ICD-10 pharmaceutical opioid dependence and 4.7% met criteria for past year ICD-10 pharmaceutical opioid dependence. Multivariate analysis found past-year dependence was independently associated with being younger, exhibiting more aberrant behaviours and having a history of benzodiazepine dependence

Conclusions: In this population of people taking opioids for CNCP, consumption of higher doses was associated with increased risk of problematic behaviours, and was more likely among people with a complex profile of physical and mental health problems.

113

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

Chronic non-cancer pain (CNCP) is common and burdensome. Low back pain, neck pain, and migraines were the first, fourth, and eighth largest contributors, respectively, to global nonfatal health burden (years lived with disability) (Vos et al. 2012). CNCP has a major impact on the individual in terms of quality of life, mental health, health status, relationships, and employment (Blyth et al. 2001; Breivik et al. 2006; Breivik, Eisenberg, and O'Brien 2013).

Short-duration randomized controlled trials have evaluated pharmaceutical opioids in the treatment of a range of CNCP conditions and have demonstrated modest attenuation of pain (Bloodworth 2005), but few studies have run for long enough to demonstrate long-term benefit of opioids for CNCP (Eisenberg, McNicol, and Carr 2006; Furlan et al. 2006; Manchikanti et al. 2011; Noble et al. 2010) such that there are reasons to question their clinical effectiveness. Despite the limitations of the evidence, there has been a considerable increase in the long-term prescribing of opioids for CNCP in a number of countries (Compton and Volkow 2006b; Degenhardt et al. 2006; Dhalla et al. 2009; Leong, Murnion, and Haber 2009a; Roxburgh et al. 2011). There has also been professional and public concern about concomitant increases in problematic opioid use and harms including dependence (Compton and Volkow 2006b, 2006a; Fischer and Argento 2012; Leong, Murnion, and Haber 2009b).

There are limited data on the patterns of opioid prescribing for individual patients over the longer term (Reid et al. 2010). Because they often exclude more complex patients and rarely follow up patients for long enough to capture behaviours that may be indicative of abuse or dependence (Lintzeris 2013), clinical trials typically find far lower rates of aberrant drug-related behaviours and opioid dependence than have been reported in some observational studies (Fishbain, Cole, Lewis, Rosomoff, and Rosomoff 2008).

114

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

The Pain and Opioid IN Treatment (POINT) cohort is designed to document patterns of pharmaceutical opioid prescribing and risk of adverse events in patients prescribed opioids for CNCP (Campbell et al. 2014). In this article, the aims were to:

1. Describe levels of current pharmaceutical opioid consumption, and the correlates of different levels of consumption; 2. Examine the prevalence of lifetime dependence and past year dependence as per ICD-10 criteria; and 3. Examine the correlates of the criteria met for pharmaceutical opioid dependence.

5.4 Methods 5.5.1 Study design and setting

The POINT study is a prospective cohort study of people across Australia who have been prescribed opioids for CNCP; the methodology (Campbell et al. 2014) and characteristics (Campbell, Nielsen, Bruno, et al. 2015) of this cohort have been described in detail elsewhere.

The study was approved by the Human Research Ethics Committee of the University of New South Wales (HREC reference: # HC12149). The study also received A1 Australian National Pharmacy Guild Approval to approach pharmacists to assist with recruitment of participants (Approval n. 815).

5.5.2 Eligibility criteria

Participants were eligible if they were: 18 years of age or older; fluent in English; mentally and physically able to complete telephone and self-complete interviews; without obvious cognitive impairment; living with CNCP; prescribed an opioid such as morphine, oxycodone, or fentanyl (Schedule 8 in the Australian classification of drugs of dependence; namely drugs that are subject to additional regulatory controls regarding manufacture, supply, distribution, possession, and use (Therapeutic Goods Administration 2013); and having taken such opioids for CNCP for more than 6 weeks. 115

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

A history of injecting drug use was not an exclusion criterion, but those currently prescribed methadone or buprenorphine as opioid substitution therapy for heroin dependence were not eligible. Those patients taking opioids for cancer pain were also excluded.

5.5.3 Measures

The measures, tools, and domains collected were based on recommendations of the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) (Dworkin et al. 2005; Turk et al. 2003). Full details of the measures used in the study have been reported elsewhere (Campbell et al. 2014); however, brief summaries of measures used in the current analyses are provided below.

5.5.3.1 Pain and pain-related measures

Participants were asked about lifetime and current history of pain conditions. Current pain severity and pain interference were measured by the Brief Pain Inventory (Cleeland 1991), as a continuous score out of 10. Relief from pain provided by current medications was measured as a continuous score from 0% to 100%. The Pain Self- Efficacy Questionnaire (Nicholas 2007; Nicholas, Asghari, and Blyth 2008) provided a score from 0-60, with higher scores indicating more confidence in managing life despite pain.

5.5.3.2 Mental health and substance use

Participants were asked if they had suffered from depression, anxiety or panic attacks or from post-traumatic stress disorder (PTSD) in the past year. Current depression and

116

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

generalized anxiety disorder were measured by the PHQ-9 and GAD-7 modules of the Patient Health Questionnaire (Kroenke et al. 2010). Validated cut-offs were used as follows: moderate to severe depression was defined as a score of ≥ 10 on the PHQ-9 (Kroenke, Spitzer, and Williams 2001); moderate to severe anxiety was defined as a score of ≥ 10 on the GAD-7 (Spitzer et al. 2006). Participants were screened for potential ICD-10 diagnoses of Borderline Personality Disorder (BPD) using the National Survey of Mental Health and Wellbeing version of the CIDI (Andrews et al. 1999). A score of ≥ 3 on the Primary Care PTSD screen (PC-PTSD) was used to indicate presence of PTSD (Prins et al. 2004).

Participants were asked about lifetime and past year use of alcohol and illicit drugs. Lifetime illicit drug, pharmaceutical opioid and alcohol use disorders (using ICD-10 dependence criteria, including tolerance and withdrawal) were assessed via the Composite International Diagnostic Interview (CIDI) (World Health Organization 2001) alcohol and illicit drug use module. A binary variable for overdose was created from the question “how many times in your life have you overdosed on any drug?”.

The Prescribed Opioids Difficulty Scale (PODS, Appendix N) was used to measure participants' current problems and concerns about using prescribed opioids (Sullivan et al. 2010). A cut-off of ≥8 was considered a moderate score in which participants would have endorsed at least two difficulties out of the 15 items (Sullivan et al. 2010). Scores on the Opioid Related Behaviours in Treatment scale (ORBIT, Appendix N), a 10-item measure of aberrant behaviours such as doctor shopping, diversion and other examples of unsanctioned use of medications, were converted into a dichotomous variable based on endorsement of at least one item (Larance et al. 2015).

5.5.3.3 Medications

Daily oral morphine equivalent (OME) doses for the opioids taken by the cohort were estimated following review and synthesis of a range of clinical guidelines (Nielsen et al.

117

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

2014). All Schedule 8, Schedule 4, and over-the-counter medications were included in the calculation. This was based on opioid use recorded in a one-week medication diary, which was completed as part of the self-complete questionnaire mailed to participants

Four groups were formed with the following cut-offs: 1-20mg/day OME, 21-90mg, 91- 199mg and more than 200mg, based on the previous research and guidelines (Manchikanti, Abdi, Atluri, Balog, Benyamin, Boswell, Brown, Bruel, Bryce, and Burks 2012; Zedler et al. 2014).

5.5.4 Data analysis

All analyses were planned a priori independent of the data. The analyses were conducted using STATA, version 12.0 (Stata Corporation, College Station, TX, USA). Means, standard deviations and t-tests were used where data were normal; medians, interquartile ranges, Hedges g for effect sizes with 95% Confidence Intervals were used to compare groups where the data were not normally distributed. OME groups were compared using multinomial logistic regression; relative risk ratios, and 95% confidence intervals were used for proportions. The category of 21–90 mg/day OME was used as the reference group in the multinomial regressions, as this range represented usual therapeutic doses and the largest group in the sample..

Two groups were defined according to whether or not they met ICD-10 criteria for pharmaceutical opioid dependence in the past year. Univariate logistic regression models were conducted to compare these groups. Relative risk and Hedges g are reported with 95% CI; a multivariate model was also conducted to determine the independence of associations, adjusted RR and 95% CI are reported. OME was not entered into the regression as we did not have an OME for the whole sample. Further, as all participants that scored moderate-high in the PODS had past 12-month dependence, this could not be included in the analysis.

118

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

5.5 Results Participants were recruited through community pharmacies. From a database of 5,745 community pharmacies, researches contacted 5,332 pharmacies across Australia (93%), and 1,868 pharmacies were willing to refer potentially eligible participants to the study (see (Campbell et al. 2014) for further detail). In total, 33% of pharmacies across Australia agreed to participate. Of those that agreed to participate, 36% of pharmacists sent through details of eligible participants to the POINT team. A random sample of pharmacists (n=71) was asked to collect data on the number and characteristics of all customers purchasing opioids during their six-week recruitment window. We found that of the total number of customers recorded as purchasing opioids in these pharmacies, 52% were female (the POINT cohort was 55% female); and 7% were 18-34 years, 55% 35-64 years and 38% 65+ years (vs. 5%, 62% and 33% respectively, in the POINT cohort). Of these customers, 63% were prescribed oxycodone (vs. 62% in the POINT cohort), 16.5% prescribed morphine (vs. 15% in the POINT cohort) fentanyl patches 21% (vs. 15% in the POINT cohort) and 24% buprenorphine patches (vs. 21% in the POINT cohort). These findings suggest a similarity between the POINT sample and people prescribed opioids for CNCP more generally.

The analysis of opioid consumption presented in this article is based on data from n=1,085 participants who returned medication diaries, from which OME mg daily doses could be determined. Analyses of pharmaceutical opioid dependence were based on interviews with n=1,424 participants who completed the CIDI to determine ICD-10 dependence.

Of the 1,424 participants who completed the CIDI, there were 299 participants who did not return their medication diary, so OMEs could not be calculated. Participants who did not return their diaries were more likely to be younger (53 vs 59, Hedges g 20.20, 95% CI 20.32–0.070), more likely to be male (52% vs 48%, RR 0.92 95% CI 0.87–0.98), unemployed (51.8% vs 42.4%, RR 1.40, 95% CI 1.14–1.71), have higher pain

119

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

interference (6.2 vs 5.6, RR 1.11, 95% CI 1.06–1.16), more past 12-month chronic conditions (2 vs 2, Hedges g, 0.24, 95% CI 0.12–0.37), were more likely to have past 12- month depression (49.2% vs 40.5%, RR 1.32 95% CI 1.08–1.61), anxiety (32.8% vs 25.1%, RR 1.34, 95% CI 1.08–1.65), and past month PTSD (15.7% vs 9.8% RR 1.51, 95% CI 1.16–1.96); and they were more likely to meet criteria for ICD-10 lifetime (13.1% vs 7.3%, RR 1.62 95% CI 1.22–2.14) and past 12-month (8.7% vs 3.7%, RR 1.90, 95% CI 1.38–2.62) pharmaceutical opioid dependence. These results indicate that the people who did not return their diary, and therefore not included in analyses of OME, had more complex comorbidities and this must be considered in the following results

5.6.1 Correlates of different levels of pharmaceutical opioid consumption

Participants were divided into four groups based on their daily OME dose: ‘less than 20mg’ group (n=95, 8.8% of the sample), ‘21-90mg’ group (n = 565, 52.1% of the sample), ‘91-199mg’ group (n=264, 24.3% of the sample) and ‘more than 200mg’ group (n=161, 14.8% of the sample, Table 5-1). Participants in the ‘91-199mg’ and ‘more than 200mg’ groups were younger than the ‘21-90mg’ group, whereas, those in the ‘less than 20mg’ group were more likely to be older than those in the ‘21-90 mg’ group.

Both the ‘91–199 mg’ and the ‘over 200 mg’ groups reported higher pain scores, higher pain interference, and lower pain self-efficacy scores than those in the ‘21–90 mg‘ group, and the ‘less than 20 mg’ group reported lower pain interference and had higher pain self-efficacy scores (see Figure 5-1). The ‘91–199 mg’ and “over 200 mg” groups reported less relief from their medications than the “21–90 mg” group. The ‘more than 200 mg’ group were less likely to report past year arthritis compared with the ’21–90 mg’ group, while the ‘91–199 mg’ group were more likely to report chronic back or neck problems than the ‘21–90 mg’ group.

120

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

Those in the ’over 200 mg’ group had 2.7 times the relative risk of being prescribed morphine compared with the ‘21–90 mg’ group; buprenorphine was less likely to be prescribed in the ‘91–199 mg’ and “more than 200 mg” groups compared with the ‘21– 90 mg’ group (Table 5-1). The “91–199” and ‘more than 200 mg’ groups were more likely to report a higher median number of prescribed current opioids, while the ‘less than 20 mg’ group reported a lower median number of prescription opioids than the ‘21–99 mg’ group. Side effects were more likely to be reported in the ‘91–199 mg” and “more than 200 mg” groups, while the ‘less than 20 mg’ group were less likely to report side effects than the ‘21–90 mg’ group (Table 5-1).

121

Table 5-1: Demographics and clinical characteristics of the POINT cohort (n=1,085) by mean daily opioid consumption (measured in oral morphine equivalent (OME) mg)

Total A. OME 1-20mg B. OME 21- C. OME 91- D. OME ≥200mg A vs B (ref) C vs B (ref) D vs B (ref) per day 90mg per day 199mg per day per day N=95 N=565 N=264 N=161 + + + N=1,085 RR, Hedges g RRR, Hedges g RRR, Hedges g (95%CI) (95%CI) (95%CI) Demographics Median age (IQR)# 59 (49-68) 67 (54-73) 61 (51.5-69.5) 58 (47-66) 53 (45.5-60) 0.06 (-0.16-0.27)+ -0.17 (-0.32--)0.02+ -0.27 (-0.45-0.10)+ % Male 42.8 (39.9-45.8) 35.8 (26.8-46.0) 40.0 (36.0-44.1) 45.8 (39.9-51.9) 52.2 (44.4-59.8) 0.83 (0.53-1.31) 1.27 (0.95-1.70) 1.63 (1.15-2.32)* % Unemployed 46.8 (43.8-49.8) 31.6 (23.0-41.6) 41.2 (37.2-45.4) 49.2 (43.2-55.3) 71.4 (63.9-77.9) 0.66 (0.41-1.05) 1.38 (1.03-1.85)* 3.56 (2.43-5.21)*** Pain and Physical Health Median years living with pain (IQR) + 11 (4.5-21) 11 (3-23) 10 (2-21) 11 (5-22) 15 (5-12) -0.06 (-0.27-0.15)+ -0.01 (-0.16-0.13) -0.14 (-0.31-0.04) Pain severity score M(SD) 5.0 (1.8) 4.4 (1.8) 4.8 (1.8) 5.4 (1.6) 5.4 (1.8) 0.89 (0.79-1.01) 1.21 (1.11-1.31)*** 1.21 (1.10-1.35)*** Pain interference M(SD) 5.6 (2.3) 4.7 (2.3) 5.3 (2.3) 6.1 (2.1) 6.2 (2.0) 0.89 (0.81-0.98)* 1.18 (1.09-1.26)*** 1.23 (1.13-1.34)*** PSEQ M(SD) 30.0 (13.2) 35.8 (13.8) 31.8 (12.9) 27.4 (12.7) 24.7 (12.5) 1.03 (1.01-1.04)** 0.97 (0.96-0.99)*** 0.96 (0.94-0.97)*** Relief from current medications Median (IQR)+ 7 (5-8) 7 (5-8) 7 (5-8) 6 (5-8) 6 (5-8) 0.02 (0.20-0.24)+ 0.16 (0.01-0.31)+ 0.19 (0.01-0.36)+ % Pain conditions in past years Arthritis or rheumatism 62.7 (59.7-65.5) 68.4 (58.4-77.0) 63.5 (59.5-67.4) 62.9 (56.9-68.5) 55.9 (48.1-63.4) 1.24 (0.78-1.97) 0.97 (0.72-1.32) 0.73 (0.51-1.04) Back or neck problems 76.3 (73.7-78.8) 68.4 (58.3-77.0) 74.2 (70.4-77.6) 81.1 (75.9-85.4) 80.7 (73.9-86.1) 0.75 (0.47-1.21) 1.49 (1.04-2.14)* 1.46 (0.95-2.26) Frequent/severe headaches 28.0 (25.4-30.8) 20.0 (13.1-29.3) 26.7 (23.2-30.5) 30.3 (25.0-36.1) 33.5 (26.7-41.2) 0.69 (0.40-1.17) 1.19 (0.86-1.64) 1.38 (0.95-2.02) Visceral 21.8 (19.5-24.4) 16.8 (10.5-25.8) 22.1 (18.9-25.7) 21.6 (17.0-27.0) 24.2 (18.2-31.5) 0.71 (0.40-1.26) 0.97 (0.68-1.38) 1.12 (0.75-1.70) Median 12m chronic pain conditions (IQR) + 2 (1-3) 2 (1-3) 2 (1-3) 2 (2-3) 2 (2-3) 0.12 (-0.09-0.34) -0.14 (-0.29-0.006) -0.12 (-0.29-0.05) Mental and physical health % Mod-severe depression 44.0 (41.0-46.9) 34.7 (25.8-44.9) 39.4 (35.5-43.6) 48.5 (42.5-54.5) 57.8 (50.0-65.2) 0.82 (0.52-1.29) 1.44 (1.08-1.94)* 2.10 (1.47-2.99)*** % Mod-severe anxiety 21.7 (19.3-24.3) 16.7 (10.3-25.9) 21.5 (18.3-25.2) 23.3 (18.6-28.9) 22.6 (16.7-29.9) 0.73 (0.40-1.32) 1.11 (0.78-1.58) 1.06 (0.69-1.63) SF12 physical health score Median (IQR) + 26.5 (22.3-31.0) 28.1 (24.3-34.0) 26.6 (22.4-31.4) 25.9 (21.6-31.6) 25.8 (22.3-29.0) -0.22 (-0.45—0.001) 0.13 (-0.02-0.28) 0.20 (-0.02-0.38) Medication use % Any past week Oxycodone 60.0 (57.0-62.9) 66.3 (56.2-75.1) 59.6 (55.6-63.6) 61.4 (55.3-67.1) 55.3 (47.5-62.8) 1.33 (0.84-2.10) 1.07 (0.80-1.45) 0.84 (0.59-1.19) % Any past week Morphine 14.8 (12.8-17.1) 7.3 (3.5-14.7) 12.0 (9.6-15.0) 15.2 (11.3-20.0) 28.6 (22.1-36.1) 0.58 (0.26-1.31) 1.31 (0.86-1.99) 2.92 (1.91-4.47)*** % Any past week Buprenorphine 22.6 (20.2-25.2) 33.7 (24.9-43.8) 31.3 (27.6-35.3) 11.4 (8.1-15.8) 3.7 (1.7-8.1) 1.11 (0.70-1.76) 0.28 (0.18-0.43)*** 0.08 (0.04-0.20)*** No. of current opioids MDN (IQR) 1 (1-1) 1 (1-1) 1 (1-1) 1 (1-1) 1 (1-2) 0.17 (-0.04-0.39) -0.11 (0.26-0.03) -0.50 (-0.68—0.33) Median years prescribed opioids (IQR) + 4 (1.6-10) 2.5 (0.6-5) 3 (1.1-8) 6 (2-13) 7.8 (3-15) 0.25 (0.04-0.47) -0.35 (-0.50—0.20) -0.52 (-0.70—0.35) % Current schedule 4 opioid use 31.6 (28.9-34.4) 17.9 (11.4-27.0) 31.9 (28.1-35.8) 35.6 (30.0-41.6) 32.3 (25.5-39.9) 0.47 (0.27-0.81)** 1.18 (0.87-1.61) 1.02 (0.70-1.48) % Current over-the-counter analgesic use 62.9 (60.0-65.8) 70.5 (60.6-78.9) 63.2 (59.1-67.1) 64.8 (58.8-70.3) 54.7 (46.9-62.2) 1.39 (0.87-2.24) 1.07 (0.79-1.45) 0.70 (0.49-1.00) No. adverse effects Median (IQR) + 4 (2-8) 2 (0-6) 4 (1-7) 5 (2-9.5) 6 (3-11) 0.34 (0..12-0.56) -0.26 (-0.41—0.12) -0.44 (-0.62—0.26) % Currently prescribed benzodiazepines 31.4 (28.7-34.3) 22.1 (14.8-31.6) 26.2 (22.7-30.0) 36.0 (30.4-42.0) 47.8 (40.2-55.6) 80.0 (0.48-1.34) 1.58 (1.16-2.17)** 2.58 (1.80-3.71)*** % Currently antidepressants 52.4 (49.5-55.4) 35.8 (26.7-46.0). 51.2 (47.0-55.3) 59.8 (53.8-65.6) 54.7 (46.9-62.2) 0.53 (0.34-0.84)** 1.42 (1.06-1.91)* 1.15 (0.81-1.64)

122

Table 5-1 (cont): Demographics and clinical characteristics of the POINT cohort (n=1,085) by mean daily opioid consumption (measured in oral morphine equivalent (OME) mg) Total A. OME 1-20mg B. OME 21- C. OME 91- D. OME ≥200mg A vs B (ref) C vs B (ref) D vs B (ref) per day 90mg per day 199mg per day per day N=95 N=565 N=264 N=161 + + + N=1,085 RR, Hedges g RRR, Hedges g RRR, Hedges g (95%CI) (95%CI) (95%CI) Substance use Illicit drug use past 12-months 12.7 (10.9-14.8) 9.5 (5.0-17.3) 10.3 (8.2-13.6) 14.0 (10.3-18.8) 21.1 (15.5-28.1) 0.91 (0.44-1.91) 1.42 (0.92-2.21) 2.34 (1.46-3.73)*** Past 12-month risky drinking (>5 standard 24.0 (21.6-26.7) 25.3 (17.5-35.0) 25.0 (21.6-28.7) 25.0 (20.1-30.6) 18.6 (13.3-25.4) 1.02 (0.62-1.68) 1.00 (0.72-1.40) 0.69 (0.44-1.07) drinks) Pharmaceutical opioid use behaviours % ICD-10 pharmaceutical opioid dependence- 7.1 (5.7-8.8) 0 5.0 (3.4-7.1) 8.4 (5.6-12.4) 16.9 (11.8-23.5) na 1.76 (0.99-3.14) 3.89 (2.22-6.82)*** lifetime % ICD-10 pharmaceutical opioid dependence – 3.6 (2.6-4.9) 0 2.3 (1.3-3.9) 5.3 (3.2-8.8) 7.5 (4.3-12.7) na 2.37 (1.10-5.13)* 3.42 (1.53-7.65)** Past 12m % past 3-month tampering 6.2 (4.9-7.8) 1.1 (0.1-7.2) 4.9 (3.4-7.1) 6.8 (4.3-10.6) 12.4 (8.1-18.5).5) 0.20 (0.3-1.52) 1.40 (0.76-2.59) 2.72 (1.49-4.97)** % past 3-month ‘doctor shopping’ 1.0 (0.6-1.8) 0 0.9 (0.4-2.1) 1.1 (0.4-3.5) 1.9 (0.6-5.6) na 1.29 (0.31-5.42) 2.13 (0.50-9.0) % past 3-month different route 0.7 (0.4-1.5) 0 0.2 (0.0-1.2) 0.8 (0.02-3.0) 3.1 (1.3-7.3) na 4.31 (0.39-47.9) 18.1 (2.10-156.5)** % used other person’s opioid medication past 2.2 (1.5-3.3) 1.1 (0.1-7.2) 1.9 (1.1-3.5) 1.5 (0.6-4.0) 4.9 (2.5-9.6) 0.54 (0.07-4.19) 0.77 (0.24-2.46) 2.63 (1.04-6.66)* 3-months % PODS intermediate-high (≥8) 61.1 (58.2-64.0) 42.1 (32.6-52.3) 57.8 (53.7-61.9) 68.5 (62.5-73.9) 72.4 (64.8-78.9) 0.53 (0.34-0.82)** 1.59 (1.16-2.17)** 1.92 (1.30-2.82)*** % Lifetime overdose on any substance 16.8 (14.7-19.1) 9.5 (5.0-17.3) 16.6 (13.8-19.9) 15.9 (12.0-20.8) 23.0 (17.1-30.1) 0.52 (0.25-1.08) 0.95 (0.64-1.41) 1.50 (0.98-2.30)

123

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

Participants in the ‘more than 200 mg’ group were more likely to screen positive for current moderate-to-severe depression than participants in the ‘21–90 mg’ group (Table 5-1). Participants in the ‘91–199 mg’ and ‘more than 200 mg’ groups were more likely to score intermediate-high on the PODS, indicating more problems with their prescription opioid use, compared with the ‘21–90 mg’ group, while the ‘less than 20 mg’ group scored lower on the PODS. The ‘91–199 mg’ group were more likely to engage in aberrant behaviours, as measured by the ORBIT scale, while the ‘less than 20 mg’ group were less likely to engage in aberrant behaviours than the ‘21–90 mg’ group. The ‘more than 200 mg’ were more likely to meet lifetime and past year criteria for ICD-10 pharmaceutical opioid dependence than the ’21–90 mg’ group. No one in the ‘0–20 mg’ group met ICD-10 criteria for lifetime or past year pharmaceutical opioid dependence (Table 5-1). Ad hoc analyses comparing the ‘91-199mg’ group and the ‘more than 200mg’ groups found no differences between the two in terms of pain severity, pain interference, time in pain, problems with opioids, aberrance, and dependence. Figure 1 presents associations and locally weighted scatterplot smoothing (LOESS) between OME and pain severity, pain medication relief, time on medication for pain, and pain and self-efficacy scores.

124

Lowess smoother Lowess smoother 1500 1500 1000 1000 totalopioiddose Total opioid dose 500 500 0 0

0 2 4 6 8 10 0 200 400 600 800 bpi average pain score Months continuously prescribed opioids bandwidth = .8 bandwidth = .8

Lowess smoother Lowess smoother 1500 1500 1000 1000 totalopioiddose totalopioiddose 500 500 0 0

0 2 4 6 8 10 0 20 40 60 how much relief did pain treatments pseq total score bandwidth = .8 bandwidth = .8

Figure 5-1: Scatterplot and LOWESS of OME by pain severity score, length of time on opioids, pain relief from medications and pain self-efficacy score

125

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

5.6.2 Correlates of ICD-10 pharmaceutical opioid dependence with past year symptoms

Among those who completed the CIDI (1,424), 8.5% (95% CI 7.2–10.1) met criteria for lifetime ICD-10 pharmaceutical opioid dependence and 4.8% (95% CI 3.8–6.0) met criteria past year dependence (Table 5-2). The median age of onset for ICD-10 opioid use disorder was 40 years (IQR, 32-49).

People who met criteria for past-year ICD-10 opioid dependence were more likely to be younger, male, unemployed, and less likely to be married or defacto (Table 5-2). Those who met criteria for past-year ICD-10 opioid use dependence were currently prescribed a higher median opioid consumption than those who did not. All participants who met criteria for past-year ICD-10 dependence also scored intermediate-to-high for problematic use associated with opioids. They were also more likely to engage in aberrant behaviours, compared with those who did not meet criteria for past year ICD-10 pharmaceutical opioid dependence. The most common aberrant behaviours among participants who met criteria for dependence were asking for an early script renewal (48.5%, n533), and asking doctor for an increase in dose (39.7%, n=27, Table 5-2).

126

Table 5-2: Demographic and clinical characteristics of the POINT cohort, according to past year ICD-10 pharmaceutical opioid dependence Total No Yes N=1417 N=1424 n = 1356 n=68 % (95%CI) % (95%CI) RR, hedges g+(95%CI) Adj RR (95%CI) Demographics Mean age (SD)+ 58 (48-67.5) 58.8 (22.2) 48.1 (10.6) 0.95 (0.93-0.97)*** 0.97 (0.95-0.99)** % male 44.4 (41.9-47.0) 43.8 (41.2-46.5) 57.4 (45.3-68.6) 1.68 (1.05-2.68)* 1.35 (0.83-2.20) % tertiary qualifications 35.2 (32.7-37.7) 35.2 (32.7-37.8) 35.3 (24.8-47.4) 1.00 (0.62-1.63) % married or de facto 53.3 (50.8-56.0) 53.6 (50.9-56.2) 48.5 (36.8-60.4) 0.82 (0.52-1.31) % Unemployed 48.9 (46.3-51.5) 47.6 (45.0-50.3) 73.5 (61.7-82.7) 2.90 (1.71-4.93)*** 1.31 (0.73-2.33) Pain -related factors Median no. pain conditions – 12m + 2 (2-3) 2 (2-3) 3 (2-3) -0.25 (-0.49- - 0.01) 1.13 (0.93-1.38) Median years living with pain + 10 (5-20) 10 (5-20) 10.5 (2-17.5) 0.04 (-0.27-0.36) + Pain severity score M(SD) 5.0 (1.8) 5.0 (1.8) 5.1 (1.6) 1.00 (0.89-1.15) Pain interference M(SD) 5.7 (2.3) 5.6 (2.3) 6.7 (1.8) 1.27 (1.12-1.43)*** 1.05 (0.91-1.21) PSEQ M(SD) 29.3 (13.5) 29.7 (13.5) 22.4 (12.3) 0.96 (0.94-0.98)*** 0.99 (0.97-1.01) Medication-related factors Median daily OME mg dose+ a 75 (36-145) 70.7 (35.0-142.5) 120 (77.9-232.9) -0.54 (-0.86—0.22) PODs intermediate-to-high scoreb 61.2 (58.4-64.0) 59.6 (56.7-62.4) 100 1 % 1 or more instance of non-adherence, past 38.3 (35.8-40.9) 36.7 (34.1-39.3) 72.1 (60.2-81.5) 4.15 (2.47-6.97)*** 2.47 (1.44-4.24)** 3-months1 Substance use % History of alcohol dependence 13.1 (11.5-15.0) 12.4 (10.7-14.3) 27.9 (18.5-39.8) 2.56 (1.55-4.23)*** 1.60 (0.58-1.90) % current daily smoking 31.2 (28.8-33.6) 29.9 (27.6-32.4) 55.9 (43.9-67.2) 2.96 (1.81-4.85)*** 1.61 (0.98-2.61) % history of illicit drug dependence 9.8 (8.4-11.5) 9.1 (7.7-10.7) 25.0 (16.1-36.7) 3.06 (1.82-5.14)*** 0.75 (0.35-1.58) % history of benzodiazepine dependence 2.2 (1.6-3.2) 1.7 (1.1-2.5) 13.2 (7.0-23.6) 6.64 (3.61-12.18)*** 3.22 (1.40-7.41)** Mental health % Past year … depression 42.3 (39.8-44.9) 41.0 (38.4-43.6) 69.1 (57.1-80.0) 3.05 (1.84-5.04)*** 1.48 (0.84-2.62) generalised anxiety disorder 26.7 (24.4-29.0) 25.8 (23.5-28.2) 44.1 (32.7-56.1) 2.17 (1.36-3.45)*** 1.02 (0.61-1.70) post-traumatic stress disorder 11.0 (9.5-17.8) 10.6 (9.1-12.4) 19.1 (11.4-30.3) 1.90 (1.07-3.41)* 0.96 (0.52-1.77) + Hedges G 95%CI 1. As measured via the ORBIT scale a OME was not included in multivariate analysis as the sample size would be reduced from 1417 to 1081. b PODS score was not included in the multivariate regression as all people who scored intermediate to moderate on the PODS also met criteria for past 12-month pharmaceutical opioid dependence 127

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

Participants who met criteria for past year dependence were more likely also to meet criteria for ICD-10 dependence on alcohol, benzodiazepines, and other illicit drugs (Table 5-2). They were also more likely to report past year depression, GAD, and PTSD. There was a strong association between OME and past year pharmaceutical opioid dependence (120mg vs. 70mg respectively, hedges g 0.54, 95%CI -0.86- -0.22, Table 5- 2). Significant predictors in bivariate analyses were entered into a multivariate logistic regression to examine the factors associated with ICD-10 lifetime pharmaceutical opioid dependence with past year symptoms (Table 5-2). In the multivariate analyses, those who met criteria for past year pharmaceutical opioid dependence were younger; more likely to engage in aberrant behaviours; and were more likely to meet criteria for ICD-10 lifetime benzodiazepine dependence. Note that the multivariate analysis did not include OME due to number that did not return the medication diary, and the complexity of the participant characteristics that did not return the diaries. We ran a post hoc multivariate analysis including OME which resulted in age (RR 0.96, 95% CI 0.93–0.99) and aberrant behaviours (RR 2.49, 95% CI 1.18–5.24) and ICD-10 lifetime benzodiazepine dependence (RR 4.31, 95%CI 1.55–12.63) remaining significant

128

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

5.6 Discussion Research is needed to address concerns about potential adverse events of chronic opioid therapy, including overdose, dependence, aberrance, and diversion (Minozzi, Amato, and Davoli 2013). The POINT study is unique with respect to the detailed data collected on actual consumption of all pharmaceutical opioids, including those obtained without prescription, i.e., over-the-counter opioids sold in pharmacies. This detailed assessment revealed some concerning patterns of opioid consumption, and clear associations between high-level consumption and a range of indicators of poorer functioning. Approximately 15% of the cohort was taking daily doses of more than 200 mg OME and around 40% of the sample was consuming 90 mg OME or more daily. Those taking higher doses had the highest rates of problems with opioid medication, aberrance, and dependence. Concerningly, participants receiving higher OME doses reported less pain relief from their medications than participants on lower consumptions. Not only were there no differences in pain relief, pain severity, time in pain between those consuming 91–199 mg OME per day and those taking more than 200 mg OME, but also there were no differences in rates of aberrance and dependence between these two groups.

Higher current opioid consumption was associated with a range of demographic and substance use characteristics. Correlates of higher opioid consumption were also consistent with factors identified in the literature as being associated with increased overdose mortality risk, including young age, male gender, lower socioeconomic status, and psychiatric comorbidity (Clarke et al. 2014; Cochran et al. 2014; Rogers et al. 2013; Sullivan 2010). They were also consistent with characteristics identified in risk screening tools for opioid prescribing (Webster and Webster 2005). The term “adverse selection” has been coined to describe this apparent contradiction in which the likelihood of a patient receiving opioid therapy increases as the number of risk factors for adverse outcomes increases (Sullivan 2010); this study found strong evidence for this, whereby those consuming higher levels of opioids were clearly those with a more complex picture of physical and mental health problems, as well as social disadvantage. Further, in this sample many of the chronic pain patients prescribed 129

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

higher doses of long-term opioids were concurrently taking other medications (e.g., benzodiazepines) in doses that are considered high-risk for adverse outcomes, and the levels of concomitant medications were higher in those taking higher amounts of opioids.

The association of higher opioid consumption with increasing levels of aberrant behaviours (e.g., tampering and nonmedical use) suggests that monitoring by prescribers is warranted (Passik and Kirsh 2008). Conversely, the finding that some behaviours such as doctor shopping were rarely reported suggest that, at least in this sample, strategies such as prescription drug monitoring would have limited ability to identify patients at risk.

A minority (8.5% of the sample) met criteria for ICD-10 lifetime pharmaceutical opioid dependence, and 4.7% had features of dependence within the past year. This is consistent with research suggesting that pharmaceutical opioid use disorders affect a minority of CNCP patients prescribed opioids (Minozzi, Amato, and Davoli 2013). In this study, past year dependence was associated with many indicators of adverse psychosocial, mental, and physical functioning; it remained independently associated with being younger and having lifetime benzodiazepine dependence, which has important clinical implications for the safety of opioid prescribing.

Universal precautions in opioid prescribing have been widely endorsed internationally and in Australian national guidelines (Gourlay, Heit, and Almahrezi 2005; Royal Australasian College of Physicians 2008), and provide a uniform approach to risk management based on the fact that chronic pain and substance use disorders often co- occur. Guidelines based on universal precautions often suggest consultation with a pain specialist in cases where high doses (usually more than 120 mg OME per day) appear to be required and in whom improvement in pain and function are not seen. Uptake of these guidelines in practice is generally low, possibly due to challenges in prescriber confidence in managing identified risks. However, this study has clearly 130

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

demonstrated that there is a significant proportion of CNCP patients taking very high doses of opioids who have multiple risk factors for potential adverse outcomes. Similarly, those meeting criteria for dependence had higher levels of most indicators of poorer well-being. There is clearly a need for increased vigilance and reassessment of the progress and functioning of CNCP patients in whom opioid consumption is considerable and problems related to opioid consumption prominent.

5.7.1 Limitations

This study has a number of strengths, which include the large national sample, with data detailed collected on actual consumption of opioids and OTC medication, thus addressing the common limitation of relying on administrative data for medication use. In addition, carefully selected and validated measures were selected based on the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (Dworkin et al. 2005; Turk et al. 2003).

Nonetheless, there are limitations. The self-report design raises potential issues of both recall and social desirability bias. However, self-report is reliable in the Pharmaceutical Opioid Use and Dependence within the POINT Cohort and is commonly used in research involving sensitive issues where care is needed to minimise concern over confidentiality and judgment (Darke 1998). All participants were informed that their responses would be de-identified and confidential, which enhances the validity of self-reported substance use.

Further, OME was not entered into the multivariate analysis to consider current consumption as an independent correlate of pharmaceutical opioid dependence. OME data were only available for 29 out of the 69 participants who met criteria for past year opioid dependence. The participants who did not return the medication diary were more likely to be younger, have more pain conditions, more mental health problems, and were more likely to meet criteria for past year pharmaceutical opioid dependence.

131

Chapter five: Pharmaceutical opioid use and dependence in the POINT sample

Therefore, the proportion of participants on higher OME doses may have been underestimated, although there is no reason to expect the associations we observed here would vary. Future POINT research will include linkage to PBS dispensing data and will allow us to obtain medication profiles for those that did not return their medication diaries.

Although the POINT cohort may not be representative of all CNCP patients receiving opioids, as noted earlier, data collected from pharmacists over the recruitment window showed striking similarity in these demographic characteristics of those receiving opioids more generally and the sample recruited here. A final limitation to note is that there are a assumptions made when OME units are used as measure of opioid consumption (Nielsen et al. 2014), although this is an accepted and widely used method to calculate opioid doses (Svendsen et al. 2011).

5.7.2 Conclusion

This study demonstrated that a significant minority of people taking opioids for CNCP were taking considerable amounts of those drugs on a daily basis, and that those consuming more had a higher number of risk factors for less favourable outcomes of opioid use. Higher opioid consumption was also strongly associated with risk for dependence. Given the psychiatric comorbidity and combinations of sedative medications that were documented in this population, the risks of high-dose consumption of pharmaceutical opioids need to be weighed against clinical evidence that patients are deriving net benefit from their use

132

Chapter six: Suicidality in the POINT sample

6 PAPER FIVE: PREVALENCE AND CORRELATES OF SUICIDAL THOUGHTS AND SUICIDE

ATTEMPTS IN PEOPLE PRESCRIBED PHARMACEUTICAL OPIOIDS FOR CHRONIC PAIN.

Gabrielle Campbell1, Raimondo Bruno2, Fiona Shand3, Wayne Hall4,5, Shane Darke1, Michael Farrell1 and Louisa Degenhardt1

1. National Drug and Alcohol Research Centre, UNSW, Australia 2. School of Medicine, University of Tasmania, Australia 3. Black Dog Institute, UNSW Australia 4. Centre for Youth Substance Abuse Research, University of Queensland, AUSTRALIA 5. National Addiction Centre, Kings College, London ENGLAND

Paper five has been accepted in Clinical Journal of Pain (Campbell, Bruno, Darke, Shand, et al. 2015)

133

Chapter six: Suicidality in the POINT sample

6.1 Copyright Statement I certify that this publication was a direct result of my research towards this PhD, and that reproduction in this thesis does not breach copyright regulations.

Campbell, G., Bruno, R., Shand, F., Hall, W., Darke, S., Farrell, M., & Degenhardt, L. (2015). Prevalence and correlates of suicidal thoughts and suicide attempts in people prescribed pharmaceutical opioids for chronic pain. Clinical Journal of Pain. Doi: 10.1097/ajp.0000000000000283

Gabrielle Campbell November, 2015

134

Chapter six: Suicidality in the POINT sample

6.2 Preamble The paper presented in the current chapter examines suicidality amongst the POINT cohort. As seen in Chapter 2, suicidal-related behaviours are high amongst people living with CNCP, and are associated with having CNCP over and above socio- demographic and mental health factors. It has previously been suggested that there are general and pain-specific risk factors in people living with CNCP. Research, to date, however, has been limited by small sample sizes, or a focus on clinical populations. Further, as mentioned previously, the NSMHWB, and other population surveys to date, have not included an extensive list of pain factors. The current paper aimed to examine the influence of general and pain-specific risk factors in people prescribed opioids for CNCP. Further, as we have seen in Chapters 2 and 4, people living with CNCP have high rates of suicide-related behaviours, especially suicidal ideation, however, only 7.4% of people who report suicidal ideation attempt suicide (ten Have et al. 2009). A recent commentary suggests that previous suicidality research has been limited in its focus on the characteristics of people who attempt suicide compared with people who did not attempt, rather than examining factors that increase the likelihood of someone with suicidal ideation making a suicide attempt (Klonsky and May 2014). This framework has been termed “ideation-to-action” (Klonsky and May 2014). The paper presented in this Chapter examines this novel approach by examining what factors are associated with elevating from ideation to attempt. To date, there have been no studies that have examined this ideation-to-action framework in a chronic pain sample.

135

Chapter six: Suicidality in the POINT sample

6.3 Abstract Objectives: The main objectives of the paper were (1) to examine the prevalence of suicidality in a large community-based chronic pain sample taking prescribed opioids for chronic pain; and (2) to examine general and pain-specific factors that predict such ideation, and the transition from ideation to making a suicide attempt (ideation-to- action).

Materials and Methods: Baseline data from the Pain and Opioids IN Treatment (POINT) study with a cohort of 1,514 community based people prescribed opioids for chronic non-cancer pain across Australia.

Results: Past 12-month suicidal ideation was reported by 36.5% of the cohort and 16.4% had made a lifetime suicide attempt (2.5% in the last 12-months), after the onset of their pain condition. Suicidal ideation in the past 12-months was independently associated with a past suicide attempt [adjusted odds ratio (AOR)=4.82; 95% confidence interval, 2.43-9.56] and past 12-month depression (AOR=4.07, 95% confidence interval, 1.88-8.78). Only a lower pain self-efficacy score was independently associated with past 12-month ideation-to-action (AOR=0.98, 95%CI0.88-0.99). Notably, only general-suicide risk factors were associated with 12-month suicidal ideation; but for past year ideation-to-action, pain-specific factors also had independent associations.

Discussion: The study is one of the first to comprehensively examine general and pain- specific risk factors for suicidality in a large chronic pain sample in which suicidal ideation was common. A low pain self-efficacy score was the only factor independently associated past 12-month ideation-to-action.

136

Chapter six: Suicidality in the POINT sample

6.4 Introduction Chronic non-cancer pain (CNCP) has been estimated to affect between 17% (Blyth et al. 2001) and 41% (Tsang et al. 2008) of the general population. is associated with poorer physical health (Eriksen et al. 2006), lower quality of life (Hagg, Fritzell, and Nordwall 2003; Ojala et al. 2014) and a greater risk of developing depression (Currie and Wang 2004) and other mental health problems (Dersh, Polatin, and Gatchel 2002; McWilliams, Cox, and Enns 2003). Although there are interventions for CNCP, it is often the case that many patients have to accept a lifetime treatment plan for the management of their pain, and that their pain is not completely alleviated. In one European study, approximately 40% of people with pain perceived that management of their pain was inadequate (Breivik et al. 2006). It is not surprising then, that rates of suicidal ideation, attempted and completed suicides are high among this group of patients (Meilman 1984). People with CNCP have almost double the risk of death by suicide (Tang and Crane 2006) and are two-to-three times more likely to experience suicidal ideation or make attempts (Tang and Crane 2006). Furthermore, approximately two-thirds of those who had attempted suicide had a history of chronic pain (Campbell, Darke, et al. 2015). As the population in developed countries ages, the incidence and prevalence of in chronic pain will increase, so research on the relationship between chronic pain and suicidality is timely.

Previous research suggests that there are both pain-specific risk factors (e.g., pain severity), and an increased comorbidity of traditional risk factors (e.g., depression) that contribute to higher rates of suicidal behaviour in chronic pain patients (Cheatle et al. 2014; Fishbain 1999). In a review by Tang et al (Tang and Crane 2006), and a more recent one by Hassett (Hassett, Aquino, and Ilgen 2014) traditional risk factors were found to be associated with suicide behaviours in people with chronic pain. These included: being female, a previous suicide attempt, psychiatric comorbidity, substance use, access to weapons, abuse in childhood, and feelings of isolation and hopelessness. Pain-specific factors associated with suicidality included: location and type of pain, severity of pain, length of pain, pain catastrophizing, access to analgesics, and sleep problems. It is unclear if the same risk factors for suicide in the general population are 137

Chapter six: Suicidality in the POINT sample

independently associated with suicide in people with chronic pain (Tang and Crane 2006). Existing suicide research in CNCP populations has involved small samples from clinical settings with a narrow range of patient characteristics.

A recent commentary suggests that previous suicidality research has been limited to comparing the characteristics of people who do and do not attempt suicide and, rather than examining factors that increase the likelihood of someone with suicidal ideation making a suicide attempt. (Klonsky and May 2014). This framework has been termed “ideation-to-action” (Klonsky and May 2014). To the authors’ knowledge, there have been no studies that have used this ideation-to-action framework in a chronic pain sample.

The Pain and Opioids IN Treatment (POINT) study is a cohort of 1,514 people in the community who have been prescribed strong opioids for chronic pain. POINT provides an opportunity to examine suicidality and chronic pain in a large community-based cohort. The aims of the current paper are to:

1. estimate the prevalence of suicidal behaviours (ideation and attempts) in a sample of community based chronic pain patients who are taking opioids; 2. determine the general and pain-specific correlates and predictors of lifetime and suicidal ideation in the past year in people who have experienced suicidal ideation after the onset of their pain; 3. determine general and pain-specific correlates and predictors of moving from suicidal thoughts to a suicide attempt; and 4. determine if there are pain specific factors associated with suicidality, over and above general pain factors.

138

Chapter six: Suicidality in the POINT sample

6.5 Materials and Methods The study was approved by the Human Research Ethics Committee of the University of New South Wales (HREC reference: # HC12149). The study also received A1 Australian National Pharmacy Guild Approval to approach pharmacists to assist with recruitment of participants (Approval n. 815). Full details of the study design have been published elsewhere (Campbell et al. 2014; Campbell, Nielsen, Bruno, et al. 2015).

6.5.1 Recruitment

POINT participants were recruited through community pharmacies across Australia and were: 18 years or older; competent in English; mentally and physically able to partake in telephone and self-complete interviews; without serious cognitive impairments; living with CNCP; and taking prescribed Schedule 8 opioids for CNCP for >6 weeks. In the current study, CNCP was defined as pain that has persisted for >3- months. Schedule 8 is an Australian classification of drugs of dependence that are subject to additional regulatory controls regarding their manufacture, supply, distribution, possession, and use (Therapeutic Goods Administration 2013). Schedule 8 opioids include morphine, oxycodone, buprenorphine, methadone, and hydromorphone. A history of injecting drug use was not an exclusion criterion, but people currently prescribed with pharmaceutical opioids for opioid substitution therapy for heroin dependence or for cancer were not eligible. Of the 2,091 participants assessed for eligibility, 90% (n=1,873) were eligible and 1,514 completed the baseline interview (n=201 refused after being deemed eligible and 100 were unable to be contacted). Data were missing for seven people.

Telephone interviews were conducted by trained interviewers, who had a minimum 3- year health or psychology degree, had undergone suicide response training, and were provided with glossaries of chronic pain medications and conditions. Interviews were conducted over the telephone and took approximately 1 to 1.5 hours to complete. A self-completion survey was also sent to participants who were reimbursed AUD40 for the baseline interview. 139

Chapter six: Suicidality in the POINT sample

6.5.2 Measures

The measures, tools, and domains collected were based on recommendations made under the auspices of the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) (Dworkin et al. 2005; Turk et al. 2003). Full details of the measures used in the study have been reported elsewhere (Campbell et al. 2014).

6.5.2.1 Suicidality

The suicidality questions were based on questions used in the 2007 National Survey of Mental Health and Wellbeing (Slade et al. 2009). These questions were adopted from the World Mental Health (WMH)-Composite International Diagnostic Interview (CIDI) (Haro et al. 2006; Kessler and Üstün 2004) suicidality module that has been used extensively in previous international studies (Nock et al. 2008; Nock and Kessler 2006).

Participants were asked about lifetime and past 12-month presence of suicidal thoughts, plans, and attempts. For each of the behaviours, participants were asked for age of onset and age of most recent behaviour. Participants who had made a suicide plan in the past 12-months, or had made a lifetime suicide attempt, were asked about the method used. These were grouped into non-poisoning (fire a gun, carbon monoxide, cut wrists/stab, jump from height, jump in front of train/bus, strangulation/hanging, suffocation, drowning), or poisoning (opioid and non-opioid) methods. People who had made a suicide attempt were asked to select one of three statements. These were: “I made a serious attempt to kill myself and it was only luck that I did not succeed”; “I tried to kill myself, but knew that the method was not foolproof”; and “my attempt was a cry for help. I did not intend to die”. Suicidality after the onset of pain was determined by using the youngest age of onset for any pain condition and the first age reported for any suicidal behaviour.

140

Chapter six: Suicidality in the POINT sample

6.5.2.2 Demographic characteristics

Data on age, sex, marital status, and current work status (dichotomised to not unemployed or unemployed) were collected. Net weekly income was binary coded to AUD400, as an income below this cut-is comparable to that from unemployment or disability benefits in Australia.

6.5.2.3 Pain and pain-related measures

Participants were asked about chronic pain conditions they had been diagnosed with or experienced in their lifetime and in the past 12-months. They were also asked the age of onset of the pain condition and how long they had been in pain. Current pain severity and pain interference were measured by the Brief Pain Inventory (Cleeland 1991), as a continuous score out of 10. A higher score indicates more pain severity or interference. The Pain Self-Efficacy Questionnaire (Nicholas 2007; Nicholas, Asghari, and Blyth 2008) produced a continuous score out of 60. A higher score indicates greater self-efficacy.

6.5.2.3 Physical health

The Short Form 12 (SF12) is based on population norms, with a mean score of 50 and a SD of 10 (Sanderson K and G. 2002). Sleep patterns were measured using the Medical Outcomes Study (MOS) Sleep scale (Spritzer and Hays 2003), providing a total score out of 100 on sleep problems, with higher scores indicating greater sleep problems.

6.5.2.4 Medications

Self-reported opioid and other medication use were collected in the interview and oral morphine equivalents (OME) were obtained from a medication diary completed over a one week period as part of the self-complete questionnaire mailed to participants. Oral morphine equivalent daily doses were estimated (Nielsen et al. 2014) using

141

Chapter six: Suicidality in the POINT sample

available references (Therapeutic Guidelines Limited 2013; Australian Medicines Handbook 2013; Faculaty of Pain Medicine ANZCA 2014).

6.5.2.5 Mental health and substance use

Participants were asked about mental health disorders they have been diagnosed with or that had been problematic in the past 12-months. These included depression, post- traumatic stress disorder (PTSD) and anxiety/panic attacks and borderline personality disorders. Age of onset was also collected. Participants were screened for potential ICD-10 diagnosis of Borderline Personality Disorder (BPD) using the National Survey of Mental Health and Wellbeing version of the CIDI (Andrews et al. 1999).

Participants were asked about lifetime and past 12-month alcohol and illicit drug use. Lifetime drug and alcohol use disorders (using ICD-10 disorders criteria) were assessed via the Composite International Diagnostic Interview (CIDI) (World Health Organization 2001) alcohol and illicit drug use module.

Social support was measured by the Medical Outcomes Study Social Support Scale (MOS-SS) (Sherbourne and Stewart 1993) which provides a score based on a converted average score out of 12 questions. The higher the average score, the higher degree of social support.

Five childhood maltreatment questions were asked, “Before the age of 16, did you experience….” sexual abuse, physical abuse, emotional abuse, physical neglect, emotional and/or the witnessing of any violence. These were based on Sansone (2012 (Sansone, Whitecar, and Wiederman 2009)). In the current study we only used data from the sexual, physical and emotional abuse questions. We combined answers to these questions to one variable of ‘any childhood abuse’.

142

Chapter six: Suicidality in the POINT sample

6.5.3 General and pain-specific suicide risk factors

In the current study, general and pain-related suicide risk factors were based on previous literature (Hassett, Aquino, and Ilgen 2014; Tang and Crane 2006). General suicide risk factors included, age, sex, previous suicide attempt, lifetime and past 12- month depression, anxiety, PTSD, BPD, ICD-10 alcohol use disorder and substance use disorder, history of childhood abuse, and lack of social support. Pain-related suicide risk factors included type and location of pain, pain severity and interference, length of time in pain, number of pain conditions (lifetime and past 12-months), sleep, OME and current prescriptions of benzodiazepines, pain and self-efficacy score, and Short Form- 12 physical health score.

6.5.4 Statistical analysis

All analyses were conducted using STATA, version 12.0. Proportions, means, and SDs were reported for normally distributed data and odds ratios (OR) with 95% confidence intervals (CI) were used to compare groups. Where data were non-normally distributed, medians, interquartile ranges (IQRs) were reported and Hedges G effect sizes with 95% CI were used to compare groups. For sex comparisons logistic regression (95% CI) and Hedges G effect sizes were used at <0.05 level of significance. Only participants who had experienced suicidality after their onset of chronic pain were included in regression analyses. Correlates and predictors of lifetime ideation and attempt were examined only for those who had experienced suicidality after the onset of their pain, so that associations between pain and suicidality would be more accurate. To examine the ideation-to-action framework, comparisons were made between those who had ideation but no attempt, and those who had made an attempt.

Four multivariate regressions were planned to examine lifetime and in the past 12- months, suicidal thoughts and ideation-to-action. Univariate analyses were conducted to determine which factors were to be entered in the multivariate. All variables significant at the univariate level (P<0.05) were entered into a multivariate model. Age, 143

Chapter six: Suicidality in the POINT sample

sex, and depression were included in the models even if they were not significant at univariate level. As four multivariate regressions were planned, a conservative approach to significance (P<0.01) was used. Secondary analyses were conducted on multivariate regressions for model comparisons to examine the contribution of pain- specific factors over and above general-suicide risk factors.

Bayesian information criterion (BIC), which indicates the best fit of separate models to the overall fit of a regression (a lower BIC indicates a better fit, without overfitting) and pseudo R2 were used to examine the contribution of pain-specific factors to suicide ideation and suicide attempts in nested regressions. Significant variables at the univariate levels (0.05) were split into general and pain-specific risk factors and the contribution of pain-specific factors (model 2), over and above general-suicide risk factors (model 1) was compared.

6.6 Results 6.6.1 Overall characteristics of the cohort

The POINT study included 1,514 people, 44% of whom were male with a mean age of 58 years (SD=13.7). Forty-eight percent of the sample were unemployed and 31% were retired. Sixty-four percent of the sample reported that their employment status had changed because of their pain condition. Fifty-four percent reported a weekly income of AUD399, (equivalent to the single-aged or disability pension).

The average pain duration was 10 years (IQR=4.5 to 20) and participants reported taking pharmaceutical opioids continuously for a median of 4 years (IQR=1.5 to 10 ). The average daily OME dose was 75 mg (IQR=36 to 150 mg). Fifty-two percent of the sample concurrently took antidepressants, 34% benzodiazepines, and 6.7% antipsychotics.

144

Chapter six: Suicidality in the POINT sample

6.6.2 Prevalence and characteristics of suicidality in the cohort

Lifetime suicidal ideation was reported by 45.9% of the cohort and 20.0% in the past- 12-months (Table 6-1). Suicidal ideation occurred after the onset of pain in 36.5% of the sample. Only 10.0% the sample experienced ideation before their pain condition, which is similar to the lifetime prevalence of suicidal thoughts in the general population (Nock et al. 2008). There were no sex differences in prevalence of suicidal ideation

A lifetime suicide attempt was reported by 20.3% of the cohort (an average of two attempts) and 2.5% had attempted suicide in the past 12-months. Females were more likely than males to report a lifetime attempt (OR=1.40; 95% CI, 1.08-1.89) and to have made an attempt after the onset of their pain (OR=1.43; 95% CI, 1.08-1.90). There were no sex differences in the prevalence of attempts in the past 12-months. There were also no sex differences in non-poisoning attempts, although females were more likely than males to plan an attempt by poisoning (OR=1.77; 95% CI, 1.20-2.62).

145

Chapter six: Suicidality in the POINT sample

Table 6-1: Lifetime and Past 12-month suicidal behaviours in the POINT cohort

Males Females Total % (N=672) %(N=842) %(N=1,514) Suicidal thoughts % (n) % (n) % (n) Ever seriously thought suicide 45.1 (300) 46.5 (386) 45.9 (686) Median age first thought (Iqr) 36 (24-49) 30 (19-42) 33 (20-45) Thought suicide past 12-months 20.0 (134) 20.0 (168) 20.0 (302) Suicidal thoughts after onset pain 37.2 (250) 35.9 (302) 36.5 (552) Planned suicide Ever made a plan for suicide 19.2 (129) 22.8 (192) 21.2 (321) Age first plan 38 (25-48) 30 (19-41) 33 (21-45) Age last plan 45 (35-55) 39 (30-50) 42.5 (30-54) Made a plan in past 12-months 6.1 (41) 6.8 (57) 6.5 (98) Planned method** Non-poisoning means 65.8 (25) 37.1 (13) 50.6 (37) Poisoning Any 21.0 (8) 60.0 (21) 39.7 (29) Take non-opioid pills 10.5 (4) 37.1 (13) 23.3 (17) Overdose on opioids 10.5 (4) 22.9 (8) 16.4 (12) Overdose on heroin 5.2 (2) 0 2.7 (2) Other 7.9 (3) 2.8 (1) 5.4 (4) Attempt suicide Ever attempt suicide 17.3 (116) 22.7 (191) 20.3 (307) Number of lifetime attempts 2 (1-3) 2 (1-3) 2 (1-3) Age first attempt 30 (20-42) 28 (18-38) 29 (20-40) Age last attempt 37 (29-51) 30.5 (25-45) 35 (27-48) Suicide attempt after onset of pain 13.7 (92) 18.5 (156) 16.4 (248) Attempted suicide past 12-months 2.2 (15) 2.7 (23) 2.5 (38) Method of suicide attempt Non-poisoning means 44.6 (33) 26.3 (31) 27.6 (20) Poisoning Any 52.7 (39) 72.0 (85) 65.1 (125) Take non-opioid pills 32.4 (24) 60.2 (71) 50.0 (96) Overdose on opioids 20.3 (15) 11.9 (14) 15.1 (29) Overdose on heroin 2.7 (2) 0 1.0 (2) Other 5.4 (4) 1.7 (2) 3.1 (6) Suicide statements Serious attempt 55.0 (61) 47.4 (91) 50.2 (152) Tried to kill myself, knew it wasn’t 24.3 (27) 17.2 (33) 19.8 (60) fool-proof Cry for help 20.7 (23) 35.4 (68) 30.0 (91) ** of those that planned/attempted Non-poisoning includes: Fire a gun, Carbon monoxide, Cut wrists/stab, Jump from height, Jump in front of train/bus, Strangulation/hanging, Suffocation, Drowning

146

Chapter six: Suicidality in the POINT sample

6.6.3 Correlates and predictors of lifetime and past 12-month suicidal ideation

Participants, who reported suicidal ideation (Table 6-2) after the onset of their pain, had significantly greater odds of a range of mental health and substance use disorders. Participants who experienced suicidal ideation in the past 12-months also reported a higher pain score, more pain interference, and a lower pain self-efficacy score, than patients who did not report ideation in the past 12-months (Table 6-3).

Variables significant at 0.05 in univariate analyses were entered into a multivariate regression for both lifetime and past 12-month suicidal ideation. In multivariate analyses (Table 6-2), participants who experienced lifetime suicidal ideation, after the onset of their chronic pain, were more likely to have been diagnosed with depression (adjusted odds ratio (AOR) =3.87; 95% CI, 2.26-6.63], to have screened positive for BPD (AOR=3.05; 95% CI, 1.80-5.15), and to report childhood abuse (AOR=1.76; 95% CI, 1.15- 2.70). None of the pain-related factors remained significantly associated with lifetime ideation in the multivariate model.

In a multivariate analysis for past 12-month suicidal ideation (Table 6-3) among people who experienced suicidality after the onset of their pain, participants who experienced past 12-month suicidal ideation were more likely to report past 12-month depression (AOR=4.07; 95% CI, 1.88-8.78). They were also more likely to have made a lifetime suicide attempt (AOR=4.82; 95% CI, 2.43-9.56). No pain-related factors remained significant in the multivariate regression.

147

Chapter six: Suicidality in the POINT sample

Table 6-2: Correlates and predictors of lifetime suicidal ideation in people that experienced ideation after the onset of their pain

Lifetime ideation

No yes OR (95%CI) Adj OR (99%CI) Adjusted N=962 N=552 Hedges G (95%CI) N=978^ P value General suicide risk factors Demographics Median age (IQR)+ 61 (51-70) 53 (45-61) 0.57 (0.46-0.67)** 0.99 (0.97-1.01) 0.075 Male (%) 43.9 45.3 0.94 (0.76-1.17) 0.84 (0.53-1.33) 0.325 % married or de facto 56.2 49.1 0.75 (0.61-0.93)** 0.88 (0.58-1.33) 0.429 % Unemployed 39.8 64.1 2.70 (2.18-3.35)*** 1.47 (0.94-2.29) 0.026 Mental health % Lifetime depression 52.5 87.5 6.33 (4.78-8.40)*** 3.86 (2.26-6.16) <0.000 % Lifetime PTSD 10.4 28.3 3.40 (2.57-4.48)*** 1.56 (0.93-2.62) 0.027 % Lifetime anxiety 26.1 56.7 3.71 (2.97-4.63)*** 1.49 (0.96-2.31) 0.019 %Borderline personality 8.04 36.8 6.66 (4.64-9.55)*** 3.03 (1.79-5.13) <0.000 disorder %experiencing 36.9 61.4 2.72 (2.19-3.38)*** 1.74 (1.14-2.67) 0.001 childhood abuse Lifetime substance use AUD use disorder 26.2 38.6 1.77 (1.41-2.21)*** 0.92 (0.57-1.47) 0.639 Substance use 27.8 37.7 1.57 (1.25-1.97)*** 1.41 (0.87-2.29) 0.065 disorder Pain-specific factors Median months living 120 (48-240) 144 (60-276) -0.07 (-0.18-0.03) with pain# (IQR) % Lifetime Back or neck 75.1 79.2 1.26 (0.99-1.63) problems % lifetime 23.6 39.3 2.10 (1.67-2.63)*** 1.25 (0.77-2.04) 0.228 Frequent/severe headaches % Lifetime Fibromyalgia 4.5 8.3 1.94 (1.26-2.98)** 0.96 (0.46-2.00) 0.887 Median no. pain 3 (2-3) 3 (2-4) -0.33 (-0.43—0.22)* 1.15 (0.95-1.40) 0.053 conditions lifetime+ (range) +Hedges G effect size (95%CI) ^Out of 1,514, 978 had completed the Borderline Screener, so multivariate regression was reduced to n=978 *<0.05 **<0.01 ***<0.001

148

Chapter six: Suicidality in the POINT sample

Table 6-3: Correlates and predictors of past 12-month suicidal ideation in people with chronic pain

Past 12-month ideation

No Yes OR (95%CI) Adj OR (99%CI) Adj p N=1211 N=302 Hedges G (95%CI)+ value General suicide risk factors Demographics Median age (range)+ 60 (49-69) 52 (43-59) 0.60 (0.47-0.73)** 0.99 (0.96-1.01) 0.172 Male (%) 55.6 55.6 1.01 (0.79-1.30) 0.95 (0.49-1.87) 0.854 % married or de facto 55.2 47.4 0.72 (0.57-0.94)* 0.85 (0.44-1.66) 0.536 % unemployed 44.1 66.9 2.55 (1.96-3.33)*** 1.34 (0.64-2.78) 0.304 Mental health % 12m depression 30.9 79.1 8.47 (6.25-11.46)*** 4.07 (1.88-8.78) <0.000 % 12m anxiety 19.4 50.0 4.16 (3.18-5.43)*** 1.17 (0.56-2.46) 0.578 % 12m PTSD 7.3 23.5 3.87 (2.75-5.46)*** 1.35 (0.57-3.17) 0.371 % Borderline 11.5 46.2 6.63 (4.56-9.62)*** 1.72 (0.83-3.55) 0.054 % reporting a history of 41.3 64.2 2.57 (1.96-3.32)*** 1.16 (0.59-2.78) 0.570 childhood abuse Lifetime suicide attempt 12.5 51.3 7.35 (5.54-9.75)*** 4.82 (2.43-9.56) <0.000 MOS social support+ 3.4 (2.5-4.1) 2.8 (1.8-3.6) 0.53 (0.39-0.67)** 1.01 (0.75-1.37) 0.922 Past 12m substance use % Alcohol… 60.5 60.6 1.00 (0.78-1.30) % substance use… 2.6 4.1 1.5 (0.76-2.95) Pain-related factors Median months living with 120 (54- 126 (60-264) 0.04 (-0.08-0.17) pain+ (range) 240) Back or neck problems 75.8 79.5 1.23 (0.91-1.68) Frequent/severe 63.0 56.0 0.75 (0.58-0.96)* 1.22 (0.55-2.73) 0.525 headaches Fibromyalgia 5.3 8.3 1.61 (1.00-2.17)* 0.43 (0.13-1.38) 0.063 Median no. pain conditions 2 (1-3) 3 (2-3) -0.34 (-0.47—0.22)* 1.12 (0.78-1.60) 0.434 12m+ (range) Current pain BPI pain score (BPI)## (SD) 5.0 (1.8) 5.5 (1.6) 1.20 (1.11-1.30)*** 0.87 (0.68-1.11) 0.133 Pain Interference (BPI)## 5.4 (2.2) 7.0 (2.0) 1.46 (1.35-1.56)*** 1.20 (0.96-1.50) 0.033 (SD) Pain Coping and Self 30.8 (13.5) 23.2 (12.0) 0.96 (0.95-0.97)*** 1.00 (0.96-1.03) 0.823 Efficacy## (SD) Physical health Sleep problems (mean) 44.4 (20.1) 59.0 (19.7) 1.03 (0.03-1.04)*** 1.01 (1.00-1.04) 0.027 SF12 physical health+ 26 (22-31) 27 (24-32) -0.14 (-0.28-0.00) Medications 67.5 (33.2- 90 (45- -0.18 (-0.33 - - 0.04)* 1.00 (1.00-1.00) 0.776 OME+ 136.5) 171.43) Current benzodiazepines 30.2 49.3 2.25 (1.74-2.91)*** 0.91 (0.46-1.80) 0.717 *<0.05 **<0.01 ***<0.001 + Oral morphine equivalents

149

Chapter six: Suicidality in the POINT sample

6.6.4 Correlates and predictors of lifetime and past 12-month ideation-to-action

Participants who experienced a lifetime or past 12-month suicide attempt were compared with those that had experienced suicidal ideation but not made an attempt to examine factors that increased the chance of an attempt. In the univariate analyses, lifetime ideation-to-action was significantly (P<0.05) associated with being male, not married or in a de facto relationship, income

In univariate analyses (Table 6-5), past year ideation-to-action was significantly (P<0.05) associated with a lower pain and self-efficacy score, poorer physical health, 12-month anxiety, and PTSD compared with those who had experienced past 12- month ideation only

Again, variables significant at the univariate level (0.05) were entered into a multivariate logistic regression. Lifetime ideation-to-action was more likely in participants with incomes

In multivariate analyses a past year ideation-to-action was significantly (P<0.01) more likely in those with a lower pain self-efficacy score compared with those with ideation alone (AOR=0.94; 95% CI, 0.88-0.98) (Table 6-5). When past 12-month history of depression was added to the multivariate analyses to determine whether it had any association it remained nonsignificant.

150

Chapter six: Suicidality in the POINT sample

Table 6-4: Ideation-to-action in people that experienced a lifetime attempt after the onset of their pain

Lifetime ideation-to-action

No Yes OR (95%CI) Adj OR (99%CI) Adj p N=304 N=248 Hedges G (95%CI) n=356 value General suicide risk factors Demographics Median age (IQR)+ 53.5 (46- 53 (43- 0.18 (0.010-0.24)* 1.00 (0.98-1.03) 0.566 61) 59.5) Male (%) 52.0 37.1 1.84 (1.30-2.58)** 1.46 (0.78-2.74) 0.122 % married or de facto 54.9 41.9 0.59 (0.42-0.83)** 0.70 (0.38-1.28) 0.064 % Unemployed 60.5 68.6 1.42 (1.00-2.02) Mental health Ever diagnosed with… Depression 82.8 94.3 3.46 (.87-6.41)*** 1.60 (0.53-4.84) 0.270 PTSD 25.4 34.5 1.54 (1.06-2.24)* 1.03 (0.53-2.01) 0.895 Anxiety 46.0 70.5 2.79 (1.96-3.99)*** 2.00 (1.05-3.83) 0.006 Borderline Personality 26.1 50.3 2.86 (1.84-4.47)*** 2.49 (1.33-4.64) <0.000 disorders % experiencing childhood 55.6 68.6 1.74 (1.22-2.47)** 1.14 (0.59-2.17) 0.614 abuse Lifetime substance use AUD use disorder 35.5 42.3 1.33 (0.94-1.88) Substance use disorder 36.5 39.1 1.11 (0.79-1.58) Pain-specific factors Median months living 132 (60- 144 -0.11 (-0.27-0.06) with pain+ (IQR) 264) (66- 276) % Lifetime back or neck 74.7 84.7 1.87 (1.23-2.89)** 1.47 (0.63-3.38) 0.239 problems %lifetime 34.9 44.8 1.51 (1.07-2.13)* 1.25 (0.62-2.51) 0.409 frequent/severe headaches % lifetime Fibromyalgia 6.9 10.1 1.51 (0.82-2.77) Median no. pain 3 (2-4) 3 (2-4) -0.26 (-0.43—0.09)* 1.03 (0.79-1.35) 0.746 conditions lifetime+ (range) *Hedges G effect size (95%CI) **Out of 552, 356 had completed the Borderline Screener, so multivariate regression was reduced to n=356 *<0.05 **<0.01 ***<0.001

6.6.5 Contribution of pain-specific factors over and above general suicide risk factors.

A nested regression was conducted to determine whether pain-related risk factors were independently associated with suicidality, over and above general-suicide risk factors (not shown). Significant variables at the univariate level were entered into the 151

Chapter six: Suicidality in the POINT sample

nested regression for each of the four multivariate regressions on suicidality. General suicide factors were entered in model 1 and pain-specific factors in model 2 to determine the contribution of pain specific factors. In the past 12-month suicidal ideation, model 1, general-suicide risk factors (BIC 546.51, pseudo R2=0.31) was the better fit, compared with the addition of model 2, pain-specific suicide risk factors (BIC 585.06, pseudo R2=0.34). In past 12-month ideation-to-action model 2, with the addition of pain-specific risk factors was the better fit (BIC 181.37, pseudo R2=0.15) over and above general-suicide risk factors (BIC 182.15, pseudo R2=0.09).

152

Chapter six: Suicidality in the POINT sample

Table 6-5: Past 12-month ideation-to-action in people with chronic pain

Past 12m ideation-to-action

No Yes OR (95%CI) Adj OR (99%CI) Adj P N=268 N=38 Hedges G (95%CI)* value General suicide risk factors Demographics Median age (range)* 52 (43-59.5) 50.5 (42-57) 0.20 (-0.14-0.54) 0.99 (0.94-1.04) 0.615 Male (%) 44.8 39.5 1.24 (0.62-2.49) 1.10 (0.32-3.74) 0.838 % married or de facto 48.9 39.5 0.68 (0.34-1.36) % unemployed 64.9 79.0 2.03 (0.89-4.60) Mental health % 12m depression 77.6 84.2 1.53 (0.61-3.85) % 12m anxiety 16.3 71.1 2.85 (1.35-5.98)** 2.85 (0.72-11.20) 0.049 % 12m PTSD 21.3 36.8 2.16 (1.05-4.44)* 2.27 (0.70-7.41) 0.074 % Borderline Personality 46.2 54.6 1.39 (0.57-3.40) Disorder % reporting a history of 63.8 63.2 0.97 (0.48-1.97) childhood abuse Lifetime suicide attempt 45.2 100 MOS social support Mean (SD) 2.91 (1.8-3.6) 2.6 (1.6-3.4) 0.25 (-0.12-0.63) Past 12m substance use % Alcohol… 61.9 52.6 0.68 (0.34-1.35) % substance use… 4.1 2.6 0.63 (0.08-5.03) Pain-specific factors Pain factors Median months living with 138 (60-264) 96 (36-234) 0.17 (-0.17-0.51) pain (range)* 12m Back or neck problems 78.4 86.8 1.82 (0.68-4.88) (%) 12m Frequent/severe 57.1 47.4 0.67 (0.34-1.33) headaches (%) 12m Fibromyalgia (%) 9.0 2.6 0.27 (0.04-2.09) Median no. pain conditions 3 (2-3) 3 (1-3) 0.11 (-0.23-0.44) 12m* (range) BPI pain score (BPI)## (SD) 5.5 (1.7) 5.8 (1.5) 1.11 (0.90-1.38) Pain Interference (BPI)## (SD) 2.9 (2.0) 7.3 (1.9) 1.16 (0.94-1.42) Pain Coping and Self Efficacy## 23.9 (11.8) 18.0 (12.1) 0.96 (0.93-0.99)* 0.94 (0.88-0.99) 0.006* (SD) Physical health Sleep problems Mean (SD) 59.3 (19.3) 57.3 (22.9) 0.00 (0.02-0.01) SF12 physical health*Median 26.9 (24.2- 24.7 (22.2- 0.44 (0.06-0.82)** 0.98 (0.88-1.10) 0.633 (IQR)+ 32.4) 29.0) Medications 90 (45-163.9) 90.9 (30- 0.05 (-0.34-0.44) OME+* 217.5) Current benzodiazepines 47.8 63.2 1.88 (0.93-3.78) + Oral morphine equivalents *Hedges G effect size (95%CI) *<0.05 **<0.01 ***<0.001

153

Chapter six: Suicidality in the POINT sample

6.7 Discussion There were a number of important findings from the current study. Firstly, suicidality was common among patients prescribed strong opioid medications for CNCP. Secondly, the adoption of the ideation-to-action framework in a chronic pain population found notable differences between those who experienced only ideation and those in the ideation-to-action group. Thirdly, pain-specific factors were associated with past 12-month attempts, over and above general-suicide risk factors. These findings contribute significantly to our growing understanding of chronic pain and suicidality.

6.7.1 Prevalence of suicidal behaviours in a sample of community based chronic pain patients who are taking opioids

Consistent with previous research (Braden and Sullivan 2008; Hassett, Aquino, and Ilgen 2014; Smith et al. 2004), there were much higher rates of all forms of suicidality in the chronic pain cohort than in the general population : lifetime suicidal ideation (45.9% vs 13.3%, respectively), lifetime suicide attempts (20.3% vs. 3.2%), past 12- month ideation (20.0% vs 2.3%) and past 12-month suicide attempt (2.5% vs. 0.4%).

Joiner’s Interpersonal Theory of suicide (IPT) (Van Orden et al. 2010) provides some answers to the high rates of suicidal behaviours in people with CNCP. Joiner’s IPT proposes that perceived burdensomeness and low belongingness are associated with suicidal desire, whereas the capability to act on the desire is acquired over time. The impact of chronic pain on an individual’s employment, physical functioning, and quality of life (Campbell, Nielsen, Bruno, et al. 2015), and the reliance on others that were prevalent in this population provides a plausible explanation of the high levels of suicidal desire in this group (Cheatle 2014a). In addition, many people live with CNCP for a number of years and often fail to receive complete relief for their pain through medications or other treatments. This may explain why some CNCP develop the capability to act on this suicidal desire over time.

154

Chapter six: Suicidality in the POINT sample

Interestingly, there were a number of gender findings in the current study that were inconsistent with the general suicide risk literature. Although females were more likely to report suicide by poisoning, there were no sex differences in the use of non- poisoning means (strangulation, shooting, drowning, jumping from a height, etc.). While female preference for poisoning is supported in the literature, generally males have been found to prefer violent methods. It is interesting to note that, although women preferred poisoning, the use of non-opioid pills (37%) was preferred over opioid medication (23%), despite all participants having ready access to opioids. Further, the current study found no sex differences in lifetime and past 12-month ideation or attempts. This is inconsistent with general-suicide research where females are more likely to report suicidal ideation and attempts compared with males (Tsirigotis, Gruszczynski, and Tsirigotis 2011). These findings appear to suggest that the addition of pain interacts with suicidality and sex; however, we are unable to test this possibility directly given that all of our participants were living with CNCP.

6.7.2 Risk factors associated with suicidal ideation

Although a minority of people with suicidal ideation make an attempt (ten Have et al. 2009) approximately 60% of people who attempt suicide report ideation in the past 12-month (Nock et al. 2009). Identification and awareness of who might be at risk is important for clinicians. Consistent with previous research, a past suicide attempt and a diagnosis of depression were associated with suicidal ideation (Nock et al. 2009; Nock et al. 2010).

Pain-specific factors, such as type, duration, and severity of pain, that have been associated with suicidality in previous research, were not supported in the current research (Braden and Sullivan 2008; Hassett, Aquino, and Ilgen 2014). In the current cohort, however, 85% of the sample reported greater than one lifetime pain condition, and had experienced pain for a mean of 10 years. The prescription of opioids may also be a proxy for chronic pain that is more detrimental and severe than pain in chronic

155

Chapter six: Suicidality in the POINT sample

pain patients who have not been prescribed opioids. In a cohort where chronic pain is more entrenched, these pain-specific risk factors appear to be less important.

While a higher OME, current prescription of benzodiazepines, or both were not independently associated with suicidal ideation or ideation-to-action, there were, however, a high proportion of participants prescribed benzodiazepines, many of whom were taking them daily (Nielsen et al. 2015). These drugs can be potentially lethal when used in combination with pharmaceutical opioids. Non-opioid poisoning was the most commonly reported method used in a previous suicide attempt. Although restriction of access to potentially lethal means has been found to be an effective measure in reducing suicide attempts (Hawton et al. 2013; Mann et al. 2005), among a sample of CNCP patients, there is a need to balance effective pain management and the ease of access to other available effective treatments. Because of the high rate of health service utilization among this group, health professionals are in a good position to monitor patient’s pain and suicidal behaviours and make informed decisions about when to prescribe medications (opioids and non-opioids), taking account of possible interactions and dosages, and the potential for overdose..

6.7.3 Factors associated with moving from ideation-to-action in a chronic pain sample

In a population where suicidality is high, it is important to be able to identify those who most likely would move from ideation-to-action. The current study provided a novel approach to determine the risk factors associated with ideation-to-action in a cohort of chronic pain patients.

The current findings support the general ideation-to-action findings by Klonsky (Klonsky and May 2014). Specifically, depression was not a predictor when examining factors that elevate the risk of moving from suicidal ideation to an attempt. Critically, the IPT posits that one of the mechanisms for acquiring the capability for suicide is

156

Chapter six: Suicidality in the POINT sample

exposure to trauma (direct or vicarious) and physical pain. There were high rates of childhood abuse, in the current cohort, and these were more elevated in those who engaged in suicidal behaviours. Previous literature has found childhood abuse and neglect to be associated with chronic pain (Davis, Luecken, and Zautra 2005) and later engagement in suicidal behaviours (Dube et al. 2001). It was an independent risk factor only for lifetime suicidal ideation.

Importantly, the current study found that poorer pain coping skills were associated with a past 12-month suicide attempt. Only in past 12-month attempts were pain- related factors independently associated after controlling for general suicide risk factors. Active, problem-focused coping styles have been found to be associated with a better adjustment to chronic pain (Jensen et al. 1991). It is promising that coping skills can be effectively modified by psychosocial therapies including, cognitive-behavioural therapy, problem solving therapy, and dialectical behaviour therapy (Mann et al. 2005). Again, good access to these treatment options is essential for this population.

6.7.4 Contribution of pain-specific factors to suicidality

Chronic pain has been associated with suicidality over and above mental health factors (Braden and Sullivan 2008; Ratcliffe et al. 2008). In the current study, pain-specific factors were not independently associated with lifetime or past 12-month ideation or lifetime attempt, as mentioned above; in only past 12-month ideation-to-action were pain-specific suicide risk more important, over and above general-suicide risk factors. It may be that in a sample of patients where chronic pain is entrenched and the impact of quality of life has been considerable, there is less variability in these pain-specific factors. Overall, it is the presence of chronic pain and the associated factors that places the person at high risk for suicidality

157

Chapter six: Suicidality in the POINT sample

6.7.5 Limitations

There were limitations of the current study that need to be considered. Firstly, the cross-sectional nature of this study means that causality cannot be examined. We examined only those that had experienced suicidality after the onset of their pain partially to address this issue. Further, the current paper presents data on suicidal ideation, plans and attempts, not completed suicides. As shown, these factors can differ between each of the behaviours, i.e. different risk factors for suicide ideation and suicide attempt. The current study still provides significant insight into risk factors associated with suicidal thoughts and attempts that can aid our identification and prevention of suicidal behaviours. Secondly, the potential biases that may be introduced by the reliance on self-report data need to be considered. Information on chronic medical illnesses and mental health problems were not verified through patient records. However, the rates of pain condition and finding were similar to those in previous research (Braden and Sullivan 2008; Smith et al. 2004) and all participants were informed that their responses would be de-identified and confidential. This has been found to enhance the validity of self-reported substance use (Darke 1998). It must be noted that there was no objective measure of serious cognitive impairment at screening, and although interviewers were trained in interview administration, no reliability checks were carried out.

Finally, there is the potential that we did not recruit a representative sample of people prescribed opioids for their chronic pain. To investigate this possibility, during recruitment we gathered additional data from a random sample of recruiting pharmacies (n=71) on the characteristics of all their opioid customers during the 6- week recruitment window of their involvement. We found that of the total number of customers recorded as purchasing opioids in these pharmacies were 52% female (the POINT cohort was 55% female); and 7% were 18 to 34 years, 55% 35 to 64 years, and 38% 65+ years (vs. 5%, 62%, and 33%, respectively, in the POINT cohort). Of these customers, 63% were prescribed oxycodone (vs. 62% in the POINT cohort), 16.5% prescribed morphine (vs. 15% in the POINT cohort), 21% prescribed fentanyl patches

158

Chapter six: Suicidality in the POINT sample

(vs. 15% in the POINT cohort), and 24% prescribed buprenorphine patches (vs. 21% in the POINT cohort). Although we cannot be sure that all the opioid customers recorded by these pharmacists had been taking these opioids for chronic pain, and for Z6 weeks, the striking similarity in these demographic and opioid prescription characteristics is reassuring. Further, there were 308 people who were eligible but did not complete the baseline interview, 201 refused after eligibility was determined and 100 were unable to be contacted after eligibility was determined. There was a median age of 58 (IQR=46 to 71) for those who refused after eligibility, similar to the cohort, and they been prescribed opioids for a median of two years (IQR=1 to 5) compared with the cohort median of 4 years. The “could not contact” group had a median age of 47.5 years (IQR=40 to 57) and those who had been prescribed opioids had a median of two years (IQR=1 to 5.8). These differences suggest that the “could not contact” was different to the cohort and the people who did not participate overall had been prescribed opioids over a shorter timeframe compared with the cohort. Without further information it is difficult to determine what these differences might mean. From our previous paper on the characteristics of the cohort (Campbell, Nielsen, Bruno, et al. 2015), it appears as though it is the younger people who have the most complex presentations in terms of mental and physical health comorbidities, so if anything; our cohort representation of the complexity of people with CNCP is an underestimation, although we do not expect any of the associations to be different.

6.7.6 Conclusion

Suicidality among people with CNCP is high. There are different risk factors associated with ideation and ideation-to-action among this sample. Understanding these factors can assist clinicians to determine treatment priorities. People with chronic pain have complex clinical profiles, having not only pain, but also mental health problems, sleep problems, low incomes, and low physical functioning. Access and affordability of multidisciplinary treatment options are essential in addressing the multitude of problems experienced by this population.

159

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7 PAPER SIX: ASSOCIATIONS OF BORDERLINE PERSONALITY WITH PAIN, PROBLEMS WITH

MEDICATIONS AND SUICIDALITY IN A COMMUNITY SAMPLE OF CHRONIC NON-CANCER

PAIN PATIENTS PRESCRIBED OPIOIDS FOR PAIN

Gabrielle Campbell1, Raimondo Bruno1-2, Shane Darke1 and Louisa Degenhardt1, 3-5

1. National Drug and Alcohol Research Centre, UNSW, Australia 2. School of Medicine, University of Tasmania, Australia 3. School of Population and Global Health, University of Melbourne, Australia 4. Murdoch Children’s Research Institute, Australia 5. Department of Global Health, School of Public Health, University of Washington, USA

Paper six has been published in General Hospital Psychiatry (Campbell, Bruno, Darke, and Degenhardt 2015)

160

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.1 Copyright Statement I certify that this publication was a direct result of my research towards this PhD, and that reproduction in this thesis does not breach copyright regulations.

Campbell, G., Bruno, R., Darke, S., & Degenhardt, L. (2015). Associations of borderline personality with pain, problems with medications and suicidality in a community sample of chronic non-cancer pain patients prescribed opioids for pain. General Hospital Psychiatry, 37(5), 434-440.

Gabrielle Campbell November, 2015

161

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.2 Preamble As seen throughout this thesis, problematic use and suicidality is common amongst people prescribed opioids for CNCP. A particularly clinically important finding from Chapter 4 was the high prevalence of people in the POINT cohort who screened positive for BPD, 19.8%. On a review of the literature, although there were a number of studies that suggested a higher prevalence of BPD in people living with CNCP compared with the general population, they were very few in number and many were limited by small sample sizes. With the well documented association of BPD with both problematic substance use and suicidality (Maloney et al. 2007; Maloney et al. 2009; Trull et al. 2000), it was clear that further investigation was warranted. This final empirical Chapter is one of the largest studies to comprehensively examine the association between BPD and problematic medication use and suicidality in people living with CNCP.

162

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.3 Abstract Objective: Borderline personality disorder (BPD) is common in patients with chronic non-cancer pain (CNCP). BPD patients often report worse pain and are more likely to abuse opioid medication. Although the prevalence of suicidality is high in both CNCP patients and those with BPD, no studies have examined the interrelationship of BPD, CNCP and suicidality. This article aims to examine the prevalence and associations of BPD in a large community sample of CNCP patients and the association with medication problems and suicidality.

Methods: Data from a national sample of 978 CNCP patients prescribed pharmaceutical opioids for CNCP. The screener from the International Classification of Diseases, version 10, International Personality Disorder Examination was used to identify patients with symptoms of BPD.

Results: One in five CNCP patients (19.1%) screened positive for BPD. BPD was associated with a number of demographic and clinical features, such as daily benzodiazepine use, and was independently associated with lifetime pharmaceutical opioid dependence [odds ratio (OR) 2.49, 95% confidence interval (95% CI) 1.42–4.38], past 12-month suicidal thoughts (OR 2.9, 95% CI 1.90–4.39) and lifetime suicide attempts (OR 3.19, 95% CI 2.16–4.72).

Conclusions: BPD symptoms were prevalent among people prescribed opioids for CNCP and are associated with a number of adverse consequences. Further, those screening positive were at elevated risk of suicidal behaviours. Careful opioid prescription monitoring and appropriate referrals by clinicians are warranted in BPD with CNCP.

163

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.4 Introduction Borderline personality disorder (BPD) is characterized by a longstanding, pervasive pattern of affective dysregulation, identity disturbances and interpersonal dysfunction and is associated with high levels of distress and suffering in the individual, as well as difficulties in functioning(Gunderson 2009). The prevalence of BPD in patients with chronic pain has been estimated at 30% (Sansone and Sansone 2012), notably higher than 2% in the general population (Jackson and Burgess 2000).

While self-injurious behaviour is common among people with BPD, 50–80% report no pain associated with this behaviour (Bohus et al. 2000). On the other hand, they have been found to be over-represented in patients with chronic pain. BPD has also been found to be associated with past-year chronic back pain, frequent/severe headaches and other chronic pain conditions (McWilliams and Higgins 2013). Sansone refers to this as the ‘pain paradox’ (Sansone and Sansone 2007), suggests that chronic pain, as opposed to self-injurious behaviours, is not under the individuals control and, thus, poorly tolerated in people with BPD.

People with BPD are often high users of health services and access general practitioners and specialists more frequently than those without BPD. They also present more frequently to pain clinics or present with more pain problems in primary care (Kinney et al. 1993; Sansone et al. 2001). Further, emerging research suggests that BPD patients may over-utilize their opioid medication in an effort to cope with chronic pain (Sansone and Sansone 2007). Due to the inflexible, pervasive and maladaptive behaviours of the patient with BPD and the high health service utilization, the BPD patient is often described as a ‘difficult patient’ (Kalira, Treisman, and Clark 2013). To date, research has been limited: specifically, a recent review found only eight studies on the association between BPD and chronic pain from 1994 to 2011 (Sansone and Sansone 2012), only a few have examined medication non-adherence and dependence (Sansone and Sansone 2012).

164

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

Another concern is the high rates of suicidal behaviours in both chronic pain and BPD. It has been estimated that chronic pain patients are two-three times more likely to engage in suicidal behaviours compared with those without chronic pain (Tang and Crane 2006). ]. Further, at least three quarters of BPD patients attempt suicide, and approximately 10% eventually complete suicide (Black et al. 2004). This rate is approximately 400 times greater than the general population (Oumaya et al. 2008). With such high rates of suicidality in patients with chronic pain and patients with BPD, and limited research to date, the association between BPD, chronic pain and suicidality warrants investigation.

The Pain and Opioids IN Treatment (POINT) study is a cohort of people in the community who have been prescribed strong opioids for chronic pain. POINT provides an opportunity to examine BPD, chronic pain and suicidality in a large community- based cohort. The aims of the current paper are to:

1. Determine the proportion of community based chronic pain patients who are taking opioids who screen positively for BPD; 2. Examine the relationship between BPD and pain and medication related factors, such as pain severity and interference, oral morphine equivalents (OME), medication non-adherence and dependence in chronic pain patients; and 3. Examine the relationship between BPD and suicidality in chronic pain patients.

7.5 Materials and Methods The study was approved by the Human Research Ethics Committee of the University of New South Wales (HREC reference: # HC12149). The study also received A1 Australian National Pharmacy Guild Approval to approach pharmacists to assist with recruitment of participants (Approval n. 815). Full details of the study design have been published elsewhere (Campbell et al. 2014) and characteristics (Campbell, Nielsen, Bruno, et al. 2015) of this cohort have been described in detail elsewhere.

165

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.5.1 Recruitment

POINT participants were recruited through community pharmacies across Australia. Of the 5,745 pharmacies with available contact information, 5,332 were contacted by researchers (93%) and 1,868 agreed to assist with recruitment (33%). Participants were eligible if they were: 18 years or older; competent in English; mentally and physically able to partake in telephone and self-complete interviews; without serious cognitive impairments; living with chronic non-cancer pain; and taking prescribed Schedule 8 opioids for CNCP for more than six weeks. Schedule 8 is an Australian classification of drugs of dependence that are subject to additional regulatory controls regarding their manufacture, supply, distribution, possession and use (Therapeutic Goods Administration 2013). Schedule 8 opioids include morphine, oxycodone, buprenorphine, methadone and hydromorphone. A history of injecting drug use (IDU) was not an exclusion criterion, but people currently prescribed pharmaceutical opioids for opioid substitution therapy (OST) for heroin dependence or for cancer were not eligible. Of the 2,725 participants referred into the study, 2,091 were assessed for eligibility (78%). 1,873 participants were deemed eligible (90%) and 1,514 completed the baseline interview (81%). Full details of the recruitment methods are published elsewhere (Campbell et al. 2014).

Phone interviews were conducted by assistants who had a minimum three-year health or psychology degree. Interviewers had received training in the survey instrument and were provided glossaries of chronic pain medications and conditions. Participants also completed self-complete surveys and a seven day medication diary. In the current study n=747 returned their self-complete survey and 724 returned their medication diary.

7.5.2 Measures

The measures, tools, and domains collected were based on recommendations made under the auspices of the Initiative on Methods, Measurement, and Pain Assessment

166

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

in Clinical Trials (IMMPACT) (Dworkin et al. 2005; Turk et al. 2003). Full details of the measures used in the study have been reported elsewhere (Campbell et al. 2014).

Of the 1,514 participants enrolled in the POINT study, 978 completed the BPD module. The BPD screener (see Appendix M, page 285, question D39-D53) was adapted from the International Personality Disorder Examination (IPDE) and was used in the Australian National Survey of Mental Health and Wellbeing (NSMHWB (Loranger et al. 1994). The performance of the BPD screener within the NSMHWB has been studied at length (Lewin et al. 2005). The screener uses the International Classification of Diseases, version 10 (ICD-10) criteria to assess whether a set of 10 borderline traits were ‘true’ or ‘false’ about the participants typical personality. Participants were classified as screening positively for BPD if they endorsed three or more of the symptom criteria, and they indicated that the symptoms interfered with their life (Loranger et al. 1994). Due to the association between BPD and suicide, the symptom ‘I've never threatened suicide or injured myself on purpose’ was removed from the classification. This only slightly reduced the sample that screened positive for BPD from n=195 to n=187.

It must be noted that there are two common methods of scoring the IPDE for BPD: firstly the simple categorical scoring method in which a score of three or more indicates that the person has ‘failed the screen for that disorder’ and should be followed up with an interview. A second scoring method is the criterion-based scoring that mirrors the ICD-10 diagnostic criterion. To screen positive for BPD using this method, the person must score three or more on the five impulsive items and score two or more on the borderline-unstable five items. They must also indicate positive to questions on the ‘pervasiveness’ and ‘associated disability’ items. These two methods have a tendency to over- and under-identify positive (simple scoring and criterion scoring, respectively) cases. A suggested scoring system is to use the simple criterion of three or more and the additional question of pervasiveness (Lewin et al. 2005). Since the categorical scoring system has been used in other studies using the IPDE for BPD in

167

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

chronic pain (Braden and Sullivan 2008; McWilliams and Higgins 2013), in the current study, we have used the categorical scoring with the addition of the pervasiveness item so that our study is comparable. In the current study, not including the suicide item for the BPD screener, 33.5% [95% confidence interval (95% CI) 30.6–36.6] would screen positive using the simple categorical scoring (a score of three or more), and 4.5% (95% CI 3.4–6.0) would screen positive using the criterion scoring. The use of the suggested method by Lewin et al. (score of three more and a positive on the pervasiveness item) resulted in 19.1% (95% CI 16.8–21.7) screening positive for BPD.

7.5.2.2 Pain and pain-related measures

Participants were asked about chronic pain conditions they suffered from in the past 12-months. They were also asked the age of onset of the pain condition and how long they had been in pain. Current pain severity and pain interference were measured by the Brief Pain Inventory (Cleeland 1991), as a continuous score out of 10. A higher score indicates more pain severity or interference. The Pain Self-Efficacy Questionnaire (Nicholas 2007; Nicholas, Asghari, and Blyth 2008) produced a continuous score out of 60, with a higher score indicating more confidence in managing life despite pain.

7.5.2.3 Physical health

The Short Form 12 (SF12) is based on population norms and measures physical and mental health functioning over the previous four weeks, with a mean score of 50 and a SD of 10 (Sanderson K and G. 2002). One item from the World Health Organisation Quality of Life (WHOQOL-BREF) questionnaire (Skevington, Lotfy, and O'Connell 2004) were used a measures of self-rated health: “How would you rate your quality of life” dichotomised to very poor/poor and neither good nor poor/good/very good. Both these were completed in the self-complete questionnaire by participants.

168

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.5.2.4 Medications

Oral morphine equivalent (OME) daily doses were estimated (Nielsen et al. 2014) using available references (Therapeutic Guidelines Limited 2013; Australian Medicines Handbook 2013; Faculaty of Pain Medicine ANZCA 2014). Self-reported opioid use was obtained from a medication diary completed over a one week period as part of the self-complete questionnaire mailed to participants. Of the 978 participants in the current study, 724 returned their questionnaire and OME could be calculated.

The Prescribed Opioids Difficulty Scale (PODS) was used to measure participants' current problems and concerns about using prescribed opioids (Sullivan 2010). A cut- off of ≥8 was considered a moderate score in which participants would have endorsed at least two difficulties out of the 15 items (Sullivan et al. 2010). Scores on the Opioid Related Behaviours in Treatment scale (ORBIT), a 10-item measure of aberrant behaviours such as doctor shopping, diversion and other examples of unsanctioned use of medications, were converted into a dichotomous variable based on endorsement of at least one item (Larance et al. 2015).

7.5.2.5 Mental health and substance use

Participants were asked if they had suffered from depression, anxiety or panic attacks or from post-traumatic stress disorder (PTSD) in the previous past years. Current depression and generalized anxiety disorder were measured by the PHQ-9 and GAD-7 modules of the Patient Health Questionnaire (Kroenke et al. 2010). Validated cut-offs were used as follows: moderate to severe depression was defined as a score of ≥ 10 on the PHQ-9 (Kroenke, Spitzer, and Williams 2001); moderate to severe anxiety was defined as a score of ≥ 10 on the GAD-7 (Spitzer et al. 2006).

Five childhood maltreatment questions were asked if they had experienced sexual abuse, physical abuse, or emotional abuse before the age of 16, based on questions by Sansone (2012 (Sansone, Whitecar, and Wiederman 2009)).

169

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.5.2.6 Suicidality

The suicidality questions were based on questions used in the 2007 NSMHWB (Slade et al. 2009). These questions were adopted from the World Mental Health (WMH) Composite International Diagnostic Interview (CIDI) (Kessler and Üstün 2004) suicidality module.

Participants were asked about lifetime and past 12-month presence of suicidal thoughts, plans and attempts. For each of the behaviours, participants were asked for age of onset and age of most recent behaviour. Participants who had made a suicide plan in the past 12-months, or had made a lifetime suicide attempt, were asked about the method used. These were grouped into non-poisoning (fire a gun, carbon monoxide, cut wrists/stab, jump from height, jump in front of train/bus, strangulation/hanging, suffocation, drowning) or poisoning (opioid and non-opioid) methods. People who had made a suicide attempt were asked to select one of three statements. These were; I made a serious attempt to kill myself and it was only luck that I did not succeed; I tried to kill myself, but knew that the method was not foolproof; and my attempt was a cry for help. I did not intend to die

170

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.5.3 Statistical analysis

All analyses were planned a priori independent of the data and were conducted using STATA, version 12.0 (Stata Corporation, College Station, TX, USA). Means, standard deviations and were used where data were normal; medians, inter-quartile ranges were used where the data were not normally distributed. Odds ratios and 95% confidence intervals were used for comparisons of those who screened positive to BPD and those that did not. A significance cut-off p< 0.05 was used in all analyses.

Hierarchical regression models were conducted to examine whether BPD was independently associated with pain severity, interference, medication non-adherence, overdose, lifetime pharmaceutical opioid dependence (Table 7-3), past 12-month suicidal thoughts and lifetime suicide attempts (Table 7-4) after controlling for: model 1: socio-demographic factors (age, sex, unemployed); model 2: socio-demographic factors, pain related factors (time in pain, number of chronic conditions, lifetime or past 12-month presence of arthritis, chronic back/ Neck problems, frequent headaches/migraines, visceral pain, fibromyalgia, pain severity, pain interference, SF 12 physical health score); model 3: socio-demographic factors, pain related factors, mental health (lifetime/12m major depression, lifetime/12m GAD, lifetime/12m PTSD); model 4: socio-demographic factors, pain related factors, mental health, substance use (lifetime alcohol use disorder, lifetime drug use disorders). The hierarchical regressions were based on a reduced sample size of 749, based on the people who had returned their medication diaries and self-complete surveys.

Suicidality was defined in current study as thoughts or attempts. In hierarchical regression models, due to small numbers, lifetime accidental overdose (n=163) was used instead of past 12-month overdose (n=25), and lifetime suicide attempts (n=192) were used instead of past 12-month attempts (n=22). Past 12-month suicidal ideation was used (n=193). Odds ratios and 95% confidence intervals and p -values are presented.

171

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.6 Results 7.6.1 Prevalence and correlates of BPD in a sample of chronic pain patients

Of the 978 participants who completed the BPD screener, 19.1% (95% CI 16.8–21.7) screened positive for BPD. BPD was associated with being younger and those who screened positively had 2.6 times the odds of being unemployed. There were no gender differences (Table 7-1).

172

Table 7-1: Socio-demographic and clinical characteristics of people with and without borderline personality disorder in a sample of people with chronic pain

Total Borderline Personality Borderline personality Comparison of those with borderline and those absent present n=187 without N=978 n=791 n % (95%CI) N (%/range/sd) n (%/range/sd) OR 95% CI P value Socio-demographics %Gender (male) 437 44.7 41.6-47.8) 348 44.0 (40.6-47.8) 89 47.6 (40.5-54.8) 1.16 0.84-1.59 0.37 Age (mean SD) 978 57.5 (13.6) 791 58.7 (13.6) 187 52.4 (11.9) 0.97 0.95-0.98 <0.001 % Unemployed 498 50.9 (47.8-54.1) 368 46.5 (43.0-50.0) 130 69.5 (62.5-75.7) 2.62 1.86-3.69 <0.001 Pain and physical health Years in pain (MDN-IQR) 978 10 (5-21) 791 11 (5-21) 187 10 (4-21) 1.00 1.00-1.00 0.16 % Arthritis 12m 602 61.6 (58.5-64.6) 491 62.1 (58.6-65.4) 111 59.4 (52.1-66.2) 0.89 0.64-1.23 0.49 % Chronic back neck 12m 767 78.4 (75.7-81.0 610 77.1 (74.1-79.9) 157 84.0 (77.9-88.6) 1.55 1.02-2.37 0.04 % Migraines and headaches 12m 271 27.7 (25.0-30.6) 193 24.4 (21.5-27.5) 78 41.7 (34.8-49.0) 2.22 1.59-3.09 <0.001 % Visceral pain 12m 260 26.6 (23.9-29.4) 198 25.0 (22.1-28.2) 62 33.2 (26.7-40.3) 1.49 1.05-2.10 0.02 % Fibromyalgia 12m 87 8.9 (7.3-10.8) 61 7.7 (6.0-9.8) 26 13.9 (9.6-19.7) 1.94 1.18-3.15 0.008 No. chronic cond 12m (Mean SD) 978 2.4 (1.1) 791 2.3 (1.1) 187 2.8 (1.2) 1.14 1.23-1.61 <0.001 BPI severity (Mean SD) 978 5.1 (1.8) 791 5.0 (1.8) 187 5.7 (1.5) 1.26 1.15-1.39 <0.001 BPI interference (Mean SD) 978 5.8 (2.3) 791 5.4 (2.3) 187 7.1 (1.8) 1.47 1.35-1.61 <0.001 PSEQ (Mean SD) 978 28.4 (13.4) 791 30.3 (13.3) 187 20.6 (10.4) 0.94 0.93-0.95 <0.001 % Poor self-reported health 266 27.2 (24.5-30.1) 181 22.9 (20.1-25.9) 85 45.5 (38.4-52.7) 2.81 2.01-3.92 <0.001 Change since treatment began 978 4.7 (1.8) 791 4.8 (1.8) 187 4.3 (1.8) 0.85 0.77-0.94 0.001 (mean) SF -12 physical score (mdn IQR) 978 26.3 (22.3-31.1) 791 26.4 (22.2-31.7) 187 26.2 (22.8-30.1) 0.99 0.96-1.01 0.31

173

Table 7-1 (cont): Socio-demographic and clinical characteristics of people with and without borderline personality disorder in a sample of people with chronic pain

Total Borderline Personality Borderline personality Comparison of those with borderline and those absent present n=187 without N=978 n=791 n % (95%CI) N (%/range/sd) n (%/range/sd) OR 95% CI P value Medications and associated problems Oral morphine equivalent 724 75 (37.5-150) 596 68.9 (34.4-135) 128 101.7 (50-180) 1.00 1.00-1.00 <0.001 mg/day* % Daily benzodiazepine 180 18.4 (16.1-21.0) 124 15.7 (13.3-18.4) 56 30.0 (23.8-37.0) 2.30 1.59-3.32 <0.001 % Current anti-depressants 504 51.5 (48.4-54.7) 369 46.7 (43.2-50.1) 135 72.2 (65.3-78.2) 2.97 2.09-4.21 <0.001 % Current antipsychotic 63 6.4 (5.1-8.2) 27 3.4 (2.3-4.9) 36 19.3 (14.2-25.6) 6.75 3.97-11.44 <0.001 PODS (intermediate-high score) 487 62.0 (58.5-65.3) 373 58.2 (54.3-62.0) 114 78.6 (71.1-84.6) 2.64 1.72-4.05 <0.001 ORBIT score Mdn (iqr) 976 0 (0-1) 790 0 (0-1) 186 1 (0-2) 1.13 1.06-1.22 0.001 % Accidental overdose ever 163 16.7 (14.5-19.1) 116 14.7 (12.4-17.3) 47 25.1 (19.4-31.9) 1.95 1.33-2.87 0.001 % Accidental overdose past 12m 25 2.6 (1.7-3.8) 14 1.8 (1.0-3.0) 11 5.9 (3.3-10.4) 3.47 1.59-7.77 0.003 *of those that returned their medication diary diary and/or self-complete survey, n=747

174

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

There were a number of differences in pain-related factors between people who screened positive for BPD and people that did not screen positive. People with BPD were more likely to report past 12-month problems with chronic back and/or neck problems, migraines and/or headaches, fibromyalgia and visceral pain. Further, BPD was associated with greater pain severity, interference and number of pain conditions within the past 12-months, compared with people that did not screen positive for BPD. BPD was also associated with poorer self-reported health and a lower impression of treatment change, though there were no differences in terms of physical functioning, as measured by the SF12, nor was there any difference in length of time in pain according to BPD screening status.

In terms of medications, BPD was associated with a higher OME, current benzodiazepines and antidepressants. People who screened positive for BPD had 2.3 times the odds of taking benzodiazepines daily and 6.8 times the odds of currently being prescribed an antipsychotic (Table 7-1). BPD was also associated with more problems associated with opioid use, more non-adherence behaviours and more side effects and overdose. People with BPD had 3.5 times the odds of reporting a past 12- month accidental overdose compared with people without BPD.

People with BPD had 9.2 times the odds of meeting criteria for current moderate-to- severe depression and 10.6 times the odds of meeting current moderate-to-severe anxiety. BPD was also associated with PTSD and childhood abuse and lower social support (Table 7-2). People with BPD had 6.6 times the odds of reporting past 12- month suicidal ideation and 7.9 times the odds of reporting a past 12-month suicide attempt, compared with people without BPD. They also had 4.9 times the odds of reporting self-harm compared with people without BPD

175

Table 7-2: Socio-demographic and clinical characteristics of people with and without borderline personality disorder in a sample of people with chronic pain

Total Borderline Personality Borderline personality Comparison of those with borderline absent present n=187 and those without N=978 n=791 n % (95%CI) N % (95%CI) n % (95%CI) OR 95% CI P value Mental heath % Mod-severe depression 460 47.0 (43.9-50.2) 301 38.1 (34.7-41.5) 159 85.0 (79.1-89.5) 9.24 6.03-14.16 <0.001 % Mod-severe anxiety 222 22.8 (20.3-25.6) 106 13.5 (11.3-16.1) 116 62.4 (55.1-69.1) 10.63 7.41-15.24 <0.000 % PTSD symptoms past month 155 15.9 (13.7-18.3) 74 9.4 (7.5-11.6) 81 43.3 (36.3-50.6) 7.40 5.09-10.78 <0.001 % Childhood sexual abuse 236 24.1 (21.5-26.9) 167 21.1 (18.4-24.1) 69 36.9 (30.2-44.1) 2.18 1.55-3.08 <0.000 % Childhood physical abuse 312 31.9 (29.0-34.9) 218 27.6 (24.6-30.8) 94 50.3 (43.1-57.4) 2.66 1.92-3.68 <0.001 % Childhood emotional abuse 400 40.9 (37.9-44.0) 282 35.7 (32.4-39.1) 118 63.1 (55.9-69.8) 3.09 2.22-4.30 <0.000 MOS social support (mean SD) 3.3 (1.1) 3.7 (1.1) 2.8 (1.1) 0.62 0.52-0.73 <0.001 Suicidality and self-harm % Past 12m suicidal thoughts 193 19.7 (17.4-22.4) 101 12.8 (10.6-15.3) 92 49.2 (42.0-56.4) 6.62 4.64-9.43 <0.000 % Past 12m suicidal attempts 22 2.3 (1.5-3.4) 8 1.0 (0.5-2.0) 14 7.5 (4.5-12.3) 7.92 3.27-19.17 <0.001 % Lifetime suicide attempt 192 19.6 (17.3-22.2) 105 13.3 (11.1-15.8) 87 46.5 (39.4-53.8) 5.68 3.99-8.09 <0.000 % Ever self-harm 95 9.8 (8.1-11.8) 49 6.3 (4.8-8.2) 46 24.7 (19.0-31.5) 4.93 3.17-7.66 <0.001 Substance use % Lifetime ICD pharm opioid 76 7.8 (6.2-9.6) 40 5.1 (3.7-6.8) 36 19.3 (14.1-25.6) 4.48 2.76-7.26 <0.000 dependence % AUD lifetime dependence 126 12.9 (10.9-15.1) 68 8.6 (6.8-10.8) 58 31.0 (24.7-38.1) 4.78 3.21-7.11 <0.001 % Past 12m alcohol – 216 22.1 (19.6-24.8) 171 21.6 (18.9-24.6) 45 24.1 (18.4-30.8) 1.15 0.79-1.67 0.47 occasional/regular risky drinking (=>5) % Lifetime illicit drug 98 10.0 (8.3-12.1) 56 7.1 (5.5-9.1) 42 22.5 (17.0-29.1) 3.80 2.45-5.89 <0.000 dependence % Past 12-month illicit drug use 27 2.8 (1.9-4.0) 17 2.2 (1.3-3.4) 10 5.4 (2.9-9.7) 2.57 1.16-5.71 0.02

176

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.6.2 BPD and the association to pain severity, interference medication non-adherence, OME and pharmaceutical opioid dependence

In hierarchical regression models (Table 7-3) adjusting for demographics (age, gender, unemployment), pain-related factors (time in pain, number of chronic conditions, pain severity, pain interference types of pain, SF12 physical health score) and mental health (lifetime 12-month major depression, lifetime/12-month GAD, lifetime/12-month PTSD), BPD was independently associated pain interference and medication non- adherence and overdose. BPD was not independently associated with accidental overdose after adjusting further for mental health, nor medication non-adherence, after adjusting for lifetime alcohol and illicit substance dependence. BPD was independently associated with lifetime pharmaceutical opioid dependence after adjusting for demographics, pain-related factors, mental health and lifetime alcohol and illicit substance use dependence. After adjusting for pain-related factors, BPD was not independently associated with pain severity or OME.

177

Pain severity1. Pain interference1. Non-adherence Score intermediate-severe on Lifetime accidental Lifetime opioid dependence* PODS (problems with opioid overdose* use) b b 95% CI P value b 95% CI P value b 95% CI P value OR 95% CI P value OR 95% CI P OR 95% CI P value value Unadjusted 0.81 0.49-1.14 <0.001 1.72 1.32-2.13 <0.001 0.89 0.49-1.30 <0.001 2.72 1.77-4.21 <0.001 1.95 1.33-2.87 0.001 4.48 2.76-7.26 <0.001 Model 1 0.67 0.34-0.99 <0.001 1.45 1.05-1.86 <0.001 0.69 0.29-1.10 <0.001 2.21 1.42-3.45 <0.001 1.71 1.16-2.53 0.007 3.72 2.26-6.12 <0.001 Model 2 -0.01 -0.30-0.27 0.97 0.94 0.63-1.26 <0.001 0.62 0.19-1.04 0.005 1.72 1.08-2.74 0.02 1.57 1.04-2.36 0.028 3.76 2.26-6.29 <0.001 Model 3 0.14 -0.16-0.44 0.35 0.51 0.19-0.84 <0.001 0.48 0.03-0.92 0.04 1.58 0.98-2.57 0.06 1.35 0.88-2.07 0.153 3.17 1.85-5.45 <0.001 Model 4** 0.13 0.32-0.58 0.56 1.44 0.88-2.37 0.15 1.09 0.69-1.72 0.705 2.52 1.43-4.47 0.002 a Orbit score b PODS intermediate/severe Model 1: socio-demographic factors (age, sex, unemployed); Model 2: socio-demographic factors, pain related factors (time in pain, number of chronic conditions, pain severity1, pain interference1 past 12-month arthritis, back/neck problems, chronic headaches/migraines, visceral pain, fibromyalgia, SF 12 physical health score); Model 3: socio-demographic factors, pain related factors, mental health (lifetime/12m major depression, lifetime/12m GAD, lifetime/12m PTSD) Model 4: socio-demographic factors, pain related factors, mental health, substance use (lifetime alcohol use disorder, lifetime drug use disorders) * Used lifetime number of pain conditions and pain conditions and lifetime mental health ** Substance use included for non-adherence, problems with medication use, lifetime overdose and lifetime opioid dependence 1 In the pain severity regression, pain interference was included and in the pain interference model, pain severity was included

178

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.6.3 Suicidality and borderline personality disorder in a sample of chronic pain patients

In hierarchical regression models, BPD was independently associated with past 12- month suicidal ideation and lifetime suicide attempts (after adjusting for age, gender, unemployment), pain-related factors (time in pain, number of chronic conditions, pain severity, pain interference, types of pain, SF12 physical health score) and mental health (lifetime/12-month major depression, lifetime/12-month GAD, lifetime/12- month PTSD) as well as substance use (Table 7-4). After adjusting for these factors, people with BPD had 2.9 times the odds of experiencing past 12-month suicidal ideation and 3.2 times the odds of experiencing a lifetime suicide attempt.

179

Table 7-4: Logistic Regression Models#: Independent associations of BPD to past 12-month suicidal ideation and lifetime suicide attempts (n=916)

Past 12-month suicidal thoughts* Lifetime suicide attempts** OR 95% CI P value OR 95% CI P value Unadjusted 6.59 4.62-9.41 <0.001 5.68 3.99-8.09 <0.001 Model 1 5.77 4.01-8.30 <0.001 5.19 3.61-7.47 <0.001 Model 2 4.25 2.91-6.23 <0.001 4.90 3.39-7.09 <0.001 Model 3 2.71 1.81-4.06 <0.001 3.51 2.39-5.17 <0.001 Model 4 2.89 1.90-4.41 <0.001 3.16 2.13-4.69 <0.001 #Model 1: socio-demographic factors (age, sex, unemployed); Model 2: socio-demographic factors, pain related factors (time in pain, number of chronic conditions, pain severity, pain interference, past 12-month arthritis, back/neck problems, chronic headaches/migraines, visceral pain, fibromyalgia SF 12 physical health score); Model 3: socio-demographic factors, pain related factors, mental health (lifetime/12m major depression, lifetime/12m GAD, lifetime/12m PTSD) ) Model 4: socio-demographic factors, pain related factors, mental health, substance use (lifetime alcohol use disorder**, lifetime drug use disorders**) * past 12 month variables **Used lifetime variables

180

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.7 Discussion The current study is one of the largest to examine BPD in a sample of community patients prescribed strong opioids for CNCP. Consistent with previous research, the prevalence of BPD (19.1%) was considerably higher than the estimated 2% in the general population. BPD was associated with a number of pain and clinical features. Importantly, the presence of BPD was independently associated with higher pain interference scores and pharmaceutical opioid dependence after controlling for demographics, pain-related factors, mental health and substance use. Further, this study is one of the first to examine the interrelationship between BPD, chronic pain and suicidality. BPD was independently associated with past 12-month suicidal ideation and lifetime suicide attempts.

The complexity of the chronic pain patient, in terms of multiple social, mental health and physical wellbeing issues (Campbell, Nielsen, Bruno, et al. 2015; Rogers et al. 2013), makes treatment complicated. With the addition of BPD, which has often been associated with a ‘difficult patient’ (Kalira, Treisman, and Clark 2013) treatment can become more difficult. While objective measures of pain are often used in health settings, the BPD patient's self-regulation difficulties and hypersensitivity to their own internal environment mean that their perception of endogenous pain is likely to be unnaturally augmented (Sansone and Sansone 2007) and determination of pain levels, and thus treatment, is difficult. Interestingly, unlike previous research (Kalira, Treisman, and Clark 2013; Sansone and Sansone 2012), BPD was not independently associated with pain severity after adjusting for other pain factors (time in pain, number of chronic conditions, pain severity, pain interference and types of pain). We believe, however, that this is one of the first studies to control for a number of other pain-related factors and this may be the reason for the inconsistency. Further, recent research suggests that the association of BPD to pain can be explained by negative affect (Tragesser, Bruns, and Disorbio 2010). In univariate analyses, BPD patients had higher pain scores, had more pain interference and generally rated their health as poorer compared with people without BPD, although there were no differences in

181

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

physical health and functioning between those who screened positive for BPD and those who did not. Catastrophizing is a common maladaptive coping strategy in both patients with BPD and chronic pain (Kalira, Treisman, and Clark 2013) and this amplification of negative feeling can lead to more intense and distressing symptoms.

BPD was independently associated with medication non-adherence and pharmaceutical opioid dependence, after adjusting for demographics, pain-related factors, mental health and substance use, though BPD was no longer associated with non-adherence after adjusting for substance use history. These findings, especially that BPD was independently associated with lifetime pharmaceutical opioid dependence over and above a history of substance dependence, raise important clinical implications. It is essential that clinicians collect detailed client histories, including mental health and substance use history, so they can develop tailored treatment plans for the management of chronic pain. It has been suggested that BPD be screened for primary care and specialist settings and a conservative management approach with regard to pharmaceutical opioids adopted (Sansone and Sansone 2012). This might entail, continual reminding that opioids will not provide complete relief, use of non- opioid medications and treatments and careful monitoring of medications (Sansone and Sansone 2007).

The current study is one of the first to examine suicidality in the chronic pain patient with BPD. While it has been documented that suicidality is more common in both the BPD and chronic pain patient separately, to the author's knowledge, this interrelationship has not been examined previously. In the current study, people with BPD had 6.6 times the odds of reporting past 12-month suicidal thoughts, and 7.9 times the odds of reporting a past 12-month attempt, compared with people without BPD. Further, BPD was independently associated with past 12-month suicidal thoughts and lifetime attempts after adjusting for general suicide risk factors, such as socio- demographics, pain-related factors, mental health and substance use

182

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

The finding of an association between BPD and greater use of other medications, especially taking benzodiazepines on a daily basis, raises some serious clinical implications. Not only is the combination of opioids and benzodiazepine medications associated with a high risk of overdose (Jann, Kennedy, and Lopez 2014; Jones, Mack, and Paulozzi 2013) but is especially problematic in a subpopulation with a high rate of suicidality. Reducing the availability of means to complete suicide has been found to be an effective prevention strategy (World Health Organization 2014). Care must be taken, though, that chronic pain patients have access to possible treatments, and this includes pharmaceutical opioids. Again, this highlights the need of careful monitoring of pharmaceutical opioids in patients with BPD and chronic pain and careful consideration of the benefits and risks of prescribing opioids and other types of medications need to be determined by clinicians. This finding also highlights the need to obtain a patient's full medication history. It is unknown, in the current study, if participants obtained all their medication from the one doctor or different doctors, i.e., GPs and psychiatrists. While prescription monitoring systems may be useful at providing a more detailed prescription history of a patient, often benzodiazepines are not included in this monitoring (Nielsen et al. 2013; Shand et al. 2013).

Referrals to other services may also be useful in reducing suicidality in patients with BPD. Two approaches have been found to reduce suicide attempts in people with BPD (Black et al. 2004), include, dialectical behaviour therapy (Binks et al. 2006; Linehan, Heard, and Armstrong 1993) and the psychoanalytic partial hospital program (Bateman and Fonagy 1999). It must be noted, though, these programs had no effect on completed suicide rates. Again, there is a need to assess the risk of suicide in patients with chronic pain, and an assessment of BPD may be useful for this in light of the finding that BPD was independently associated with suicidality. As the chronic pain patient has many complex social, physical and mental health issues, a tailored, multifaceted and multidisciplinary treatment approach is necessary, as in this case, the ‘one size fits all’ approach is not appropriate..

183

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

7.7.1 Limitations

As with any research, there are limitations in the current study that need to be considered. Firstly, while the IPDE Questionnaire has been found to perform satisfactorily (Lewin et al. 2005), caution should be used when interpreting results with an understanding that the IPDE Questionnaire was designed to be a screener and not a diagnostic tool and a positive indication suggests the need for a second-stage interview with a mental health clinician, rather than an indication of diagnoses. In the current study, we use the term BPD as indicating a positive screen to BPD symptoms from the IPDE, not an actual diagnosis. Secondly, the cross-sectional nature of this study means that causality cannot be examined, although through the longitudinal data collected from the POINT study, it is hoped that some of these questions may be answered.

Further, the current paper presents data on suicidal ideation and attempts, not completed suicides. There are different risk factors associated with each of the suicidal behaviours (Nock and Kessler 2006), for example, ideation compared with completed suicides. It must also be acknowledged that, due to all participants in the current sample suffering from CNCP, the current study is unable to determine whether chronic pain in and of itself increases the chances of problematic opioid use and suicidality in people with BPD. Future research is necessary to delineate these factors. Further, due to small numbers, past 12-month suicide attempts could not be included in the multivariate regressions and lifetime attempts were used instead. Obviously, it would be ideal to use more recent behaviours. It may also be argued that the relationship between BPD and suicidality is an artefact of the self-harm item in the BPD. In the current study, this item was not used and the BPD was still associated with suicide, which is consistent with previous studies (Maloney et al. 2009). Finally, the potential biases that may be introduced by the reliance on self-report data need to be considered. Information on chronic medical illnesses or other health problems were not verified through patient records. However, the rates of pain conditions and findings were similar to those in previous research and all participants were informed

184

Chapter seven: Borderline personality disorder features in the POINT sample: associations with problematic opioid use and suicidality

that their responses would be de-identified and confidential. This has been found to enhance the validity of self-reported substance use (Darke 1998).

7.7.2 Conclusions

In this study of chronic pain, patients prescribed strong opioids, one in five screened positive for BPD. The chronic pain patient with BPD symptoms presents with a number of important clinical features including being independently associated with pharmaceutical opioid dependence and suicidal behaviours. Additionally, the high rate of daily benzodiazepine use suggests the need for detailed histories of chronic pain patients to be collected and caution when prescribing opioids and other medications.

185

Chapter eight: Discussion and implications

8 DISCUSSION AND IMPLICATIONS This thesis consisted of: an examination of the prevalence and nature of CNCP and suicidality in the Australia population using a nationally representative survey, and an examination of the socio-demographic and clinical characteristics, and associations with problematic opioids use, suicidality and BPD in a large community sample of people prescribed opioids for CNCP. The research presented is of major clinical and public health interest. CNCP has a major impact on not only the individual in terms of mental health, physical functioning, economic implications, and quality of life, but also an impact on the public in terms of health care utilisation, loss in productivity and associated costs. Additionally, the long-term use of opioids for CNCP treatment has been controversial, due to limited long-term data on their effectiveness and an increase in adverse consequences, such as, medication non-adherence, dependence and overdose (Chou et al. 2015). Finally, although there has been much concern about the misuse, abuse and overdose related to opioids and CNCP, another area that that has received far less attention, though of great clinical importance, is the association of suicide and CNCP (Cheatle 2014b).

The aims of this chapter are to briefly summarise the findings of the chapters comprising of empirical studies and discuss the implications in terms of the major themes of the thesis. These are: sociodemographic and clinical characteristics of people living with CNCP and people prescribed opioids for CNCP (Chapter 2 and 4), suicidality and CNCP (Chapters 2 and 6), pharmaceutical opioid dose and problematic opioid use in patients prescribed opioids for CNCP (Chapter 5) and BPD and CNCP (Chapter 7). Strengths and limitations of the overall thesis will be discussed and clinical and research recommendations will be detailed.

186

Chapter eight: Discussion and implications

8.1 Socio-demographic and clinical characteristics of people living with CNCP and prescribed opioids for CNCP Chapter 2 utilised data from the NSMHWB to examine prevalence of CNCP and associations at the population level. The paper presented in Chapter 2 estimated the prevalence of any chronic pain (i.e. arthritis, back/neck problems and/or migraines) to be 38%, which equates to an estimated 6.2 million people in Australia.

People living with CNCP were more likely to be female, had not completed formal schooling, were unemployed, met criteria for DSM-IV Major Depression, Generalised Anxiety Disorder, Post-Traumatic Stress Disorder and substance use disorders, compared with those not experiencing from CNCP. These results are consistent with previous research that has found that people living with CNCP experience a number of socio-economic factors such as unemployment, and are more likely to experience from mental health problems such as depression, anxiety and PTSD, as well as substance use disorders, than people without CNCP (Blyth et al. 2001; Breivik et al. 2006; Strine and Hootman 2007).

In Chapter 4, the paper described the socio-demographic and clinical characteristics of people prescribed opioids for CNCP, the POINT study (described in Chapter 3). The paper presented in Chapter 4 found, compared with those in the retirement age- group, younger people (i.e. in prime working age, or nearing retirement) were more likely to suffer from problems such as, higher unemployment, a greater proportion of people on incomes less than the aged or disability pension (i.e.

187

Chapter eight: Discussion and implications

The findings from the studies presented in Chapter 2 and Chapter 4 demonstrate the complexities and comorbidities in people living with CNCP in the multitude of financial, physical and mental health of problems faced by people living with CNCP. They are consistent with previous research that has found that not only do people with CNCP have more comorbid problems, compared with the general population (Blyth et al. 2001; Breivik et al. 2006; Strine and Hootman 2007), but also people prescribed opioids for CNCP have more complex profiles and greater comorbidities compared with those not prescribed opioids (Rogers et al. 2013; Breivik et al. 2006). Guidelines suggest caution with prescribing opioids (Chou, Fanciullo, Fine, Adler, et al. 2009), especially people with risk factors such as a history of substance use, or severe psychological problems. It has been found, however, that those with the greatest risk factors associated with adverse events are more likely to be prescribed opioids and a higher doses (Sullivan and Howe 2013). The term to describe this has been coined ‘adverse selection’ (Sullivan and Howe 2013). Although it is probable that ‘adverse selection’ is occurring, the current study and the one by Rogers (2013) are based on a cross-sectional samples, it is important that future epidemiological studies examine this phenomenon to determine whether these problems develop after the introduction of prescribed opioids, or if they are in fact evident before.

The findings in Chapter 2 and 4 support the biopsychosocial theory discussed in Chapter 1. People with pain not only suffer from the physical consequences, but there are also a number or psychological problems, such as mental health and substance use problems and social problems such as a lack of social support, low employment and poor income. As mentioned in Chapter 1, a multidisciplinary approach is the most recommended treatment that encompasses and attempts to treat the multitude of factors. The multidisciplinary approach compromises of specialists from a number of fields, including, pain specialists, psychologists/psychiatrist, physiotherapists/occupational therapists and in some cases addiction specialist. The aim is to approach treatment and pain management in a holistic approach with frequent communication between specialists. This approach usually compromises medication management, graded physical exercise and cognitive and behavioural 188

Chapter eight: Discussion and implications

techniques for management of stress and pain (Schatman 2012). In systematic reviews multidisciplinary approaches have been found to be effective in the treatment of chronic pain (Scascighini et al. 2008) and chronic low back pain (Kamper, Apeldoorn, Chiarotto, Ostelo, et al. 2014). Due to the financial constraints often experienced by people living with CNCP, and as evident in this thesis, it is important that these treatments are widely accessible and affordable.

8.2 Suicidality and chronic non-cancer pain As seen in previous section, people living with CNCP suffer from a number financial, social support, physical functioning and mental health problems. Emerging research suggests there is a link between suicidality and CNCP. The factors associated with suicidality, and whether these include pain-specific factors, remained unclear. Chapter 2 examined the association between suicidality and CNCP at the population level, and Chapter 6 examined suicidality in people prescribed opioids for CNCP and the independent contribution of pain-specific factors such as pain severity, time in pain and pain interference.

Consistent with previous epidemiological studies (Braden and Sullivan 2008; Ratcliffe et al. 2008), the paper presented in Chapter 2 found people living with CNCP were two-three times more likely to engage in suicidal behaviours, compared with those not suffering from CNCP in the general population. The presence of any pain was also independently associated with suicidality over and above socio-demographic and mental health factors. Additionally, the paper in Chapter 2 was the first, which we are aware of, to estimate the contribution of CNCP to suicide rates. Approximately, 50- 60% of people who engaged in suicidal behaviours had a history of CNCP. The paper presented in Chapter 2 was the first to examine suicidality and CNCP in the Australian population.

189

Chapter eight: Discussion and implications

Although previous studies have found a link between certain types of pain conditions and suicidality, over and above mental health problems (Braden and Sullivan 2008; Ilgen et al. 2008; Ratcliffe et al. 2008; Tang and Crane 2006), the study presented in Chapter 2 did not find an association between migraines and suicidality, over and above mental health. It is important to note, however, as mentioned in Chapter 2, there were high rates of comorbid pain conditions. For example, half of people with arthritis also reported suffering from chronic back/neck problems or migraines. Additionally, as presented in Chapter 4, 85% of the POINT sample had experienced more than one pain condition in their life. It may, therefore, be difficult to examine specific pain conditions, except in clinical pain populations where specific conditions can be determined. The high rate of comorbid pain conditions in people living with CNCP suggests that it is the presence of chronic pain, rather than the specific conditions that are associated with elevated rates of suicidality.

Although previous suicide research has identified the importance of mental health factors and their contribution to suicide, the finding in Chapter 2 of chronic pain being independently associated with suicide, over and above mental health, suggests the need for clinicians to be aware of this heightened risk in people presenting with chronic pain, irrespective of their mental health status.

The paper in Chapter 2 presented the suicidality rates for the Australian general population, with and without CNCP, and the paper in Chapter 6 presented suicidality rates in a community sample of people prescribed opioids for pain (the POINT sample). The same measure was used in both the NSMHWB (Chapter 2) and the POINT sample (Chapter 6) and Figure 8-1 compares suicidality across the three samples.

190

Chapter eight: Discussion and implications

0 General Australian Population Australian population in CNCP POINT sample prescribed opioids for CNCP

Lifetime suicide ideation Lifetime suicide plan Lifetime suicide attempt Past 12 month suicide ideation Past 12 month suicide plan Past 12 month suicide attempt

Figure 8-1: Comparison of suicidality in the Australian general population (with and without chronic pain) and the POINT sample

As can be seen in Figure 8-1, rates of suicidality are far greater in those living with CNCP who have been prescribed opioids. These differences are striking and have not been well documented before. It is of concern that people with high rates of suicide- related behaviours have access to potentially lethal opioids. Chapter 6 found that people who had experienced suicidality were more likely to be prescribed higher OME and also concurrently prescribed benzodiazepines. Whilst these were not independently associated with suicidality, the potentially lethal combination of these central nervous system depressant medications is of concern. Furthermore, although the use of opioids in a suicide plan or attempt was not as common as other methods, despite the whole cohort having access to opioids, non-opioid poisoning was the most commonly reported method used in previous suicide attempts. Restriction of access to potentially lethal means has been found to be an effective measure in reducing suicide attempts (Hawton et al. 2013; Mann et al. 2005). This raises important clinical implications in terms of the use of opioids for CNCP in persons with high rates of suicidality. There is a need to balance the benefits and risks associated with opioids for CNCP, especially in people with high rates of suicidality (Madadi and Persaud 2014). In

191

Chapter eight: Discussion and implications

a number of studies examining opioid-related deaths (Darke, Duflou, and Torok 2011; Madadi et al. 2013; Roxburgh et al. 2011) it has been found that suicides as a result of opioids often occur in people prescribed opioids for CNCP. Although, in the current research, the use of opioids was not as common as other methods for suicide, future research is necessary to determine the extent of the problem, and the impact of limiting opioids to a potentially at-risk group.

Researchers have suggested the importance of general and pain-specific factors in suicidality in people with CNCP (Hassett, Aquino, and Ilgen 2014; Hooley, Franklin, and Nock 2014; Tang and Crane 2006). At the time of writing this thesis, however, no studies had examined the independent importance of these factors in a large community sample. The paper presented in Chapter 6 was the first, to our knowledge, to examine the influence of general (e.g. comorbid mental health problems, previous suicide attempt and gender) and pain-specific factors (such as severity, time in pain, pain interference). The study found that pain-specific factors were less important indicators of suicidal ideation compared with general suicide risk factors such as mental health. It may be that in a sample of patients where chronic pain is entrenched and the impact on quality of life has been considerable, pain-specific factors become less important. Overall, it is the presence of chronic pain and associated factors, such as low employment, high rates of mental health and poor physical functioning, that places the person at high risk for suicidality.

The study presented in Chapter 6 was the first to employ the ideation-to-action framework (Klonsky and May 2014), i.e. to determine factors that led to progression form suicidal ideation to attempt, in a CNCP sample. In a population where suicidality is high, it is useful to be able to identify those that might be more likely to progress from ideation-to-action, as only a minority of people with suicidal ideation make an attempt (Klonsky and May 2014). The study in Chapter 6 found a lower pain self- efficacy score, indicating lower confidence in coping with pain, as the only independent factor associated with progression from past 12-month suicide ideation

192

Chapter eight: Discussion and implications

to suicide attempt. Active, problem-focused coping styles have been found to be associated with a better adjustment to chronic pain (Jensen et al. 1991). Promisingly, coping skills have been found to be effectively modified by psychosocial therapies including, cognitive-behavioural therapy, problem solving therapy, and dialectical behaviour therapy (Mann et al. 2005). To date, however, there has been limited research examining the impact that these therapies have on people with high rates of suicidality and CNCP. Future research needs to examine the effectiveness of these approaches.

The findings of this thesis have important clinical implications in the need for more research and attention to be given to suicide and CNCP in Australia and internationally. Educating health professionals, such as, primary care physicians, pain specialist and mental health professionals in the unique contribution of pain to suicidality, over and above mental health factors is urgently required. Further, more focus on suicide prevention in this area is necessary, and further research is needed to determine the impact of prescribing and limiting opioids to people with high rates of suicidality and CNCP.

8.3 Pharmaceutical opioid use and problematic use of opioids in people living with chronic non-cancer pain The paper in Chapter 5 describes pharmaceutical opioid dose in the POINT cohort. Concerningly, 15% of the cohort was taking daily doses of more than 200mg OME and around 40% were consuming 90mg OME or more daily The study found a clear association between high-level consumption and a range of indicators of poorer functioning, including poor mental health and problematic opioid use (such as higher scores on the PODS, ORBIT and ICD-10 dependence). Although clinical guidelines suggest that opioid dose should be determined on a case-by-case basis (Merrill et al. 2012), caution has been recommended at higher doses (Chou, Fanciullo, Fine, Adler, et al. 2009; Manchikanti, Abdi, Atluri, Balog, Benyamin, Boswell, Brown, Bruel, Bryce, Burks, et al. 2012), higher doses have been associated with an increase in adverse 193

Chapter eight: Discussion and implications

outcomes, such as abuse, misuse and dependence (Chou et al. 2015). It is clear that in the POINT cohort, people being prescribed the highest doses were the ones with the most complex histories, and were also more likely to report medication non- adherence, prescription opioid dependence and concerns around their opioid use. The association of higher opioid consumption with increasing levels of non-adherence (e.g. tampering and non-medical use), dependence and patient-centred problems suggests that monitoring of higher doses by prescribers is warranted (Passik and Kirsh 2008).

As mentioned in Chapter 1 there are a number of different ways that problematic use has been measured in people prescribed opioids for CNCP. The current thesis was one the first to examine the variety of problematic behaviours in the one sample (Vowles et al. 2015). A number of measures were used, including: measures of abuse and dependence according the diagnostic criteria of the ICD-10, the ORBIT for medication non-adherence and aberrant behaviours and the PODS for patient-centred concerns.

The paper presented in Chapter 5 found that a minority met criteria for ICD-10 dependence (lifetime 8.4%, past 12-month 4.8%). Past 12-month dependence was associated with being younger, engaging in more non-adherent behaviours and a history of benzodiazepine dependence. Further, a higher OME was associated with greater likelihood of dependence.

Medication non-adherence and aberrance amongst people prescribed opioids for CNCP was measured in Chapter 5. As mentioned in Chapter 1, many of the measures designed to measure aberrant behaviours and non-adherence have been limited in scope (Smith, Paillard, et al. 2015). In the paper presented in Chapter 4, 38% of the sample reported the engagement in at least one non-adherent or aberrant behaviour in the preceding three months. When the individual items of the measure were presented separately, however, these behaviours were more likely to be behaviours indicative of under or overusing opioids relative to a doctor’s advice (medication non- adherence), rather than aberrant behaviours such as ‘doctor shopping’ and diversion. 194

Chapter eight: Discussion and implications

It is important that these differences are understood to determine which people are likely to engage in non-adherence due to ‘addiction’ issues, and those that are self- regulating their medication due to being undertreated with medications.

The levels of dependence and medication non-adherence in the POINT sample are similar to the rates found by Vowles et al (2015). Although different terminology was used, Vowels et al (2015) found that 20-30% of people prescribed opioids for CNCP misused their opioids and the opioid “addiction” rate was approximately 10%. As the authors of the review noted, there is wide variability in defining problematic use and also in reported rates of problematic use (Vowles et al. 2015). To date, measures of misuse and abuse of opioids in CNCP patients have found to be inadequate (Smith, Paillard, et al. 2015). In this thesis, the ORBIT was used, which has been developed to address the shortcomings of previous instruments (Larance et al. 2015). Future research is needed to determine the usefulness of the ORBIT in comparison to other measures, and ability to predict poor treatment outcomes.

Another issue is the use and classification of misuse, abuse and related events (MARE) (Smith et al. 2013). In a recent systematic review, published after the undertaking of this thesis, the Analgesic, Anaesthetic and Addiction Clinical Trials, Translations, Innovations, Opportunities and Networks (ACTTION) expert panel reviewed the literature and have created mutually exclusive and exhaustive classifications and definitions of MARE’s to increase accuracy and consistency in reporting of MARE events in clinical trials (Smith et al. 2013). Future research examining the categories and developing sound instruments is warranted to understand the nature and extent of problematic opioid use in CNCP patients.

While most problematic opioid behaviour has focused on the perspective of the clinician, the PODS measures patient-centred problems and concerns. Two-thirds of the POINT sample scored moderate-to-high for patient-centred problems and concerns associated with their opioid use. Further, all participants who met criteria for past 12- 195

Chapter eight: Discussion and implications

month ICD-10 dependence, scored moderate-to-high on the PODS. This high rate of patient-centred concerns suggests that many people prescribed opioids for CNCP have a number of self-identified problems with their medication. Importantly, those on opioid doses of less than 20 OME were less likely to score moderate-to-high than those on 21-90mg, whereas, those at the highest doses (91-199 OME and over 200) were significantly more likely to score moderate-to-high. It is therefore important that these concerns are discussed with health professionals and routinely screened for and addressed throughout treatment.

The results of the study presented in Chapter 5 suggest the need for education and training of pain specialists in the area of addiction, to be able to identify those at risk of developing problems, and also the need to include addiction specialist in the multidisciplinary approach to pain. Long-term research is lacking (Chou et al. 2015) and is necessary to identify the benefits and risks of opioid use for CNCP. Although this thesis did not provide prospective data, the findings were still important as many of the issues had not previously been examined, including all in one sample, in detail before and provide a basis on which future research can be developed. The POINT study is a two-year prospective study and it is hoped that the data collected will enable better understanding of causality in terms of problematic opioid use. With the higher OME doses associated with greater mental health problems and poorer functioning, and greater problematic use, research is also necessary to understand the concept of ‘adverse selection’.

8.4 Borderline Personality Disorder and chronic non-cancer pain As mentioned in Chapter 1, emerging research suggests that BPD is associated with CNCP (Sansone and Sansone 2012). Chapter 7 presented one of the largest and most comprehensive studies to date, examining the association of BPD and pain, problematic opioid use and suicidality in people prescribed opioids for CNCP. Screening positive for BPD was associated with a number of pain and clinical features. Importantly, the presence of BPD was independently associated with higher pain

196

Chapter eight: Discussion and implications

interference scores and pharmaceutical opioid dependence after controlling for demographics, pain-related factors, mental health and substance use. Further, this study was one of the first to examine the interrelationship between BPD, chronic pain and suicidality. BPD was independently associated with past 12-month suicidal ideation and lifetime suicide attempts in people living with CNCP.

The complexity of issues faced by many chronic pain patients, in terms of multiple social, mental health and physical wellbeing issues (Campbell, Nielsen, Bruno, et al. 2015; Rogers et al. 2013), makes treatment complicated. Treatment can become even more difficult in people with borderline personality features. Catastrophizing is a common maladaptive coping strategy in patients with both BPD and chronic pain (Kalira, Treisman, and Clark 2013) and this amplification of negative feeling can lead to more intense and distressing symptoms. Further, the finding of an association between BPD and greater use of other medications, especially taking benzodiazepines on a daily basis, raises some serious clinical implications. Not only is the combination of opioids and benzodiazepine medications associated with a high-risk of overdose (Jann, Kennedy, and Lopez 2014; Jones, Mack, and Paulozzi 2013) but is especially problematic in a sub-population with a high rate of suicidality.

It has been suggested that BPD be screened for in primary care and specialist settings and a conservative management approach with regard to pharmaceutical opioids adopted (Kalira, Treisman, and Clark 2013; Sansone and Sansone 2012). A recent systematic review found that Dialectal Behaviour Therapy (DBT) therapy can be effective in the treatment of BPD and again suggests the need for a multidisciplinary approach encompassing mental health professionals (Stoffers et al. 2012). DBT is similar to CBT, however there is a focus on how the person interacts and responds to others, as opposed to in CBT where is it mainly about the individuals thoughts and behaviours.. As mentioned previously, reducing the availability of means to complete suicide has been found to be an effective prevention strategy (Mann et al. 2005; World Health Organization 2014). There is, however, a need to balance the risks and benefits

197

Chapter eight: Discussion and implications

of opioids for CNCP in this sub-group and the need for greater access and affordability of other treatments, such as psychological treatments.

8.5 Strengths and limitations There are a number of strengths and limitations in the current thesis. Although these have been detailed in previous chapters, it is important to consider the major strengths and limitations of the thesis as a whole.

8.5.1 Strengths

A major strength of this thesis is the use of data from two samples, a population-based sample and a smaller clinical population. Although the paper presented in Chapter 2 is population-based and provides detailed information on DSM-IV mental health and substance use problems, the data collected on chronic pain and pain-related factors were less detailed, and limited the ability to examine the effect of pain-specific factors such as pain severity, interference and length of time in pain at the population level. On the other hand, the POINT study, presented in Chapters 3-7 was designed to examine important factors in chronic pain and carefully selected and validated measures were selected based on the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) (Dworkin et al. 2005; Turk et al. 2003). The combination of these two data sources enabled an examination of population prevalence of chronic pain and suicidality and the POINT study consisted of detailed clinical measures, in a number of areas such as socio-demographics, pain measures, physical functioning, mental health, social support, and substance use. The combination of these approaches provides a detailed picture of the current status of CNCP in Australia.

Definitions of problematic use have been difficult (Smith et al. 2013). The current thesis examined problematic use in a number of ways, and the methods were well reported in the studies. The papers presented in Chapter 4, 5 and 7 used a number of 198

Chapter eight: Discussion and implications

methods of examining problematic use. These included using the CIDI, a structured clinical interview, in examining the ICD and DSM criteria for substance dependence and harmful use/abuse, the ORBIT scale to measure medication non-adherence and aberrant behaviours, and the PODS to measure patient centred concerns with opioid medication. The POINT study is one of the largest to examine all these different measures of problematic use of opioids in a community sample of people prescribed them for CNCP. Previous studies on problematic use of opioids in CNCP have often occurred in pain clinics (which focus on a clinical population), in clinical trials (which often exclude individuals with complex comorbidities) or through administrative datasets (although useful at the population level, can often have limited detailed patients histories), limiting their representativeness and generalisability.

Further, the POINT study included a 7-day medication diary that participants self- completed. This diary included all medications consumed over the preceding 7-days, including over-the-counter medication, and also the doses consumed for each day. The collection of medications obtained without prescription, i.e. over-the-counter opioids sold in pharmacies, which are often not included in routinely collected administration data but may contribute to risk, and therefore addressing a common limitation when relying on administrations data. Further, unlike administrative datasets, which are often based on dispensed medications, the POINT study is unique in its ability to capture actual consumption.

Although there may be a concern of the representativeness, and therefore generalisability of the POINT study, additional data were collected during recruitment to address these issues, which are detailed in Chapter 6. Demographic data collected comparing all opioid customers who attended recruiting pharmacies while the POINT sample was being recruited had striking similarities in terms of gender, age-group and opioid prescribed, and is therefore, reassuring. Additionally, a comparison of people who completed the study and those who were eligible but did not enrol in study (See

199

Chapter eight: Discussion and implications

Chapter 6), found similarities between the groups and again is reassuring that the POINT cohort is similar to community pain patients prescribed opioids.

With emerging research in the field examining suicidality and CNCP, many of the studies have been limited by small clinical samples, or large studies with less detailed information of pain-specific factors collected. The paper presented in Chapter 6 is one of the largest, to date, to examine the independent associations of general and pain- specific risk factors of suicidality in people living with CNCP. Further, the paper presented in Chapter 6 is also the first that employ the ideation-to-action framework in people living with CNCP. In a sample of people where suicidal ideation is high, it is important to be able to identify factors that may increase ones risk of attempting suicide compared with those who have the thoughts but do not act on those thoughts. Finally, the paper presented in Chapter 7 has been one of the largest and most comprehensive in examining the association of BPD, problematic opioid use and suicidality in CNCP.

8.5.2 Limitations

Despite the strengths of this thesis, there are limitations that need to be noted. One of the major limitations is the potential biases that may be introduced by the reliance on self-report data. Information on chronic medical illnesses or other health problems was not verified through patient records or biological and physical assessments. The rates of pain conditions and findings were, however, similar to those in previous research (Blyth et al. 2001; Johannes et al. 2010; Reid et al. 2011). Further, with regards to pain, in the absence of a gold-standard measure of pain, self-report has been the accepted method of collections (Stomski, Mackintosh, and Stanley 2010). Further, with regards to substance misuse, all participants were informed that their responses would be de-identified and confidential, which enhances the validity of self- reported substance use (Darke 1998).

200

Chapter eight: Discussion and implications

Another limitation is that all papers presented cross-sectional data which, although useful for examining associations, does not allow for determination of causality. The findings in the current thesis are, however, important in that they examined many issues, such as suicidality and BPD in much greater details then previously. The findings of this thesis are therefore important in determining the important factors that require further research. Researchers have called for the need for prospective data in both examining opioid use in CNCP and suicide and CNCP (Chou et al. 2015; Tang and Crane 2006) and the findings of this thesis will be useful in determining the focus of research in prospective studies.

The current thesis used ICD-10 criteria for opioid dependence. Although this diagnostic system is commonly used in illicit substance use research, some have argued that it is not relevant to the pain medicine field (Ballantyne 2015; Smith, Paillard, et al. 2015). It has been suggested that using traditional models of dependence, i.e. ICD and DSM, with patients who are being prescribed opioids are inadequate in determining dependence amongst this sample (Smith et al. 2013). Tolerance and withdrawal often occur in patients prescribed opioids, though there may be no sign of “addiction”. Patients may develop physiological dependence, though not psychological dependence. A consensus statement by the American Academy of Pain Medicine, American Pain Society and The American Society of Addiction Medicine recommend the following definitions in prescribed medication for pain (Savage et al. 2001). ‘Addiction’, as one or more of the following, impaired control over drug use, compulsive use, continued use despite harm, and cravings. ‘Physical dependence’ is a state of adaptation that is manifested by withdrawal after reduction or cessation of medication. ‘Tolerance’, as a state of adaptation in which exposure induces changes which diminish the drug’s effects over time. Further, recent recommendations of the ACTTION and definitions of mutually exclusive categories for problematic use are in line with those mentioned above (Smith, Paillard, et al. 2015). It is important future research examines the importance of the definitions in determining treatment outcomes in people prescribed opioids for CNCP.

201

Chapter eight: Discussion and implications

Finally, the current thesis provided one of the largest and most extensive examinations of BPD in CNCP and associations with problematic use and suicidality (Chapter 7). Although the IPDEQ has been found to perform satisfactorily (Lewin et al. 2005), caution should be used when interpreting results with an understanding that the IPDEQ was designed to be a screener and not a diagnostic tool. Future research should use clinical structured measures to determine the prevalence of BPD in people living with CNCP and its association with problematic use opioid use and suicidality.

8.6 Clinical recommendations

8.6.1 Multidisciplinary approach to CNCP treatment

As seen throughout this thesis, many people suffering from CNCP also encounter a number of other problems such as psychological disturbances, poor physical functioning and substance use problems. In line with the biopsychosocial theory, the multidisciplinary approach compromises of specialists from a number of fields, (including, pain specialists, mental health professionals and physical therapists) and pain management is seen as a holistic approach with frequent communication between the specialists. Systematic reviews suggest this approach is effective (Eccleston and Morley 2014; Noble et al. 2010) and guidelines recommend when pain is accompanied by these comorbidities, a multidisciplinary approach should be employed (Chou 2009).

It is important to note, however, that access and affordability of this approach needs to be considered. Although most opioid medications and other medications in Australia are subsidised through the Pharmaceutical Benefits Scheme (a scheme that provides affordable access to a wide range of medications), access to pain clinics is more difficult, with long waiting times in public clinics and is costly for private clinics (Hogg et al. 2012). There is a need to increase the number of multidisciplinary pain treatment programs and increase the affordability and accessibility of these treatments.

202

Chapter eight: Discussion and implications

8.6.2 Education and training of general practitioners

With a lack of speciality services, a lack of specialist in the pain medicine field, and access and affordability issues, many patients present to primary care for treatment and management of their CNCP. Many primary care physicians, however, lack confidence in their ability to treat the complexity of patients with CNCP (Peppin et al. 2015; Jamison et al. 2014; Spitz et al. 2011). It is important that primary care physicians receive appropriate and specialised education and training in the management of CNCP especially being able to target multidisciplinary treatments to those sub-groups with greater comorbidities.

Further, the primary care physician has limited time and training or access to resources to effectively evaluate, treat and monitor CNCP patients (Cheatle and Barker 2014). Although there are a number of guidelines and recommendations for opioid prescribing in CNCP to assist primary care physicians, uptake of these guidelines are generally low (Holliday et al. 2013) possibly due to challenges in prescriber confidence in managing identified risks. Appropriate training in CNCP and addiction medicine is essential for primary care physicians to treat patients with CNCP. A recent review found that even short addiction medicine training programs can be effective in improving knowledge, skills and attitudes (Ayu et al. 2015). Short courses on addiction medicine and CNCP should be made widely available to already practising GPs, although both should begin to be taught as core competencies in medical training.

Finally, as has been previously mentioned, one of the most effective means in reducing suicide rates is the education of primary care physicians (Mann et al. 2005). This thesis demonstrates, however, that it is not just mental health that is associated with suicidality, but also the presence chronic pain. It is important that primary care physicians receive appropriate training in identifying patients most at risk of suicidality.

203

Chapter eight: Discussion and implications

8.6.3 Guidelines for opioid prescribing in CNCP

Despite there being a lack of long-term data on the effectiveness of opioids for CNCP an expert panel has suggested that they may be effective therapy for carefully selected and monitored CNCP patients (Chou, Fanciullo, Fine, Adler, et al. 2009). In a systematic review and critical appraisal of opioid prescribing for CNCP (Nuckols et al. 2014) there are a number of universal recommendations that are relevant to the current thesis.

1. Patient selection and risk stratification A detailed assessment of the patient at the initial appointment is necessary to be able to tailor an appropriate treatment plan (Chou, Fanciullo, Fine, Adler, et al. 2009). Detailed information of a patient’s physical, mental health and substance use history, including family history of substance problems. Also as identified in this thesis screening for suicidality and BPD is needed. These measures need to be brief and freely available to reduce the burden on primary care physicians.

A recent model, the complexity model (Peppin et al. 2015), proposes that patients be stratified by need, with complicated cases requiring the need of a pain specialists, less complex cases be managed by primary care physicians with support of the pain specialists, and those with the lowest risk profiles can be easily and effectively managed by primary care physicians. Differing from other risk measures, such as the opioid risk tool, the complexity model stratifies patients on a number of physical, medical and psychological factors. These domains and risk stratification are based on literature which has identified factors that are associated with poorer treatment outcomes. It is important to note that this model has yet to be tested and, as demonstrated by this thesis, it is important to also include suicidality as a risk factor, which is currently missing from the model. Since limiting access to lethal means is an effective suicide prevention strategy (Mann et al. 2005), it is important that this is screened for and the risk and benefits of prescribing opioids determined. Nonetheless, encompassing the individual and their comorbidities in CNCP management, as proposed by the biopsychosocial model, is certainly a step forward. 204

Chapter eight: Discussion and implications

After a detailed initial assessment it is necessary to stratify the person’s risk of adverse events such as problematic opioid use, poor treatment outcomes and suicidality. If, after careful assessment, opioid therapy is seen as part of the treatment plan, it is important that high-risk patients, those with histories of substance use or psychiatric difficulties, are monitored more frequently (Chou, Fanciullo, Fine, Adler, et al. 2009). As seen throughout this thesis and in other research (Rogers et al. 2013), many people receiving the highest doses also have the greatest substance use histories or more mental health problems. It is therefore important, that those with the greatest risk factors, i.e. history of substance use and mental health problems are monitored more frequently (Chou, Fanciullo, Fine, Adler, et al. 2009). It is important that future research empirically tests this assertion, and whether more frequent monitoring is useful in this group of high-risk people.

2. Constant monitoring and revaluation It is important for primary care physicians to understand that many of the comorbidities faced by CNCP patients are dynamic. Although a person may have a low risk profile at the beginning of treatment, it is possible that this can change throughout treatment, especially if long-term opioid therapy continues for a number of years. It is, therefore, useful for primary care physicians assess the 5A’s of pain treatment on a regular basis. The 5 as (as seen in Chapter 3) are, activity, analgesia, adverse effects, aberrant behaviours and affect. These should be re-evaluated regularly and treatment amended accordingly.

Another important aspect that needs monitoring is the use of other medications, especially benzodiazepines. As seen in Chapter 4, the use of prescription benzodiazepines was high, with 17% reporting the use of them daily, despite guidelines recommending caution of concurrent benzodiazepines and opioids (Nuckols et al. 2014). It is unclear from the current thesis if the participants were receiving these from the primary care physician, or from other health professionals, such as psychiatrists. The implementation of prescription drug monitoring programmes, that

205

Chapter eight: Discussion and implications

also monitored benzodiazepines, would aid in identifying patients that were on risky concurrent use of medications (Jann, Kennedy, and Lopez 2014).

3. Upper dosing thresholds In the POINT cohort studied in this thesis, over 40% were on opioid doses over 90 OME, including 15% on over 200 OME. Both these thresholds have been associated with adverse outcomes such as medication non-adherence, aberrant behaviours and overdose (Chou et al. 2015; Nuckols et al. 2014). In cases where patients are receiving a high dose of opioids, it is important that they are monitored frequently, including improvements in pain and physical functioning to determine the benefits and risks of such high doses.

8.7 Research recommendations This thesis has attempted to address issues of problematic opioid use. As has been mentioned, however, there are a number of ways of determining problematic opioid use. The recent recommendations by the ACTTION group of definitions of mutually exclusive categories for problematic use (Smith, Paillard, et al. 2015), including addiction, physical dependence and tolerance is certainly a step forward. A clinician is to determine the intent of the patient before deciding which MARE category tis most appropriate.. Following these recommendations for addiction, physical dependence and tolerance; 25.7% of the POINT cohort met these criteria for addiction, 32.4% for physical dependence and 17.3% for tolerance, though determining the intent of the patient, and therefore the most appropriate MARE category is not possible in the POINT sample. Compared with 8.4% of the POINT cohort meeting lifetime ICD-10 criteria for pharmaceutical opioid dependence, however, it is clear that future research is necessary to determine which definitions are most clinically useful for identifying those with poorer long-term outcomes, i.e. addiction. Prospective studies, including the POINT study, will be important in identifying the most clinically useful definitions for poor treatment outcomes. Finally, as has been mentioned, measures of aberrant behaviour and non-adherence have been limited, there is a need to further develop

206

Chapter eight: Discussion and implications

and validate appropriate measures of medication non-adherence and aberrant behaviours.

Although this thesis has made a substantial contribution in delineating the factors associated with problematic opioid use, suicidality and BPD in people living with CNCP, future research will be necessary to aid in better understanding of these factors. The data presented in this thesis was cross-sectional and causality is difficult to determine. The POINT study is a two-year prospective study and future research examining the trajectories and benefits and risks of long-term outcomes of opioids for CNCP in the POINT cohort will greatly add to our knowledge and understanding. Additionally, this thesis has increased our understanding of the association of CNCP and suicidality. It is important for future suicide prevention research to examine the impact of limiting opioids to people living with CNCP with a high-risk of suicidality, and the effectiveness of psychological therapies on improving one’s ability to cope with pain, and thereby potentially reducing their risk of suicide.

Finally, as mentioned, the study of BPD, suicidality and problematic opioid use in CNCP (Chapter 7) found some important findings though as mentioned, these are preliminary as a screener for BPD was used in the current thesis. Future research should examine the prevalence and associated characteristics of people living with CNCP diagnosed with BPD.

207

Chapter nine: Conclusions

9 CONCLUSIONS The overall aim of this thesis was to address gaps in knowledge and understanding of the complexities of people living with CNCP, and on problematic opioid use and suicidality. Although there has been a large amount of public and professional concern surrounding the use of opioids in CNCP and suicidality in CNCP, research to date has been limited. The thesis utilised nationally representative data to estimate the prevalence of CNCP in the Australian population and the association of suicidality and CNCP. The thesis demonstrated that, not only is CNCP prevalent in the community, but that people living with CNCP have much higher levels of mental health comorbidities, and are more likely to engage in suicide-related behaviours. The thesis also utilised data from a large community sample of people prescribed opioids for CNCP and is one of the largest and most detailed studies to date. This research clearly shows that CNCP patients prescribed opioids have a great number of medical and mental health comorbidities, and have complex clinical profiles that need to be addressed in their treatment plans.

A significant minority reported problematic opioid use (non-adherence, patient centred concerns and problems and dependence), and the thesis also found an apparent dose-response associated problematic opioid use. Problematic opioid use was associated with being younger and a history of substance disorders. Future research is necessary to determine which definitions of problematic use are the most useful in determining poor clinical outcomes.

Further, suicidality was common in people living with CNCP, was especially elevated in people prescribed opioids for CNCP. When suicidality is common it is important to determine those at greatest risk of a suicide attempt, therefore, this thesis was the first to use the ideation-to-action framework in people with CNCP to determine factors associated with elevating from suicidal ideation to attempt. A person’s ability to cope with pain was the only independent factor associated with elevating from ideation to

208

Chapter nine: Conclusions

attempt in the previous 12-months. Encouragingly, cognitive coping ability can be modified through psychological treatment and this finding supports the use of a multidisciplinary approach to treatment. Finally, BPD was found to be common in this sample and was independently associated with prescription opioid dependence and suicidality. The screening of patients is essential in order to determine the most appropriate treatment plan to CNCP.

Overall, this thesis supports the biopsychosocial theory of pain and the necessity of a multidisciplinary approach to treatment. Screening and stratifying patients, not only in terms of adverse outcomes, but also in terms of their overall physical, medical and psychological complexity, at their initial treatment, and throughout treatment, and careful monitoring of treatment goals is necessary to ensure the best, and most appropriate treatment plan is utilised.

209

Chapter ten: References

10 REFERENCES Access Economics. 2007. The high price of pain: the economic impact of persistent pain in Australia (MBF Foundation).

American Academy of Pain Medicine and American Pain Society. 1997. "The use of opioids for the treatment of chronic pain: a consensus statement from the American Academy of Pain Medicine and the American Pain Society." In.

American Psychiatric Association. 2000. Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) (American Psychiatric Association: Washington DC).

———. 2001. Practice guideline for the treatment of patients with borderline personality disorder (American Psychiatric Pub).

———. 2013. Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition) (American Psychiatric Association: Washington, DC).

Andrews, G., Hall, W., Teesson, M., and Henderson, S. 1999. "The mental health of Australians: National Survey of Mental Health and Wellbeing." In. Canberra: Commonwealth Department of Health and Aged Care.

Asmundson, G. J., and Katz, J. (2009). 'Understanding the co‐occurrence of anxiety disorders and chronic pain: state‐of‐the‐art', Depression and anxiety, 26(10), 888-901.

Australian Bureau of Statistics. 2008. "National Survey of Mental Health and Wellbeing 2007: Summary of Results (Cat. No. 4326.0)." In. Canberra: ABS.

———. 2009a. "National Survey of Mental Health and Wellbeing: Users' Guide 2007 (Cat. No. 4326.0)." In. Canberra: ABS.

———. 2009b. "Technical Manual: National Survey of Mental Health and Wellbeing, Confidentialised Unit Record Files (Cat. No.4326.0)." In. Canberra: ABS.

———. 2013. "2071.0 - Reflecting a Nation: Stories from the 2011 Census, 2012–2013 " In. Canberra: Australian Bureau of Statistics,.

Australian Medicines Handbook. 2013. Australian Medicines Handbook (Adelaide).

Ayu, A. P., Schellekens, A. F. A., Iskandar, S., Pinxten, L., and De Jong, C. A. J. (2015). 'Effectiveness and Organization of Addiction Medicine Training Across the Globe', European Addiction Research, 21(5), 223-39.

Baca, C., and Grant, K. (2007). 'Mortality from opioid analgesics must not be ignored.[comment]', Pain, 128(3), 288; author reply 88-9.

210

Chapter ten: References

Bair, M. J., Robinson, R. L., Katon, W., and Kroenke, K. (2003). 'Depression and pain comorbidity: a literature review', Arch Intern Med, 163(20), 2433-45.

Ballantyne, J. C. (2006). 'Opioids for chronic pain: taking stock', Pain, 125, 3-4.

Ballantyne, J. C. (2007). 'Regulation of opioid prescribing', British Medical Journal, 334(7598), 811-12.

———. (2015). 'Assessing the prevalence of opioid misuse, abuse, and addiction in chronic pain', Pain, 156(4), 567-8.

Ballantyne, J. C., and LaForge, K. S. (2007). 'Opioid dependence and addiction during opioid treatment of chronic pain', Pain, 129(3), 235-55.

Ballantyne, J. C., and Shin, N. S. (2008). 'Efficacy of Opioids for Chronic Pain: A Review of the Evidence', The Clinical Journal of Pain, 24(6), 469-78

Balousek, S., Plane, M. B., and Fleming, M. (2007). 'Prevalence of interpersonal abuse in primary care patients prescribed opioids for chronic pain', Journal of General Internal Medicine, 22(9), 1268-73.

Banta-Green, C. J., Merrill, J. O., Doyle, S. R., Boudreau, D. M., and Calsyn, D. A. (2009). 'Measurement of opioid problems among chronic pain patients in a general medical population', Drug and Alcohol Dependence, 104(1-2), 43-49.

Banta-Green, C. J., Von Korff, M., Sullivan, M. D., Merrill, J. O., Doyle, S. R., and Saunders, K. (2010). 'The prescribed opioids difficulties scale: a patient- centered assessment of problems and concerns', Clinical Journal of Pain, 26(6), 489-97.

Bateman, A., and Fonagy, P. (1999). 'Effectiveness of partial hospitalization in the treatment of borderline personality disorder: a randomized controlled trial', Am J Psychiatry, 156(10), 1563-9.

Belcher, J., Nielsen, S., Campbell, G., Bruno, R., Hoban, B., Larance, B., Lintzeris, N., and Degenhardt, L. (2014). 'Diversion of prescribed opioids by people living with chronic pain: results from an Australian community sample', Drug Alcohol Rev, 33(1), 27-32.

Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D., Ahluvalia, T., Stokes, J., Handelsman, L., Medrano, M., Desmond, D., and Zule, W. (2003). 'Development and validation of a brief screening version of the Childhood Trauma Questionnaire', Child Abuse Negl, 27(2), 169-90.

Beubler, E., Jaksch, W., Devulder, J., Le Poloin, B., Bo Honsen, O., Meynadier, J., Muller- Schwefe, G., Zenz, M., Mac Sullivan, R., O'Brien, T., Eisenberg, E., Mercodante, S., Ventafriddo, V., Varrossi, G., Vielvoye-Kerkmeer, A., Zylicz, B., Breivik, H., Krajnik, M., Lopez, J. C., Puig, M., Rhodin, A., Borgeat, A., Collett, B., Hanna, M., Hunt, T., and Simpson, K. (2006). 'The white paper on opioids and pain: A pan- 211

Chapter ten: References

European challenge: The European white paper on the use of opioids in chronic pain management', Journal of Pain & Palliative Care Pharmacotherapy, 20(3), 79-87.

Binks, C., Fenton, M., McCarthy, L., Lee, T., Adams, C. E., and Duggan, C. (2006). 'Psychological therapies for people with borderline personality disorder', The Cochrane Library.

Bjorner, J., and Turner-Bowker, D. 2009. 'SF-36 and SF-12 Health Surveys.' in M. Kattan (ed.), Encyclopedia of Medical Decision Making (SAGE Publications, Inc.: Thousand Oaks, CA).

Black, D. W., Blum, N., Pfohl, B., and Hale, N. (2004). 'Suicidal Behavior in Borderline Personality Disorder: Prevalence, Risk Factors, Prediction, and Prevention', Journal of Personality Disorders, 18(3), 226-39.

Blanch, B., Pearson, S. A., and Haber, P. S. (2014). 'An overview of the patterns of prescription opioid use, costs and related harms in Australia', Br J Clin Pharmacol, 78(5), 1159-66.

Bloodworth, D. (2005). 'Issues in opioid management', American Journal of Physical Medicine & Rehabilitation, 84(3 Suppl), S42-55.

Blyth, F. M., March, L. M., Brnabic, A. J., and Cousins, M. J. (2004). 'Chronic pain and frequent use of health care', Pain, 111(1), 51-58.

Blyth, F. M., March, L. M., Brnabic, A. J., Jorm, L. R., Williamson, M., and Cousins, M. J. (2001). 'Chronic pain in Australia: a prevalence study', Pain, 89(2-3), 127-34.

Blyth, F. M., March, L. M., and Cousins, M. J. (2003). 'Chronic pain-related disability and use of analgesia and health services in a Sydney community', Medical Journal of Australia, 179(2), 84-87.

Bohnert, A. S., Ilgen, M. A., Trafton, J. A., Kerns, R. D., Eisenberg, A., Ganoczy, D., and Blow, F. C. (2014). 'Trends and regional variation in opioid overdose mortality among veterans health administration patients, fiscal year 2001 to 2009', The Clinical Journal of Pain, 30(7), 605-12.

Bohus, M., Limberger, M., Ebner, U., Glocker, F. X., Schwarz, B., Wernz, M., and Lieb, K. (2000). 'Pain perception during self-reported distress and calmness in patients with borderline personality disorder and self-mutilating behavior', Psychiatry Research, 95(3), 251-60.

Booker, C. L., Harding, S., and Benzeval, M. (2011). 'A systematic review of the effect of retention methods in population-based cohort studies', BMC Public Health, 11, 249.

212

Chapter ten: References

Braden, J. B., and Sullivan, M. D. (2008). 'Suicidal Thoughts and Behavior Among Adults With Self-Reported Pain Conditions in the National Comorbidity Survey Replication', The Journal of Pain, 9(12), 1106-15.

Breivik, H., Collett, B., Ventafridda, V., Cohen, R., and Gallacher, D. (2006). 'Survey of chronic pain in Europe: Prevalence, impact on daily life, and treatment', European Journal of Pain, 10(4), 287-87.

Breivik, H., Eisenberg, E., and O'Brien, T. (2013). 'The individual and societal burden of chronic pain in Europe: the case for strategic prioritisation and action to improve knowledge and availability of appropriate care', BMC Public Health, 13, 1229.

Breslau, N. (1992). 'Migraine, suicidal ideation, and suicide attempts', Neurology, 42(2), 392-92.

Bridge, J. A., Goldstein, T. R., and Brent, D. A. (2006). 'Adolescent suicide and suicidal behavior', J Child Psychol Psychiatry, 47(3-4), 372-94.

Calati, R., Laglaoui Bakhiyi, C., Artero, S., Ilgen, M., and Courtet, P. (2015). 'The impact of physical pain on suicidal thoughts and behaviors: Meta-analyses', Journal of Psychiatric Research, 71, 16-32.

Campbell, G., Bruno, R., Darke, S., and Degenhardt, L. (2015). 'Associations of borderline personality with pain, problems with medications and suicidality in a community sample of chronic non-cancer pain patients prescribed opioids for pain', General Hospital Psychiatry, 37(5), 434-40.

Campbell, G., Bruno, R., Darke, S., Shand, F., Hall, W., Farrell, M., and Degenhardt, L. (2015). 'Prevalence and Correlates of Suicidal Thoughts and Suicide Attempts in People Prescribed Pharmaceutical Opioids for Chronic Pain', Clinical Journal of Pain.

Campbell, G., Darke, S., Bruno, R., and Degenhardt, L. (2015). 'The prevalence and correlates of chronic pain and suicidality in a nationally representative sample', Australian and New Zealand Journal of Psychiatry, 49(9), 803-11.

Campbell, G., Mattick, R., Bruno, R., Larance, B., Nielsen, S., Cohen, M., Lintzeris, N., Shand, F., Hall, W. D., Hoban, B., Kehler, C., Farrell, M., and Degenhardt, L. (2014). 'Cohort protocol paper: the Pain and Opioids In Treatment (POINT) study', BMC Pharmacol Toxicol, 15(17), 17.

Campbell, G., Nielsen, S., Bruno, R., Lintzeris, N., Cohen, M., Hall, W., Larance, B., Mattick, R. P., and Degenhardt, L. (2015). 'The Pain and Opioids IN Treatment study: characteristics of a cohort using opioids to manage chronic non-cancer pain', Pain, 156(2), 231-42.

Campbell, G., Nielsen, S., Larance, B., Bruno, R., Mattick, R., Hall, W., Lintzeris, N., Cohen, M., Smith, K., and Degenhardt, L. (2015). 'Pharmaceutical Opioid Use 213

Chapter ten: References

and Dependence among People Living with Chronic Pain: Associations Observed within the Pain and Opioids in Treatment (POINT) Cohort', Pain Medicine, 16(9), 1745-58.

Cavanagh, J. T., Carson, A. J., Sharpe, M., and Lawrie, S. M. (2003). 'Psychological autopsy studies of suicide: a systematic review', Psychol Med, 33(3), 395-405.

Centers for Disease -Control Prevention. (2008). 'Adult use of prescription opioid pain medications - Utah, 2008', MMWR - Morbidity & Mortality Weekly Report, 59(6), 153-7.

Centre for Disease Control and Prevention. 2006. "Overview of Implementation Protocols for Testing the Washington Group Short Set of Questions on Disability." In, edited by Washington Group on Disability Statistics (WG). Washington DC: Centre for Disease Control and Prevention.

Cheatle, M. D. (2011). 'Depression, Chronic Pain, and Suicide by Overdose: On the Edge', Pain Medicine, 12, S43-S48.

Cheatle, M. D. (2014a). 'Assessing suicide risk in patients with chronic pain and depression', J Fam Pract, 63(6 Suppl), S6-S11.

Cheatle, M. D. (2014b). 'Suicidal ideation in patients with chronic pain: The risk–benefit of pharmacotherapy', Pain, 155(12), 2446-47.

Cheatle, M. D., and Barker, C. (2014). 'Improving opioid prescription practices and reducing patient risk in the primary care setting', Journal of pain research, 7, 301.

Cheatle, M. D., Wasser, T., Foster, C., Olugbodi, A., and Bryan, J. (2014). 'Prevalence of suicidal ideation in patients with chronic non-cancer pain referred to a behaviorally based pain program', Pain Physician, 17(3), E359-67.

Chou, R. (2009). '2009 Clinical Guidelines from the American Pain Society and the American Academy of Pain Medicine on the use of chronic opioid therapy in chronic noncancer pain: what are the key messages for clinical practice?', Pol Arch Med Wewn, 119(7-8), 469-77.

Chou, R., Ballantyne, J. C., Fanciullo, G. J., Fine, P. G., and Miaskowski, C. (2009). 'Research Gaps on Use of Opioids for Chronic Noncancer Pain: Findings From a Review of the Evidence for an American Pain Society and American Academy of Pain Medicine Clinical Practice Guideline', The Journal of Pain, 10(2), 147- 59.e15.

Chou, R., Fanciullo, G. J., Fine, P. G., Adler, J. A., Ballantyne, J. C., Davies, P., Donovan, M. I., Fishbain, D. A., Foley, K. M., Fudin, J., Gilson, A. M., Kelter, A., Mauskop, A., O'Connor, P. G., Passik, S. D., Pasternak, G. W., Portenoy, R. K., Rich, B. A., Roberts, R. G., Todd, K. H., and Miaskowski, C. (2009). 'Clinical Guidelines for

214

Chapter ten: References

the Use of Chronic Opioid Therapy in Chronic Noncancer Pain', The Journal of Pain, 10(2), 113-30.e22.

Chou, R., Fanciullo, G. J., Fine, P. G., Miaskowski, C., Passik, S. D., and Portenoy, R. K. (2009). 'Opioids for chronic noncancer pain: prediction and identification of aberrant drug-related behaviors: a review of the evidence for an American Pain Society and American Academy of Pain Medicine clinical practice guideline', Journal of Pain, 10(2), 131-46.

Chou, R., Turner, J. A., Devine, E. B., Hansen, R. N., Sullivan, S. D., Blazina, I., Dana, T., Bougatsos, C., and Deyo, R. A. (2015). 'The effectiveness and risks of long-term opioid therapy for chronic pain: a systematic review for a National Institutes of Health pathways to prevention workshop', Annals of Internal Medicine.

Clarke, H., Soneji, N., Ko, D. T., Yun, L., and Wijeysundera, D. N. (2014). 'Rates and risk factors for prolonged opioid use after major surgery: population based cohort study', BMJ, 348, g1251.

Cleeland, C. (1991). 'The Brief Pain Inventory (BPI)'.

Cleeland, C. S., and Ryan, K. M. (1994). 'Pain assessment: global use of the Brief Pain Inventory', Ann Acad Med Singapore, 23(2), 129-38.

Cochran, B. N., Flentje, A., Heck, N. C., Van Den Bos, J., Perlman, D., Torres, J., Valuck, R., and Carter, J. (2014). 'Factors predicting development of opioid use disorders among individuals who receive an initial opioid prescription: mathematical modeling using a database of commercially-insured individuals', Drug Alcohol Depend, 138, 202-8.

Cohen, M. L., and Wodak, A. D. (2010). 'The judicious use of opioids in managing chronic noncancer pain', Medicine Today, 11, 10-18.

Compton, W. M., and Volkow, N. D. (2006a). 'Abuse of prescription drugs and the risk of addiction', Drug and Alcohol Dependence, 83, S4-S7.

———. (2006b). 'Major increases in opioid analgesic abuse in the United States: concerns and strategies', Drug and Alcohol Dependence, 81(2), 103-07.

Corran, T. M., Farrell, M. J., Helme, R. D., and Gibson, S. J. (1997). 'The classification of patients with chronic pain: age as a contributing factor', Clinical Journal of Pain, 13(3), 207-14.

Coughlan, G. M., Ridout, K. L., Williams, A. C., and Richardson, P. H. (1995). 'Attrition from a pain management programme', Br J Clin Psychol, 34 ( Pt 3), 471-9.

Currie, S. R., and Wang, J. (2004). 'Chronic back pain and major depression in the general Canadian population', Pain, 107(1-2), 54-60.

215

Chapter ten: References

Darke, S. (1998). 'Self-report among injecting drug users: a review', Drug and Alcohol Dependence, 51(3), 253-63.

Darke, S., Duflou, J., and Torok, M. (2011). 'Toxicology and Characteristics of Fatal Oxycodone Toxicity Cases in New South Wales, Australia 1999–2008*', Journal of Forensic Sciences, 56(3), 690-93.

Darke, S., Williamson, A., Ross, J., Teesson, M., and Lynskey, M. (2004). 'Borderline personality disorder, antisocial personality disorder and risk-taking among heroin users: findings from the Australian Treatment Outcome Study (ATOS)', Drug and Alcohol Dependence, 74(1), 77-83.

Davis, D. A., Luecken, L. J., and Zautra, A. J. (2005). 'Are reports of childhood abuse related to the experience of chronic pain in adulthood?: a meta-analytic review of the literature', The Clinical Journal of Pain, 21(5), 398-405.

Degenhardt, L., Black, E., Breen, C., and Bruno, R. (2006). 'Trends in morphine prescriptions, illicit morphine use and associated harms among regular injecting drug users in Australia', Drug and Alcohol Review, 25(5), 403-12.

Degenhardt, L., Bruno, R., Lintzeris, N., Hall, W., Nielsen, S., Larance, B., Cohen, M., and Campbell, G. (2015). 'Agreement between definitions of pharmaceutical opioid use disorders and dependence in people taking opioids for non-cancer chronic pain (POINT): a cohort study', Lancet Psychiatry, 2(4), 314-22.

Demyttenaere, K., Bruffaerts, R., Lee, S., Posada-Villa, J., Kovess, V., Angermeyer, M. C., Levinson, D., de Girolamo, G., Nakane, H., Mneimneh, Z., Lara, C., de Graaf, R., Scott, K. M., Gureje, O., Stein, D. J., Haro, J. M., Bromet, E. J., Kessler, R. C., Alonso, J., and Von Korff, M. (2007). 'Mental disorders among persons with chronic back or neck pain: Results from the world mental health surveys', Pain, 129(3), 332-42.

Dersh, J., Polatin, P. B., and Gatchel, R. J. (2002). 'Chronic pain and psychopathology: research findings and theoretical considerations', Psychosom Med, 64(5), 773- 86.

Dhalla, I. A., Mamdani, M. M., Sivilotti, M. L., Kopp, A., Qureshi, O., and Juurlink, D. N. (2009). 'Prescribing of opioid analgesics and related mortality before and after the introduction of long-acting oxycodone', CMAJ Canadian Medical Association Journal, 181(12), 891-6.

Donald, M., Dower, J., Correa-Velez, I., and Jones, M. (2006). 'Risk and protective factors for medically serious suicide attempts: a comparison of hospital-based with population-based samples of young adults', Australian and New Zealand Journal of Psychiatry, 40(1), 87-96.

Dube, S. R., Anda, R. F., Felitti, V. J., Chapman, D. P., Williamson, D. F., and Giles, W. H. (2001). 'Childhood abuse, household dysfunction, and the risk of attempted

216

Chapter ten: References

suicide throughout the life span: Findings from the adverse childhood experiences study', JAMA, 286(24), 3089-96.

Dulit, R. A., Fyer, M. R., Haas, G. L., Sullivan, T., and Frances, A. J. (1990). 'Substance use in borderline personality disorder', Am J Psychiatry, 147(8), 1002-7.

Dunn, K. M., Saunders, K. W., Rutter, C. M., Banta-Green, C. J., Merrill, J. O., Sullivan, M. D., Weisner, C. M., Silverberg, M. J., Campbell, C. I., Psaty, B. M., and Von Korff, M. (2010). 'Opioid prescriptions for chronic pain and overdose: a cohort study.[Summary for patients in Ann Intern Med. 2010 Jan 19;152(2):I-42; PMID: 20083811]', Annals of Internal Medicine, 152(2), 85-92.

Dworkin, R. H., Turk, D. C., Farrar, J. T., Haythornthwaite, J. A., Jensen, M. P., Katz, N. P., Kerns, R. D., Stucki, G., Allen, R. R., Bellamy, N., Carr, D. B., Chandler, J., Cowan, P., Dionne, R., Galer, B. S., Hertz, S., Jadad, A. R., Kramer, L. D., Manning, D. C., Martin, S., McCormick, C. G., McDermott, M. P., McGrath, P., Quessy, S., Rappaport, B. A., Robbins, W., Robinson, J. P., Rothman, M., Royal, M. A., Simon, L., Stauffer, J. W., Stein, W., Tollett, J., Wernicke, J., and Witter, J. (2005). 'Core outcome measures for chronic pain clinical trials: IMMPACT recommendations', Pain, 113(1–2), 9-19.

Eccleston, C., and Morley, S. (2014). 'Psychological therapy for adults with longstanding distressing pain and disability', Cochrane Database of Systematic Reviews.

Eccleston, C., Williams, A. C. d. C., and Morley, S. (2009). 'Psychological therapies for the management of chronic pain (excluding headache) in adults', The Cochrane Library.

Edwards, R. R., Smith, M. T., Kudel, I., and Haythornthwaite, J. (2006). 'Pain-related catastrophizing as a risk factor for suicidal ideation in chronic pain', Pain, 126(1–3), 272-79.

Eisenberg, E., McNicol, E. D., and Carr, D. B. (2006). 'Efficacy of mu‐opioid agonists in the treatment of evoked neuropathic pain: Systematic review of randomized controlled trials', European Journal of Pain, 10(8), 667-67.

El-Gabalawy, R., Katz, L. Y., and Sareen, J. (2010). 'Comorbidity and associated severity of borderline personality disorder and physical health conditions in a nationally representative sample', Psychosom Med, 72(7), 641-47.

Elliott, A., Smith, B., Penny, K., Smith, W., and Chambers, W. (1999). 'The epidemiology of chronic pain in the community', Lancet, 354, 1248 - 252.

Engel, G. (1977). 'The need for a new medical model: a challenge for biomedicine', Science, 196(4286), 129-36.

Eriksen, J., Sjogren, P., Bruera, E., Ekholm, O., and Rasmussen, N. K. (2006). 'Critical issues on opioids in chronic non-cancer pain: an epidemiological study', Pain, 125(1-2), 172-9. 217

Chapter ten: References

Euser, A. M., Zoccali, C., Jager, K. J., Dekker, F. W., Euser, A. M., Zoccali, C., Jager, K. J., and Dekker, F. W. (2009). 'Cohort studies: prospective versus retrospective', Nephron, 113(3), c214-7.

Faculaty of Pain Medicine ANZCA. 2014. "Opioid Conversion to oral Morphine Equivalent Daily Dose (oMEDD)." In.

Fischer, B., and Argento, E. (2012). 'Prescription opioid related misuse, harms, diversion and interventions in Canada: a review', Pain Physician, 15(3), ES191- ES203.

Fishbain, D., Cole, B., Lewis, J., Rosomoff, H., and Rosomoff, R. (2008). 'What Percentage of Chronic Nonmalignant Pain Patients Exposed to Chronic Opioid Analgesic Therapy Develop Abuse/Addiction and/or Aberrant Drug Related Behaviors? A Structured Evidence Based Review', Pain Medicine, 9(4), 444-59.

Fishbain, D. A. (1999). 'The association of chronic pain and suicide', Seminars in clinical neuropsychiatry, 4(3), 221-7.

Fishbain, D. A., Cole, B., Lewis, J., Rosomoff, H. L., Rosomoff, R. S., Fishbain, D. A., Cole, B., Lewis, J., Rosomoff, H. L., and Rosomoff, R. S. (2008). 'What percentage of chronic nonmalignant pain patients exposed to chronic opioid analgesic therapy develop abuse/addiction and/or aberrant drug-related behaviors? A structured evidence-based review', Pain Medicine, 9(4), 444-59.

Fishbain, D. A., Cutler, R., Rosomoff, H. L., and Rosomoff, R. S. (1997). 'Chronic pain- associated depression: antecedent or consequence of chronic pain? A review', The Clinical Journal of Pain, 13(2), 116-37.

Fishbain, D. A., Lewis, J. E., and Gao, J. (2014). 'The pain suicidality association: A narrative review', Pain Medicine (United States), 15(11), 1835-49.

Fisher, B. J., Haythornthwaite, J. A., Heinberg, L. J., Clark, M., and Reed, J. (2001). 'Suicidal intent in patients with chronic pain', Pain, 89(2-3), 199-206.

Flor, H., and Turk, D. C. 2011. Chronic Pain: An Integrated Biobehavioral Approach (IASP Press).

Frankenburg, F. R., Fitzmaurice, G. M., and Zanarini, M. C. (2014). 'The use of prescription opioid medication by patients with borderline personality disorder and axis ii comparison subjects: A 10-year follow-up study', Journal of Clinical Psychiatry, 75(4), 357-61.

Frankenburg, F. R., and Zanarini, M. C. (2004). 'The association between borderline personality disorder and chronic medical illnesses, poor health-related lifestyle choices, and costly forms of health care utilization', The Journal of clinical psychiatry, 65(12), 1660-65.

218

Chapter ten: References

Franklin, G. M., Stover, B. D., Turner, J. A., Fulton-Kehoe, D., Wickizer, T. M., Disability Risk Identification Study, C., Franklin, G. M., Stover, B. D., Turner, J. A., Fulton- Kehoe, D., and Wickizer, T. M. (2008). 'Early opioid prescription and subsequent disability among workers with back injuries: the Disability Risk Identification Study Cohort', Spine, 33(2), 199-204.

Fredheim, O. M., Skurtveit, S., Breivik, H., and Borchgrevink, P. C. (2010). 'Increasing use of opioids from 2004 to 2007 - pharmacoepidemiological data from a complete national prescription database in Norway', European Journal of Pain: Ejp, 14(3), 289-94.

Furlan, A. D., Sandoval, J. A., Mailis-Gagnon, A., and Tunks, E. (2006). 'Opioids for chronic noncancer pain: a meta-analysis of effectiveness and side effects', CMAJ Canadian Medical Association Journal, 174(11), 1589-94.

Gaskin, D. J., and Richard, P. (2012). 'The economic costs of pain in the United States', J Pain, 13(8), 715-24.

Gatchel, R. J. (2004). 'Comorbidity of chronic pain and mental health disorders: the biopsychosocial perspective', American Psychologist, 59(8), 795-805.

Gatchel, R. J., Peng, Y. B., Peters, M. L., Fuchs, P. N., and Turk, D. C. (2007). 'The biopsychosocial approach to chronic pain: scientific advances and future directions', Psychol Bull, 133(4), 581-624.

Glajchen, M. (2001). 'Chronic pain: treatment barriers and strategies for clinical practice', The Journal of the American Board of Family Practice, 14(3), 211-8.

Goodacre, L., and McArthur, M. 2013. Rheumatology Practice in Occupational Therapy: Promoting Lifestyle Management (John Wiley & Sons).

Gourlay, D. L., Heit, H. A., and Almahrezi, A. (2005). 'Universal Precautions in Pain Medicine: A Rational Approach to the Treatment of Chronic Pain', Pain Medicine, 6(2), 107-12.

Gross, R., Olfson, M., Gameroff, M., and et al. (2002). 'BOrderline personality disorder in primary care', Archives of Internal Medicine, 162(1), 53-60.

Gunderson, J. G. 2009. Borderline personality disorder: A clinical guide (American Psychiatric Pub).

Gureje, O., Von Korff, M., Simon, G. E., and Gater, R. (1998). 'Persistent pain and wellbeing: a World Health Organization study in primary care', JAMA, 280(2), 147- 51.

Hagg, O., Fritzell, P., and Nordwall, A. (2003). 'The clinical importance of changes in outcome scores after treatment for chronic low back pain', Eur Spine J, 12(1), 12-20.

219

Chapter ten: References

Hariharan, J., Lamb, G. C., Neuner, J. M., Hariharan, J., Lamb, G. C., and Neuner, J. M. (2007). 'Long-term opioid contract use for chronic pain management in primary care practice. A five year experience', Journal of General Internal Medicine, 22(4), 485-90.

Haro, J. M., Arbabzadeh-Bouchez, S., Brugha, T. S., de Girolamo, G., Guyer, M. E., Jin, R., Lepine, J. P., Mazzi, F., Reneses, B., Vilagut, G., Sampson, N. A., and Kessler, R. C. (2006). 'Concordance of the Composite International Diagnostic Interview Version 3.0 (CIDI 3.0) with standardized clinical assessments in the WHO World Mental Health surveys', Int J Methods Psychiatr Res, 15(4), 167-80.

Harris, E. C., and Barraclough, B. (1997). 'Suicide as an outcome for mental disorders. A meta-analysis', The British Journal of Psychiatry, 170(3), 205-28.

Hartung, D. M., Middleton, L., Haxby, D. G., Koder, M., Ketchum, K. L., and Chou, R. (2007). 'Rates of adverse events of long-acting opioids in a state Medicaid program.[Erratum appears in Ann Pharmacother. 2007 Sep;41(9):1552]', Annals of Pharmacotherapy, 41(6), 921-8.

Hasegawa, K., Brown, D. F., Tsugawa, Y., and Camargo, C. A. 2014. "Epidemiology of emergency department visits for opioid overdose: a population-based study." In Mayo Clinic Proceedings, 462-71. Elsevier.

Hassett, A. L., Aquino, J. K., and Ilgen, M. A. (2014). 'The risk of suicide mortality in chronic pain patients', Curr Pain Headache Rep, 18(8), 436.

Hawton, K., Bergen, H., Simkin, S., Dodd, S., Pocock, P., Bernal, W., Gunnell, D., and Kapur, N. (2013). 'Long term effect of reduced pack sizes of paracetamol on poisoning deaths and liver transplant activity in England and Wales: interrupted time series analyses', BMJ: British Medical Journal, 346.

Hinkley, B. S., and Jaremko, M. E. (1994). 'Effects of pain duration on psychosocial adjustment in orthopedic patients: The importance of early diagnosis and treatment of pain', Journal of Pain and Symptom Management, 9(3), 175-85.

Hogg, M. N., Gibson, S., Helou, A., DeGabriele, J., and Farrell, M. J. (2012). 'Waiting in pain: a systematic investigation into the provision of persistent pain services in Australia', Med J Aust, 196(6), 386-90.

Holliday, S., Magin, P., Dunbabin, J., Oldmeadow, C., Henry, J. M., Lintzeris, N., Attia, J., Goode, S., and Dunlop, A. (2013). 'An evaluation of the prescription of opioids for chronic nonmalignant pain by Australian general practitioners', Pain Med, 14(1), 62-74.

Hooley, J. M., Franklin, J. C., and Nock, M. K. (2014). 'Chronic pain and suicide: Understanding the association', Current Pain and Headache Reports, 18(8).

220

Chapter ten: References

Horne, R., Weinman, J., and Hankins, M. (1999). 'The beliefs about medicines questionnaire: The development and evaluation of a new method for assessing the cognitive representation of medication', Psychology & Health, 14(1), 1-24.

Hurst, H., and Bolton, J. (2004). 'Assessing the clinical significance of change scores recorded on subjective outcome measures', Journal of Manipulative and Physiological Therapeutics, 27(1), 26-35.

Ilgen, M. A., Kleinberg, F., Ignacio, R. V., Bohnert, A. S., Valenstein, M., McCarthy, J., Blow, F. C., and Katz, I. (2013a). 'Noncancer pain conditions and risk of suicide', JAMA Psychiatry, 70(7), 692-97.

Ilgen, M. A., Kleinberg, F., Ignacio, R. V., Bohnert, A. S., Valenstein, M., McCarthy, J. F., Blow, F. C., and Katz, I. R. (2013b). 'Noncancer pain conditions and risk of suicide', JAMA Psychiatry, 70(7), 692-97.

Ilgen, M. A., Kleinberg, F., Ignacio, R. V., and et al. (2013). 'NOncancer pain conditions and risk of suicide', JAMA Psychiatry, 70(7), 692-97.

Ilgen, M. A., Zivin, K., Austin, K. L., Bohnert, A. S., Czyz, E. K., Valenstein, M., and Kilbourne, A. M. (2010). 'Severe pain predicts greater likelihood of subsequent suicide', Suicide Life Threat Behav, 40(6), 597-608.

Ilgen, M. A., Zivin, K., McCammon, R. J., and Valenstein, M. (2008). 'Pain and suicidal thoughts, plans and attempts in the United States', General Hospital Psychiatry, 30(6), 521-27.

Jackson, H. J., and Burgess, P. M. (2000). 'Personality disorders in the community: a report from the Australian National Survey of Mental Health and Wellbeing', Social psychiatry and psychiatric epidemiology, 35(12), 531-38.

Jamison, R. N., Sheehan, K. A., Scanlan, E., Matthews, M., and Ross, E. L. (2014). 'Beliefs and attitudes about opioid prescribing and chronic pain management: survey of primary care providers', J Opioid Manag, 10(6), 375-82.

Jann, M., Kennedy, W. K., and Lopez, G. (2014). 'Benzodiazepines a major component in unintentional prescription drug overdoses with opioid analgesics', Journal of pharmacy practice, 27(1), 5-16.

Jensen, M. P., Turner, J. A., Romano, J. M., and Karoly, P. (1991). 'Coping with chronic pain: a critical review of the literature', Pain, 47(3), 249-83.

Jimenez-Rodríguez, I., Garcia-Leiva, J. M., Jimenez-Rodriguez, B. M., Condés-Moreno, E., Rico-Villademoros, F., and Calandre, E. P. Suicidal ideation and the risk of suicide in patients with fibromyalgia: a comparison with non-pain controls and patients suffering from low-back pain (Neuropsychiatr Dis Treat. 2014;10:625- 30. doi:10.2147/NDT.S57596.).

221

Chapter ten: References

Johannes, C. B., Le, T. K., Zhou, X., Johnston, J. A., and Dworkin, R. H. (2010). 'The prevalence of chronic pain in United States adults: results of an Internet-based survey', J Pain, 11(11), 1230-9.

Joiner, T. 2009. Why people die by suicide (Harvard University Press).

Jones, C. M., Mack, K. A., and Paulozzi, L. J. (2013). 'Pharmaceutical overdose deaths, United States, 2010', JAMA, 309(7), 657-59.

Kalira, V., Treisman, G. J., and Clark, M. R. (2013). 'Borderline personality disorder and chronic pain: a practical approach to evaluation and treatment', Current Pain and Headache Reports, 17(8), 1-8.

Kalso, E., Edwards, J. E., Moore, R. A., and McQuay, H. J. (2004). 'Opioids in chronic non-cancer pain: systematic review of efficacy and safety', Pain, 112(3), 372-80.

Kamper, S., Apeldoorn, A., Chiarotto, A., Ostelo, R., Guzman, J., and van Tulder, M. (2014). 'Multidisciplinary treatment for back pain', Health.

Kamper, S., Apeldoorn, A., Chiarotto, A., Smeets, R., Ostelo, R., Guzman, J., and van Tulder, M. (2014). 'Multidisciplinary biopsychosocial rehabilitation for chronic low back pain (Review)'.

Kanzler, K. E., Bryan, C. J., McGeary, D. D., and Morrow, C. E. (2012). 'Suicidal Ideation and Perceived Burdensomeness in Patients with Chronic Pain', Pain Practice, 12(8), 602-09.

Kessler, R. C., and Üstün, T. B. (2004). 'The World Mental Health (WMH) Survey Initiative version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI)', International Journal of Methods in Psychiatric Research, 13(2), 93-121.

Kinney, R. K., Gatchel, R. J., Polatin, P. B., Fogarty, W. T., and Mayer, T. G. (1993). 'Prevalence of psychopathology in acute and chronic low back pain patients', Journal of Occupational Rehabilitation, 3(2), 95-103.

Klonsky, E. D., and May, A. M. (2014). 'Differentiating suicide attempters from suicide ideators: a critical frontier for suicidology research', Suicide and Life- Threatening Behavior, 44(1), 1-5.

Kreidler, S., M., Muller, K. M., Grunwald, G., Ringham, B. M., Coker-Dukowitz, Z., Sakhadeo, U. R., Barón, A. E., and Glueck, D. H. (2013). 'GLIMMPSE: Online Power Computation for Linear Models with and without a Baseline Covariate', Journal of Statistical Software, 54(10).

Kroenke, K., Spitzer, R., and Williams, J. B. W. (2001). 'The PHQ-9: Validity of a Brief Depression Severity Measure', Journal of General Internal Medicine, 16(9), 606- 13.

222

Chapter ten: References

Kroenke, K., Spitzer, R. L., Williams, J. B. W., and Lowe, B. (2010). 'The Patient Health Questionnaire Somatic, Anxiety, and Depressive Symptom Scales: a systematic review', General Hospital Psychiatry, 32(4), 345-59.

Larance, B., Bruno, R., Lintzeris, N., Black, E., Degenhardt, L., Brown, A., Nielsen, S., Dunlop, A., Holland, R., Cohen, M., and Mattick, R. P. (2015). 'Development of the Opioid Related Behaviours In Treatment (ORBIT) scale', Drug and Alcohol Dependence, Accepted 6th November, 2015.

Leong, M., Murnion, B., and Haber, P. (2009a). 'Examination of opioid prescribing in Australia from 1992 to 2007', Internal Medicine Journal, 39(10), 676-81.

Leong, M., Murnion, B., and Haber, P. S. (2009b). 'Examination of opioid prescribing in Australia from 1992 to 2007', Internal Medicine Journal, 39(10), 676-81.

Lewin, T. J., Slade, T., Andrews, G., Carr, V. J., and Hornabrook, C. W. (2005). 'Assessing personality disorders in a national mental health survey', Soc Psychiatry Psychiatr Epidemiol, 40(2), 87-98.

Linehan, M. M., Heard, H. L., and Armstrong, H. E. (1993). 'Naturalistic follow-up of a behavioral treatment for chronically parasuicidal borderline patients', Arch Gen Psychiatry, 50(12), 971-74.

Lintzeris, N. (2013). 'Commentary on Minozzi et al. (2013): Diagnosing opioid dependence in the context of long-term opioid use for chronic pain—counting shades of grey in a black and white world', Addiction, 108(4), 699-700.

Lipman, A. G. (2007). 'Treatment options for chronic pain management: opioids revisited', Managed Care, 16(2 Suppl 3), 5-9.

Loranger, A. W., Sartorius, N., Andreoli, A., Berger, P., Buchheim, P., Channabasavanna, S. M., Coid, B., Dahl, A., Diekstra, R. F., Ferguson, B., and et al. (1994). 'The International Personality Disorder Examination. The World Health Organization/Alcohol, Drug Abuse, and Mental Health Administration international pilot study of personality disorders', Arch Gen Psychiatry, 51(3), 215-24.

Madadi, P., Hildebrandt, D., Lauwers, A. E., and Koren, G. (2013). 'Characteristics of opioid-users whose death was related to opioid-toxicity: a population-based study in Ontario, Canada', PloS one, 8(4), e60600.

Madadi, P., and Persaud, N. (2014). 'Suicide by means of opioid overdose in patients with chronic pain', Current Pain and Headache Reports, 18(11), 1-4.

Maloney, E., Degenhardt, L., Darke, S., Mattick, R. P., and Nelson, E. (2007). 'Suicidal behaviour and associated risk factors among opioid‐dependent individuals: a case–control study', Addiction, 102(12), 1933-41.

223

Chapter ten: References

Maloney, E., Degenhardt, L., Darke, S., and Nelson, E. C. (2009). 'Impulsivity and borderline personality as risk factors for suicide attempts among opioid- dependent individuals', Psychiatry Research, 169(1), 16-21.

Manchikanti, L., Abdi, S., Atluri, S., Balog, C. C., Benyamin, R. M., Boswell, M. V., Brown, K. R., Bruel, B. M., Bryce, D. A., and Burks, P. A. (2012). 'American Society of Interventional Pain Physicians (ASIPP) guidelines for responsible opioid prescribing in chronic non-cancer pain: Part 2--guidance', Pain Physician, 15(3 Suppl), S67-116.

Manchikanti, L., Abdi, S., Atluri, S., Balog, C. C., Benyamin, R. M., Boswell, M. V., Brown, K. R., Bruel, B. M., Bryce, D. A., Burks, P. A., Burton, A. W., Calodney, A. K., Caraway, D. L., Cash, K. A., Christo, P. J., Damron, K. S., Datta, S., Deer, T. R., Diwan, S., Eriator, I., Falco, F. J., Fellows, B., Geffert, S., Gharibo, C. G., Glaser, S. E., Grider, J. S., Hameed, H., Hameed, M., Hansen, H., Harned, M. E., Hayek, S. M., Helm, S., 2nd, Hirsch, J. A., Janata, J. W., Kaye, A. D., Kaye, A. M., Kloth, D. S., Koyyalagunta, D., Lee, M., Malla, Y., Manchikanti, K. N., McManus, C. D., Pampati, V., Parr, A. T., Pasupuleti, R., Patel, V. B., Sehgal, N., Silverman, S. M., Singh, V., Smith, H. S., Snook, L. T., Solanki, D. R., Tracy, D. H., Vallejo, R., and Wargo, B. W. (2012). 'American Society of Interventional Pain Physicians (ASIPP) guidelines for responsible opioid prescribing in chronic non-cancer pain: Part 2--guidance', Pain Physician, 15(3 Suppl), S67-116.

Manchikanti, L., Ailinani, H., Koyyalagunta, D., Datta, S., Singh, V., Eriator, I., Sehgal, N., and Shah, R. (2011). 'A systematic review of randomized trials of long-term opioid management for chronic non-cancer pain', Pain Physician, 14, 91-121.

Mann, J., Apter, A., Bertolote, J., and et al. (2005). 'Suicide prevention strategies: A systematic review', JAMA, 294(16), 2064-74.

Martell, B. A., O'Connor, P. G., Kerns, R. D., Becker, W. C., Morales, K. H., Kosten, T. R., and Fiellin, D. A. (2007). 'Systematic review: opioid treatment for chronic back pain: prevalence, efficacy, and association with addiction', Annals of Internal Medicine, 146(2), 116-27.

Martin, A., Rief, W., Klaiberg, A., and Braehler, E. (2006). 'Validity of the Brief Patient Health Questionnaire Mood Scale (PHQ-9) in the general population', General Hospital Psychiatry, 28(1), 71-77.

Mattick, R. P., and Clarke, J. C. (1998). 'Development and validation of measures of social phobia scrutiny fear and social interaction anxiety', Behav Res Ther, 36(4), 455-70.

Maven Marketing. 2012. 'http://www.mavenmarketing.com.au/list-of-pharmacies-in- australia/'.

McCracken, L., and Iverson, G. (2001). 'Disrupted sleep patterns and daily functioning in patients with chronic pain', Pain research & management: the journal of the

224

Chapter ten: References

Canadian Pain Society= journal de la societe canadienne pour le traitement de la douleur, 7(2), 75-79.

McWilliams, L. A., Cox, B. J., and Enns, M. W. (2003). 'Mood and anxiety disorders associated with chronic pain: an examination in a nationally representative sample', Pain, 106(1-2), 127-33.

McWilliams, L. A., Goodwin, R. D., and Cox, B. J. (2004). 'Depression and anxiety associated with three pain conditions: Results from a nationally representative sample', Pain, 111(1-2), 77-83.

McWilliams, L. A., and Higgins, K. S. (2013). 'Associations Between Pain Conditions and Borderline Personality Disorder Symptoms: Findings From the National Comorbidity Survey Replication', The Clinical Journal of Pain, 29(6), 527-32.

Meilman, P. W. (1984). 'Chronic pain: The nature of the problem', Journal of Orthopaedic & Sports Physical Therapy, 5(6), 307-08.

Menefee, L. A., Cohen, M. J., Anderson, W. R., Doghramji, K., Frank, E. D., and Lee, H. (2000). 'Sleep disturbance and nonmalignant chronic pain: a comprehensive review of the literature', Pain Medicine, 1(2), 156-72.

Merrill, J. O., Von Korff, M., Banta-Green, C. J., Sullivan, M. D., Saunders, K. W., Campbell, C. I., and Weisner, C. (2012). 'Prescribed opioid difficulties, depression and opioid dose among chronic opioid therapy patients', General Hospital Psychiatry, 34(6), 581-87.

Merskey, H., and Bogduk, N. 1994. "Task Force on Taxonomy of the International Association for the Study of Pain. Classification of chronic pain: descriptions of chronic pain syndromes and definition of pain terms. Seattle: IASP, 1994." In.: ISBN 978-0-931092-05-3.

Merskey, H. E. (1986). 'Classification of chronic pain: Descriptions of chronic pain syndromes and definitions of pain terms', Pain.

Minozzi, S., Amato, L., and Davoli, M. (2013). 'Development of dependence following treatment with opioid analgesics for pain relief: a systematic review', Addiction, 108(4), 688-98.

Molton, I. R., and Terrill, A. L. (2014). 'Overview of persistent pain in older adults', American Psychologist, 69(2), 197.

Moriarty, O., McGuire, B. E., and Finn, D. P. (2011). 'The effect of pain on cognitive function: a review of clinical and preclinical research', Prog Neurobiol, 93(3), 385-404.

Morin, C. M., Gibson, D., and Wade, J. (1998). 'Self-reported sleep and mood disturbance in chronic pain patients', The Clinical Journal of Pain, 14(4), 311-14.

225

Chapter ten: References

Moscicki, E. K. (1997). 'Identification of suicide risk factors using epidemiologic studies', Psychiatr Clin North Am, 20(3), 499-517.

Mościcki, E. K. (1994). 'Gender differences in completed and attempted suicides', Annals of Epidemiology, 4(2), 152-58.

Moulin, D. E., Amireh, R., Sharpe, W. K. J., Boyd, D., Merskey, H., and Iezzi, A. (1996). 'Randomised trial of oral morphine for chronic non-cancer pain', The Lancet, 347(8995), 143-47.

Moulin, D. E., Clark, A. J., Speechley, M., and Morley-Forster, P. K. (2002). 'Chronic pain in Canada - Prevalence, treatment, impact and the role of opioid analgesia', Pain Research and Management, 7 (4), 179-84.

Murphy B, Herrman H, Hawthorne G, Pinzone T, and Evert H. 2000. "Australian WHOQoL instruments: User’s manual and interpretation guide." In, edited by Australian WHOQoL Field Study Centre. Melbourne, Australia: Australian WHOQoL Field Study Centre.

National Drug and Alcohol Research Centre. 2012. "A Review of Opioid Prescribing in Tasmania: A Blueprint for the Future." In. Sydney: NDARC, University of New South Wales.

Nicholas, M. K. 1989. "Self-efficacy and chronic pain." In Annual conference of the British Psychological Society. St. Andrews: British Psychological Society.

Nicholas, M. K. (2007). 'The pain self-efficacy questionnaire: Taking pain into account', European Journal of Pain, 11(2), 153-63.

Nicholas, M. K., Asghari, A., and Blyth, F. M. (2008). 'What do the numbers mean? Normative data in chronic pain measures', Pain, 134(1-2), 158-73.

Nielsen, S., Bruno, R., Degenhardt, L., Stoove, M. A., Fischer, J. A., Carruthers, S. J., and Lintzeris, N. (2013). 'The sources of pharmaceuticals for problematic users of benzodiazepines and prescription opioids', Med J Aust, 199(10), 696-9.

Nielsen, S., Degenhardt, L., Hoban, B., and Gisev, N. 2014. "Comparing opioids: A guide to estimating oral morphine equivalents (OME) in research. NDARC Technical Report No. 329." In.: National Drug and Alcohol Research Centre, UNSW Australia.

Nielsen, S., Lintzeris, N., Bruno, R., Campbell, G., Larance, B., Hall, W., Hoban, B., Cohen, M., and Degenhardt, L. (2015). 'Benzodiazepine use amongst chronic pain patients prescribed opioids: associations with pain, physical and mental health and health service utilization', Pain Medicine, doi: 10.1111/pme.12594.

Noble, M., Treadwell, J. R., Tregear, S. J., Coates, V. H., Wiffen, P. J., Akafomo, C., and Schoelles, K. M. (2010). 'Long-term opioid management for chronic noncancer pain', Cochrane Database of Systematic Reviews(1), CD006605. 226

Chapter ten: References

Nock, M. K., Borges, G., Bromet, E. J., Alonso, J., Angermeyer, M., Beautrais, A., Bruffaerts, R., Chiu, W. T., de Girolamo, G., and Gluzman, S. (2008). 'Cross- national prevalence and risk factors for suicidal ideation, plans and attempts', The British Journal of Psychiatry, 192(2), 98-105.

Nock, M. K., Hwang, I., Sampson, N., Kessler, R. C., Angermeyer, M., Beautrais, A., Borges, G., Bromet, E., Bruffaerts, R., and de Girolamo, G. (2009). 'Cross- national analysis of the associations among mental disorders and suicidal behavior: findings from the WHO World Mental Health Surveys', PLoS Medicine, 6(8), e1000123.

Nock, M. K., Hwang, I., Sampson, N. A., and Kessler, R. C. (2010). 'Mental disorders, comorbidity and suicidal behavior: Results from the National Comorbidity Survey Replication', Mol Psychiatry, 15(8), 868-76.

Nock, M. K., and Kessler, R. C. (2006). 'Prevalence of and risk factors for suicide attempts versus suicide gestures: analysis of the National Comorbidity Survey', Journal of abnormal psychology, 115(3), 616.

Nordt, C., Warnke, I., Seifritz, E., and Kawohl, W. (2015). 'Modelling suicide and unemployment: a longitudinal analysis covering 63 countries, 2000–11', The Lancet Psychiatry, 2(3), 239-45.

Nuckols, T. K., Anderson, L., Popescu, I., Diamant, A. L., Doyle, B., Di Capua, P., and Chou, R. (2014). 'Opioid prescribing: a systematic review and critical appraisal of guidelines for chronic pain', Ann Intern Med, 160(1), 38-47.

Ojala, T., Hakkinen, A., Karppinen, J., Sipila, K., Suutama, T., and Piirainen, A. (2014). 'The Dominance of Chronic Pain: A Phenomenological Study', Musculoskeletal Care, n/a-n/a.

Okifuji, A., and Benham, B. (2011). 'Suicidal and Self‐Harm Behaviors in Chronic Pain Patients', Journal of Applied Biobehavioral Research, 16(2), 57-77.

Osborn, M., and Smith, J. A. (2015). 'The personal experience of chronic benign lower back pain: An interpretative phenomenological analysis', British Journal of Pain, 9(1), 65-83.

Oumaya, M., Friedman, S., Pham, A., Abou, A. T., Guelfi, J., and Rouillon, F. (2008). '[Borderline personality disorder, self-mutilation and suicide: literature review]', L'Encephale, 34(5), 452-58.

Passik, S. D., and Kirsh, K. L. (2008). 'The interface between pain and drug abuse and the evolution of strategies to optimize pain management while minimizing drug abuse', Experimental and Clinical Psychopharmacology, 16(5), 400-4.

Passik, S. D., Kirsh, K. L., Whitcomb, L., Portenoy, R. K., Katz, N. P., Kleinman, L., Dodd, S. L., and Schein, J. R. (2004). 'A new tool to assess and document pain

227

Chapter ten: References

outcomes in chronic pain patients receiving opioid therapy', Clin Ther, 26(4), 552-61.

Paulozzi, L. J., Budnitz, D. S., and Xi, Y. (2006). 'Increasing deaths from opioid analgesics in the United States', Pharmacoepidemiology and Drug Safety, 15(9), 618-27.

Peppin, J. F., Cheatle, M. D., Kirsh, K. L., and McCarberg, B. H. (2015). 'The Complexity Model: A Novel Approach to Improve Chronic Pain Care', Pain Medicine, 16(4), 653-66.

Pletcher, M. J., Kertesz, S. G., Sidney, S., Kiefe, C. I., Hulley, S. B., Pletcher, M. J., Kertesz, S. G., Sidney, S., Kiefe, C. I., and Hulley, S. B. (2006). 'Incidence and antecedents of nonmedical prescription opioid use in four US communities. The Coronary Artery Risk Development in Young Adults (CARDIA) prospective cohort study', Drug & Alcohol Dependence, 85(2), 171-6.

Plosker, G. L., and Lyseng-Williamson, K. A. (2012). 'Buprenorphine 5, 10 and 20 mug/h transdermal patch: a guide to its use in chronic non-malignant pain', CNS Drugs, 26(4), 367-73.

Prins, A., Ouimette, P., Kimerling, R., Camerond, R. P., Hugelshofer, D. S., Shaw- Hegwer, J., Thrailkill, A., Gusman, F. D., and Sheikh, J. I. (2004). 'The primary care PTSD screen (PC-PTSD): development and operating characteristics', Primary Care Psychiatry, 9(1), 9-14.

Racine, M., Choiniere, M., and Nielson, W. R. (2014). 'Predictors of suicidal ideation in chronic pain patients: an exploratory study', Clinical Journal of Pain, 30(5), 371- 8.

Ratcliffe, G. E., Enns, M. W., Belik, S.-L., and Sareen, J. (2008). 'Chronic Pain Conditions and Suicidal Ideation and Suicide Attempts: An Epidemiologic Perspective', The Clinical Journal of Pain, 24(3), 204-10 10.1097/AJP.0b013e31815ca2a3.

Reid, K. J., Harker, J., Bala, M. M., Truyers, C., Kellen, E., Bekkering, G. E., and Kleijnen, J. (2011). 'Epidemiology of chronic non-cancer pain in Europe: narrative review of prevalence, pain treatments and pain impact', Current Medical Research & Opinion, 27(2), 449-62.

Reid, M. C., Engles-Horton, L. L., Weber, M. B., Kerns, R. D., Rogers, E. L., O'Connor, P. G., Reid, M. C., Engles-Horton, L. L., Weber, M. B., Kerns, R. D., Rogers, E. L., and O'Connor, P. G. (2002). 'Use of opioid medications for chronic noncancer pain syndromes in primary care', Journal of General Internal Medicine, 17(3), 173-9.

Reid, M. C., Henderson, C. R., Jr., Papaleontiou, M., Amanfo, L., Olkhovskaya, Y., Moore, A. A., Parikh, S. S., and Turner, B. J. (2010). 'Characteristics of older adults receiving opioids in primary care: treatment duration and outcomes', Pain Med, 11(7), 1063-71.

228

Chapter ten: References

Rejas, J., Ribera, M. V., Ruiz, M., and Masrramón, X. (2007). 'Psychometric properties of the MOS (Medical Outcomes Study) Sleep Scale in patients with neuropathic pain', European Journal of Pain, 11(3), 329-40.

Rogers, K. D., Kemp, A., McLachlan, A. J., and Blyth, F. (2013). 'Adverse Selection? A Multi-Dimensional Profile of People Dispensed Opioid Analgesics for Persistent Non-Cancer Pain', PloS one, 8(12), e80095.

Rosenblum, A., Marsch, L. A., Joseph, H., and Portenoy, R. K. (2008). 'Opioids and the treatment of chronic pain: controversies, current status, and future directions', Experimental and Clinical Psychopharmacology, 16(5), 405.

Rowbotham, M. C. (2006). 'Pharmacologic Management of Complex Regional Pain Syndrome', The Clinical Journal of Pain, 22(5), 425-29 10.1097/01.ajp.0000194281.74379.01.

Roxburgh, A., Bruno, R., Larance, B., and Burns, L. (2011). 'Prescription of opioid analgesics and related harms in Australia', Medical Journal of Australia, 195(5), 280-84.

Royal Australasian College of Physicians. 2008. "Prescription Opioid Policy: Improving management of chronic non-malignant pain and prevention of problems associated with prescription opioid use." In. Sydney: Royal Australasian College of Physicians.

Rust, K., and Rao, J. (1996). 'Variance estimation for complex surveys using replication techniques', Statistical Methods in Medical Research, 5(3), 283-310.

Rustoen, T., Wahl, A. K., Hanestad, B. R., Lerdal, A., Paul, S., and Miaskowski, C. (2005). 'Age and the experience of chronic pain: differences in health and quality of life among younger, middle-aged, and older adults', Clinical Journal of Pain, 21(6), 513-23.

Saastamoinen, P., Leino-Arjas, P., Laaksonen, M., and Lahelma, E. (2005). 'Socio- economic differences in the prevalence of acute, chronic and disabling chronic pain among ageing employees', Pain, 114(3), 364-71.

Sanderson K, and G., A. (2002). 'The SF-12 in the Australian population: cross- validation of item selection.', Australian & New Zealand Journal of Public Health, 26, 343-45.

Sansone, R. A., and Sansone, L. A. (2007). 'Borderline personality and the pain paradox', Psychiatry (Edgmont), 4(4), 40.

———. (2012). 'Chronic pain syndromes and borderline personality', Innovations in clinical neuroscience, 9(1), 10.

229

Chapter ten: References

Sansone, R. A., Watts, D. A., and Wiederman, M. W. (2014). 'Pain, pain catastrophizing, and history of intentional overdoses and attempted suicide', Pain Pract, 14(2), 30.

Sansone, R. A., Whitecar, P., Meier, B. P., and Murry, A. (2001). 'The prevalence of borderline personality among primary care patients with chronic pain', General Hospital Psychiatry, 23(4), 193-97.

Sansone, R. A., Whitecar, P., and Wiederman, M. W. (2009). 'The Prevalence of Childhood Trauma Among Those Seeking Buprenorphine Treatment', Journal of Addictive Diseases, 28(1), 64-67.

Savage, S., Covington, E., Heit, H., Hunt, J., Joranson, D., and Schnoll, S. (2001). 'Definitions related to the use of opioids for the treatment of pain: A consensus document from the American Academy of Pain Medicine, the American Pain Society, and the American Society of Addiction Medicine', American Pain Society Advocacy and Policy.

Savage, S. R. (1999). 'Opioid therapy of chronic pain: Assessment of consequences', Acta Anaesthesiologica Scandinavica, 43(9), 909-17.

Scascighini, L., Toma, V., Dober-Spielmann, S., and Sprott, H. (2008). 'Multidisciplinary treatment for chronic pain: a systematic review of interventions and outcomes', Rheumatology, 47(5), 670-78.

Schatman, M. (2012). 'Interdisciplinary chronic pain management: international perspectives', Pain: Clinical Updates, 20(7), 1-5.

Schneider, B. (2009). 'Substance use disorders and risk for completed suicide', Arch Suicide Res, 13(4), 303-16.

Scott, K. M., Hwang, I., Chiu, W.-T., Kessler, R. C., Sampson, N. A., Angermeyer, M., Beautrais, A., Borges, G., Bruffaerts, R., de Graaf, R., Florescu, S., Fukao, A., Maria Haro, J., Hu, C., Kovess, V., Levinson, D., Posada-Villa, J., Scocco, P., and Nock, M. K. (2010). 'Chronic Physical Conditions and Their Association With First Onset of Suicidal Behavior in the World Mental Health Surveys', Psychosom Med, 72(7), 712-19.

Sees, K. L., and Clark, H. W. (1993). 'Opioid use in the treatment of chronic pain: Assessment of addiction', Journal of Pain and Symptom Management, 8(5), 257-64.

Senate Community Affairs Committee Secretaria. 2010. "The Hidden Toll: Suicide in Australia." In, edited by Senate Community Affairs Committee Secretariat. Canberra: Senate Printing Unit.

Shand, F. L., Campbell, G., Hall, W., Lintzeris, N., Cohen, M., and Degenhardt, L. (2013). 'Real-time monitoring of Schedule 8 medicines in Australia: evaluation is essential', The Medical journal of Australia, 198(2), 80. 230

Chapter ten: References

Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T., Baker, R., and Dunbar, G. C. (1998). 'The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10', J Clin Psychiatry, 59 Suppl 20, 22-33;quiz 34-57.

Sherbourne, C. D., and Stewart, A. 1993. "The MOS Social Support Survey." In, edited by RAND Corporation. Santa Monica, CA: RAND Corporation.

Silverman, M. M., Berman, A. L., Sanddal, N. D., O'Carroll P, W., and Joiner, T. E. (2007). 'Rebuilding the tower of Babel: a revised nomenclature for the study of suicide and suicidal behaviors. Part 2: Suicide-related ideations, communications, and behaviors', Suicide Life Threat Behav, 37(3), 264-77.

Skevington, S. M., Lotfy, M., and O'Connell, K. A. (2004). 'The World Health Organization's WHOQOL-BREF quality of life assessment: Psychometric properties and results of the international field trial. A Report from the WHOQOL Group', Quality of Life Research, 13(2), 299-310.

Slade, T., Johnston, A., Oakley Browne, M. A., Andrews, G., and Whiteford, H. (2009). '2007 National Survey of Mental Health and Wellbeing: methods and key findings', Australian and New Zealand Journal of Psychiatry, 43(7), 594-605.

Smith, K., Mattick, R. P., Bruno, R., Nielsen, S., Cohen, M., Campbell, G., Larance, B., Farrell, M., and Degenhardt, L. (2015). 'Factors associated with the development of depression in chronic non-cancer pain patients following the onset of opioid treatment for pain', J Affect Disord, 184, 72-80.

Smith, M. T., Edwards, R. R., Robinson, R. C., and Dworkin, R. H. (2004). 'Suicidal ideation, plans, and attempts in chronic pain patients: factors associated with increased risk', Pain, 111(1–2), 201-08.

Smith, M. T., and Haythornthwaite, J. A. (2004). 'How do sleep disturbance and chronic pain inter-relate? Insights from the longitudinal and cognitive-behavioral clinical trials literature', Sleep Medicine Reviews, 8(2), 119-32.

Smith, S. M., Dart, R. C., Katz, N. P., Paillard, F., Adams, E. H., Comer, S. D., Degroot, A., Edwards, R. R., Haddox, J. D., Jaffe, J. H., Jones, C. M., Kleber, H. D., Kopecky, E. A., Markman, J. D., Montoya, I. D., O'Brien, C., Roland, C. L., Stanton, M., Strain, E. C., Vorsanger, G., Wasan, A. D., Weiss, R. D., Turk, D. C., and Dworkin, R. H. (2013). 'Classification and definition of misuse, abuse, and related events in clinical trials: ACTTION systematic review and recommendations', Pain, 154(11), 2287-96.

Smith, S. M., Paillard, F., McKeown, A., Burke, L. B., Edwards, R. R., Katz, N. P., Papadopoulos, E. J., Rappaport, B. A., Slagle, A., Strain, E. C., Wasan, A. D., Turk, D. C., and Dworkin, R. H. (2015). 'Instruments to Identify Prescription

231

Chapter ten: References

Medication Misuse, Abuse, and Related Events in Clinical Trials: An ACTTION Systematic Review', The Journal of Pain, 16(5), 389-411.

Somers, T. J., Kurakula, P. C., Criscione-Schreiber, L., Keefe, F. J., and Clowse, M. E. B. (2012). 'Self-efficacy and pain catastrophizing in systemic lupus erythematosus: Relationship to pain, stiffness, fatigue, and psychological distress', Arthritis Care & Research, 64(9), 1334-40.

Spiegel, B., Schoenfeld, P., and Naliboff, B. (2007). 'Systematic review: the prevalence of suicidal behaviour in patients with chronic abdominal pain and irritable bowel syndrome', Alimentary Pharmacology & Therapeutics, 26(2), 183-93.

Spitz, A., Moore, A. A., Papaleontiou, M., Granieri, E., Turner, B. J., and Reid, M. C. (2011). 'Primary care providers' perspective on prescribing opioids to older adults with chronic non-cancer pain: a qualitative study', BMC Geriatr, 11, 35.

Spitzer, R. L., Kroenke, K., Williams, J. B. W., and Lowe, B. (2006). 'A brief measure for assessing generalized anxiety disorder: the GAD-7', Archives of Internal Medicine, 166(10), 1092-7.

Spritzer, K. L., and Hays, R. D. 2003. "MOS Sleep Scale: A Manual for Use and Scoring, Version 1.0. ." In. Los Angeles, CA. Accessible at: http://www.rand.org/health/surveys_tools/mos/mos_sleep.html. Accessed on 02/04/14. .

Stewart, W. F., Ricci, J. A., Chee, E., Morganstein, D., and Lipton, R. (2003). 'Lost productive time and cost due to common pain conditions in the US workforce', JAMA, 290(18), 2443-54.

Stoffers, J. M., Völlm, B., Rücker, G., Timmer, A., Huband, N., and Lieb, K. (2012). 'Psychological therapies for people with borderline personality disorder', Cochrane Database of Systematic Reviews(8).

Stomski, N. J., Mackintosh, S., and Stanley, M. (2010). 'Patient self-report measures of chronic pain consultation measures: a systematic review', The Clinical Journal of Pain, 26(3), 235-43.

Strine, T. W., and Hootman, J. M. (2007). 'US national prevalence and correlates of low back and neck pain among adults', Arthritis Care & Research, 57(4), 656-65.

Suicide Prevention Australia. 2012. "Chronic Illness, Chronic Pain and Suicide Prevention Position Statement." In, edited by Suicide Prevention Australia. Sydney.

Sullivan, M., Von Korff, M., Banta-Green, C., Merrill, J., and Saunders, K. (2010). 'Problems and concerns of patients receiving chronic opioid therapy for chronic non-cancer pain. (19 refs.)', PAIN®, 149, 345-53.

232

Chapter ten: References

Sullivan, M. D. (2010). 'Who gets high-dose opioid therapy for chronic non-cancer pain?', PAIN®, 151(3), 567-68.

Sullivan, M. D., and Howe, C. Q. (2013). 'Opioid therapy for chronic pain in the United States: promises and perils', PAIN®, 154, S94-S100.

Svendsen, K., Borchgrevink, P., Fredheim, O., Hamunen, K., Mellbye, A., and Dale, O. (2011). 'Choosing the unit of measurement counts: the use of oral morphine equivalents in studies of opioid consumption is a useful addition to defined daily doses', Palliat Med, 25(7), 725-32.

Tan, G., Jensen, M. P., Thornby, J. I., and Shanti, B. F. (2004). 'Validation of the brief pain inventory for chronic nonmalignant pain', The Journal of Pain, 5(2), 133-37.

Tang, N. K. Y., and Crane, C. (2006). 'Suicidality in chronic pain: a review of the prevalence, risk factors and psychological links', Psychological Medicine, 36(5), 575-86. ten Have, M., de Graaf, R., Van Dorsselaer, S., Verdurmen, J., van't Land, H., Vollebergh, W., and Beekman, A. (2009). 'Incidence and course of suicidal ideation and suicide attempts in the general population', Canadian journal of psychiatry. Revue canadienne de psychiatrie, 54(12), 824-33.

Tepper, S. J. (2012). 'Opioids Should Not Be Used in Migraine', Headache: The Journal of Head and Face Pain, 52, 30-34.

Terrill, A. L., Molton, I. R., Ehde, D. M., Amtmann, D., Bombardier, C. H., Smith, A. E., and Jensen, M. P. (2014). 'Resilience, age, and perceived symptoms in persons with long-term physical disabilities', Journal of Health Psychology, 1359105314532973.

Therapeutic Goods Administration. 2013. "Standard for the Uniform Scheduling of Medicines and Poisons (SUSMP)." In http://www.tga.gov.au/industry/scheduling-poisons-standard.htm. Accessed on 05/09/2013, edited by Australian Government Department of Health and Ageing.

Therapeutic Guidelines Limited. 2013. 'eTG complete: Analgesic Guidelines . Melbourne'. http://etg.tg.com.au/complete/.

Timonen, M., Viilo, K., Hakko, H., Sarkioja, T., Ylikulju, M., Meyer-Rochow, V. B., Vaisanen, E., and Rasanen, P. (2003). 'Suicides in persons suffering from rheumatoid arthritis', Rheumatology, 42(2), 287-91.

Tragesser, S. L., Bruns, D., and Disorbio, J. M. (2010). 'Borderline Personality Disorder Features and Pain: The Mediating Role of Negative Affect in a Pain Patient Sample', The Clinical Journal of Pain, 26(4), 348-53.

233

Chapter ten: References

Trull, T. J., Sher, K. J., Minks-Brown, C., Durbin, J., and Burr, R. (2000). 'Borderline personality disorder and substance use disorders: A review and integration', Clinical psychology review, 20(2), 235-53.

Tsang, A., Von Korff, M., Lee, S., Alonso, J., Karam, E., Angermeyer, M. C., Borges, G. L. G., Bromet, E. J., de Girolamo, G., de Graaf, R., Gureje, O., Lepine, J.-P., Haro, J. M., Levinson, D., Oakley Browne, M. A., Posada-Villa, J., Seedat, S., and Watanabe, M. (2008). 'Common Chronic Pain Conditions in Developed and Developing Countries: Gender and Age Differences and Comorbidity With Depression-Anxiety Disorders', The Journal of Pain, 9(10), 883-91.

Tsirigotis, K., Gruszczynski, W., and Tsirigotis, M. (2011). 'Gender differentiation in methods of suicide attempts', Med Sci Monit, 17(8), PH65-70.

Turk, D. C., Dworkin, R. H., Allen, R. R., Bellamy, N., Brandenburg, N., Carr, D. B., Cleeland, C., Dionne, R., Farrar, J. T., Galer, B. S., Hewitt, D. J., Jadad, A. R., Katz, N. P., Kramer, L. D., Manning, D. C., McCormick, C. G., McDermott, M. P., McGrath, P., Quessy, S., Rappaport, B. A., Robinson, J. P., Royal, M. A., Simon, L., Stauffer, J. W., Stein, W., Tollett, J., and Witter, J. (2003). 'Core outcome domains for chronic pain clinical trials: IMMPACT recommendations', Pain, 106(3), 337-45.

Turk, D. C., Dworkin, R. H., Revicki, D., Harding, G., Burke, L. B., Cella, D., Cleeland, C. S., Cowan, P., Farrar, J. T., Hertz, S., Max, M. B., and Rappaport, B. A. (2008). 'Identifying important outcome domains for chronic pain clinical trials: An IMMPACT survey of people with pain', Pain, 137(2), 276-85.

Turk, D. C., Okifuji, A., and Scharff, L. (1995). 'Chronic pain and depression: role of perceived impact and perceived control in different age cohorts', Pain, 61(1), 93-101.

Van Orden, K. A., Witte, T. K., Cukrowicz, K. C., Braithwaite, S. R., Selby, E. A., and Joiner Jr, T. E. (2010). 'The interpersonal theory of suicide', Psychological review, 117(2), 575.

Vandenbroek, P. 2013. "Labour Stats 101 labour force: a quick guide." In, edited by Parliamentary Library. Canberra: Parliament of Australia.

Vos, T., Flaxman, A. D., Naghavi, M., Lozano, R., Michaud, C., Ezzati, M., Shibuya, K., Salomon, J. A., Abdalla, S., Aboyans, V., Abraham, J., Ackerman, I., Aggarwal, R., Ahn, S. Y., Ali, M. K., Alvarado, M., Anderson, H. R., Anderson, L. M., Andrews, K. G., Atkinson, C., Baddour, L. M., Bahalim, A. N., Barker-Collo, S., Barrero, L. H., Bartels, D. H., Basanez, M. G., Baxter, A., Bell, M. L., Benjamin, E. J., Bennett, D., Bernabe, E., Bhalla, K., Bhandari, B., Bikbov, B., Bin Abdulhak, A., Birbeck, G., Black, J. A., Blencowe, H., Blore, J. D., Blyth, F., Bolliger, I., Bonaventure, A., Boufous, S. A., Bourne, R., Boussinesq, M., Braithwaite, T., Brayne, C., Bridgett, L., Brooker, S., Brooks, P., Brugha, T. S., Bryan-Hancock, C., Bucello, C., Buchbinder, R., Buckle, G. R., Budke, C. M., Burch, M., Burney, P., Burstein, R.,

234

Chapter ten: References

Calabria, B., Campbell, B., Canter, C. E., Carabin, H., Carapetis, J., Carmona, L., Cella, C., Charlson, F., Chen, H. L., Cheng, A. T. A., Chou, D., Chugh, S. S., Coffeng, L. E., Colan, S. D., Colquhoun, S., Colson, K. E., Condon, J., Connor, M. D., Cooper, L. T., Corriere, M., Cortinovis, M., de Vaccaro, K. C., Couser, W., Cowie, B. C., Criqui, M. H., Cross, M., Dabhadkar, K. C., Dahiya, M., Dahodwala, N., Damsere-Derry, J., Danaei, G., Davis, A., De Leo, D., Degenhardt, L., Dellavalle, R., Delossantos, A., Denenberg, J., Derrett, S., Des Jarlais, D. C., Dharmaratne, S. D., Dherani, M., Diaz-Torne, C., Dolk, H., Dorsey, E. R., Driscoll, T., Duber, H., Ebel, B., Edmond, K., Elbaz, A., Ali, S. E., Erskine, H., Erwin, P. J., Espindola, P., Ewoigbokhan, S. E., Farzadfar, F., Feigin, V., Felson, D. T., Ferrari, A., Ferri, C. P., Fevre, E. M., Finucane, M. M., Flaxman, S., Flood, L., Foreman, K., Forouzanfar, M. H., Fowkes, F. G. R., Franklin, R., Fransen, M., Freeman, M. K., Gabbe, B. J., Gabriel, S. E., Gakidou, E., Ganatra, H. A., Garcia, B., Gaspari, F., Gillum, R. F., Gmel, G., Gosselin, R., Grainger, R., Groeger, J., Guillemin, F., Gunnell, D., Gupta, R., Haagsma, J., Hagan, H., Halasa, Y. A., Hall, W., Haring, D., Haro, J. M., Harrison, J. E., Havmoeller, R., Hay, R. J., Higashi, H., Hill, C., Hoen, B., Hoffman, H., Hotez, P. J., Hoy, D., Huang, J. J., Ibeanusi, S. E., Jacobsen, K. H., James, S. L., Jarvis, D., Jasrasaria, R., Jayaraman, S., Johns, N., Jonas, J. B., Karthikeyan, G., Kassebaum, N., Kawakami, N., Keren, A., Khoo, J. P., King, C. H., Knowlton, L. M., Kobusingye, O., Koranteng, A., Krishnamurthi, R., Lalloo, R., Laslett, L. L., Lathlean, T., Leasher, J. L., Lee, Y. Y., Leigh, J., Lim, S. S., Limb, E., Lin, J. K., Lipnick, M., Lipshultz, S. E., Liu, W., Loane, M., Ohno, S. L., Lyons, R., Ma, J. X., Mabweijano, J., MacIntyre, M. F., Malekzadeh, R., Mallinger, L., Manivannan, S., Marcenes, W., March, L., Margolis, D. J., Marks, G. B., Marks, R., Matsumori, A., Matzopoulos, R., Mayosi, B. M., McAnulty, J. H., McDermott, M. M., McGill, N., McGrath, J., Medina-Mora, M. E., Meltzer, M., Mensah, G. A., Merriman, T. R., Meyer, A. C., Miglioli, V., Miller, M., Miller, T. R., Mitchell, P. B., Mocumbi, A. O., Moffitt, T. E., Mokdad, A. A., Monasta, L., Montico, M., Moradi-Lakeh, M., Moran, A., Morawska, L., Mori, R., Murdoch, M. E., Mwaniki, M. K., Naidoo, K., Nair, M. N., Naldi, L., Narayan, K. M. V., Nelson, P. K., Nelson, R. G., Nevitt, M. C., Newton, C. R., Nolte, S., Norman, P., Norman, R., O'Donnell, M., O'Hanlon, S., Olives, C., Omer, S. B., Ortblad, K., Osborne, R., Ozgediz, D., Page, A., Pahari, B., Pandian, J. D., Rivero, A. P., Patten, S. B., Pearce, N., Padilla, R. P., Perez-Ruiz, F., Perico, N., Pesudovs, K., Phillips, D., Phillips, M. R., Pierce, K., Pion, S., Polanczyk, G. V., Polinder, S., Pope, C. A., Popova, S., Porrini, E., Pourmalek, F., Prince, M., Pullan, R. L., Ramaiah, K. D., Ranganathan, D., Razavi, H., Regan, M., Rehm, J. T., Rein, D. B., Remuzzi, G., Richardson, K., Rivara, F. P., Roberts, T., Robinson, C., De Leon, F. R., Ronfani, L., Room, R., Rosenfeld, L. C., Rushton, L., Sacco, R. L., Saha, S., Sampson, U., Sanchez-Riera, L., Sanman, E., Schwebel, D. C., Scott, J. G., Segui-Gomez, M., Shahraz, S., Shepard, D. S., Shin, H., Shivakoti, R., Singh, D., Singh, G. M., Singh, J. A., Singleton, J., Sleet, D. A., Sliwa, K., Smith, E., Smith, J. L., Stapelberg, N. J. C., Steer, A., Steiner, T., Stolk, W. A., Stovner, L. J., Sudfeld, C., Syed, S., Tamburlini, G., Tavakkoli, M., Taylor, H. R., Taylor, J. A., Taylor, W. J., Thomas, B., Thomson, W. M., Thurston, G. D., Tleyjeh, I. M., Tonelli, M., Towbin, J. R. A., Truelsen, T., Tsilimbaris, M. K., Ubeda, C., Undurraga, E. A., van der Werf, M. J., van Os, J., Vavilala, M. S., Venketasubramanian, N., Wang, M. R., Wang, W. Z., Watt, K., Weatherall, D. J., 235

Chapter ten: References

Weinstock, M. A., Weintraub, R., Weisskopf, M. G., Weissman, M. M., White, R. A., Whiteford, H., Wiersma, S. T., Wilkinson, J. D., Williams, H. C., Williams, S. R. M., Witt, E., Wolfe, F., Woolf, A. D., Wulf, S., Yeh, P. H., Zaidi, A. K. M., Zheng, Z. J., Zonies, D., Lopez, A. D., and Murray, C. J. L. (2012). 'Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010', Lancet, 380(9859), 2163- 96.

Vowles, K. E., McEntee, M. L., Julnes, P. S., Frohe, T., Ney, J. P., and van der Goes, D. N. (2015). 'Rates of opioid misuse, abuse, and addiction in chronic pain: a systematic review and data synthesis', Pain, 156(4), 569-76.

Wallston, K. A., Wallston, B. S., and DeVellis, R. (1978). 'Development of the multidimensional health locus of control (MHLC) scales', Health Education Monographs, 6, 160-70.

Webster, L. R., and Webster, R. M. (2005). 'Predicting Aberrant Behaviors in Opioid‐ Treated Patients: Preliminary Validation of the Opioid Risk Tool', Pain Medicine, 6(6), 432-42.

Weissman, M. M., Bland, R. C., Canino, G. J., Greenwald, S., Hwu, H. G., Joyce, P. R., Karam, E. G., Lee, C. K., Lellouch, J., Lepine, J. P., Newman, S. C., Rubio-Stipec, M., Wells, J. E., Wickramaratne, P. J., Wittchen, H. U., and Yeh, E. K. (1999). 'Prevalence of suicide ideation and suicide attempts in nine countries', Psychol Med, 29(1), 9-17.

Wilsey, B. L., Fishman, S. M., Tsodikov, A., Ogden, C., Symreng, I., and Ernst, A. (2008). 'Psychological Comorbidities Predicting Prescription Opioid Abuse among Patients in Chronic Pain Presenting to the Emergency Department', Pain Medicine, 9(8), 1107-17.

Wilson, K. G., Kowal, J., Henderson, P. R., McWilliams, L. A., and Péloquin, K. (2013). 'Brief Report: Chronic Pain and the Interpersonal Theory of Suicide', Rehabilitation Psychology, 58(1), 111-15.

Wittink, H. M., Rogers, W. H., Lipman, A. G., McCarberg, B. H., Ashburn, M. A., Oderda, G. M., and Carr, D. B. (2006). 'Older and Younger Adults in Pain Management Programs in the United States: Differences and Similarities', Pain Medicine, 7(2), 151-63.

Wolfe, F., Hassett, A. L., Walitt, B., and Michaud, K. (2011). 'Mortality in fibromyalgia: a study of 8,186 patients over thirty-five years', Arthritis Care Res, 63(1), 94-101.

Wolff, R. F., Reid, K., di Nisio, M., Aune, D., Truyers, C., Hernandez, A. V., Misso, K., Riemsma, R., and Kleijnen, J. (2012). 'Systematic review of adverse events of buprenorphine patch versus fentanyl patch in patients with chronic moderate- to-severe pain', Pain Manag, 2(4), 351-62.

236

Chapter ten: References

World Health Organization. 1998. "Obesity: preventing and managing the global epidemic." In. Geneva: World Health Organization.

———. 2001. "Composite International Diagnostic Interview, Version 3.0." In. Geneva: World Health Organization.

———. 2014. "Preventing Preventing suicide: A global imperative." In, edited by World Health Organization. Luxemburgh

Zedler, B., Xie, L., Wang, L., Joyce, A., Vick, C., Kariburyo, F., Rajan, P., Baser, O., and Murrelle, L. (2014). 'Risk Factors for Serious Prescription Opioid‐Related Toxicity or Overdose among Veterans Health Administration Patients', Pain Medicine, 15(11), 1911-29.

237

Chapter eleven: Appendices

11 APPENDICES

Appendix A: Pain conditions by age-groups

Weighted% < 25 years 25-44 years 45-64 years 65+ years n=1471 n=2928 n=2537 n=1905 % 95%CI % 95%CI % 95%CI % 95%CI Arthritis 0.7 0.4-1.0 4.9 3.7-6.5 22.3 20.0-24.8 42.7 40.2-45.1 Migraines 5.9 4.7-7.6 8.8 7.6-10.2 9.6 8.1-11.4 6.4 5.3-7.7 Back/neck 11.4 9.7-13.4 24.3 22.2-26.5 37.5 34.6-40.6 34.3 31.9-36.7 problems Any pain 15.5 13.5-17.7 30.8 28.5-33.2 50.1 47.0-53.1 59.4 56.9-61.8

238

Chapter eleven: Appendices

Appendix B: Prevalence of suicidality by pain condition Arthritis Migraines Back/neck problems n=1,577 n=783 n=2,487 % 95% CI % 95% CI % 95% CI Ever thought 19.8 16.8-22.8 16.2 13.5-18.9 45.3 41.5-49.1 suicide Thought suicide 18.4 10.9-26.0 18.2 12.1-24.2 48.9 40.4-57.4 12m Ever plan suicide 23.2 17.7-28.9 16.8 12.5-21.0 48.3 41.5-55.2 Plan suicide 12m 17.6 6.4-28.8 15.2 6.6-23.8 54.3 39.9-68.8 Ever attempt 22.3 15.6-29.1 18.6 13.3-23.8 49.1 41.9-56.3 suicide Attempt suicide 13.1 10.1-26.2 20.6 3.3-37.9 52.1 32.9-71.2 12m

239

Appendix C: Prevalence of chronic pain amongst people who had experienced lifetime and past 12-month suicidality, by age- groups

< 25 years 25-44 years 45-64 years 65+ years n % 95%CI n % 95%CI n % 95%CI n % 95%CI Lifetime… suicidal thoughts 60 36.8 28.7-45.7 229 51.5 45.4-57.5 306 66.7 60.7-72.3 105 69.4 60.6-77.0 suicide plan 31 51.0 37.6-64.2 78 55.8 43.9-67.0 102 67.4 56.0-77.0 28 80.0 63.8-90.0 suicide attempt 28 52.8 38.4-66.6 75 57.7 46.4-68.3 82 71.0 58.5-80.7 19 76.9 58.1-88.8 Past 12-month… suicidal thoughts 18 42.3 27.4-58.8 48 47.1 34.5-60.0 57 83.1 69.6-91.4 18 75.1 53.7-88.7 suicide plan 10 71.2 4.34-88.9 15 51.9 31.5-71.6 14 83.3 50.2-96.1 3 64.5 20.7-92.7 suicide attempt 10 61.4 36.1-81.8 8 58.0 27.7-83.3 4 90.0 45.2-99.0 2 100 na

240

Appendix D: Factors associated with past 12-month suicidal behaviours in people with chronic pain1

Any suicidal behaviour n=239 Demographics OR 95% CI p value Age over 60 0.41 0.20-0.83 0.014 Gender 1.38 0.83-2.32 0.214 Completed school 0.87 0.52-1.44 0.591 Employed 0.58 0.33-1.01 0.057 Married 0.39 0.21-0.70 0.002 Mental health Past 12m depression 3.51 1.91-6.45 <0.001 Past 12m GAD 2.11 1.07-4.17 0.031 Past 12m PTSD 2.52 1.36-4.67 0.003 Past 12m any alcohol disorder 2.53* 1.13-5.63 0.023 Past 12m any drug disorders 2.00 0.77-5.24 0.155 Physical health Poor/Fair self-reported health 2.57 1.54-4.29 <0.001 1. Regressions were run with the removal of age and gender to determine if there was an effect for controlling for age and gender in the regressions and in the survey weightings. There were no differences. Bold is significant

241

Appendix E - Glossary of conditions that may lead to chronic pain

CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Achalasia Throat/Chest Pain The esophagus has problems moving properly, causing difficulty swallowing, regurgitation, and pain. No Visceral known cause; can be treated but not cured (Park W, 2005) (Lake & Wong, 2006). Alcohol dependence Peripheral Pain Alcohol dependence can contribute to peripheral neuropathy - pain in extremities such as hands and feet Neuropathic (Muller, Koch, Von Specht, Volker, & Munch, 1985). Ankylosing spondylitis (aka Back/Joint Pain Arthritis of the spine. It causes swelling between your vertebrae, which are the disks that make up your Deep Somatic Bechterew’s disease, spine, and in the joints between your spine and pelvis. Ankylosing spondylitis is an autoimmune disease. Bechterew syndrome and This means your immune system, which normally protects your body from infection, attacks your body's Marie Struempell disease) own tissues. The disease is more common and more severe in men. It often runs in families (American Chronic Pain Association, 2012) (London Pain Clinic, 2006) . Arachnoiditis Back Pain A pain disorder caused by the inflammation of the arachnoid, one of the membranes that surround and Neuropathic protect the nerves of the spinal cord. Inflammation can sometimes lead to the formation of scar tissue and adhesions, which cause the spinal nerves to “stick” together. If arachnoiditis begins to interfere with the function of one or more of these nerves, it can cause a number of symptoms, including numbness, tingling, and a characteristic stinging and burning pain in the lower back or legs (American Chronic Pain Association, 2012). Bursitis Joint Pain Bursitis is a painful condition that affects the small fluid-filled pads — called bursae — that act as cushions Deep Somatic among your bones and the tendons and muscles near your joints. Bursitis occurs when a bursa becomes inflamed. The most common locations for bursitis are in the shoulders, elbows or hips. Bursitis often occurs in joints that perform frequent repetitive motion (American Chronic Pain Association, 2012). Carpal Tunnel Syndrome Peripheral Pain Swelling in the wrist tunnel irritates the median nerve. CTS causes tingling, numbness and pain over the Neuropathic thumb, first and middle fingers (Jacques, 2009). Central Pain Syndrome Generalised Pain Neurological condition caused by damage to or dysfunction of the central nervous system (CNS), which Neuropathic includes the brain, brainstem, and spinal cord. This syndrome can be caused by: stroke, multiple sclerosis, tumors, epilepsy, brain or spinal cord trauma, or Parkinson's disease. Central pain syndrome may affect a large portion of the body or may be more restricted to specific areas, such as hands or feet. Pain is typically constant, may be moderate to severe in intensity, and is often made worse by touch, movement, emotions, and temperature changes. There are several common pain sensations: burning, "pins and needles;" pressing, lacerating, or aching pain; and brief, intolerable bursts of sharp pain akin to a dental probe on an exposed nerve. Central pain syndrome often begins shortly after a causative injury/damage, but may be 242

delayed by months/years (American Chronic Pain Association, 2012) CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Cervical Dystonia/Spasmodic Neck Pain Cervical dystonia, also called spasmodic torticollis, is a painful condition in which your neck muscles Deep somatic Torticollis contract involuntarily, causing your head to twist or turn to one side. A rare disorder that can occur at any age, cervical dystonia most often occurs in middle-aged people (Jankovic J. , 2006) Chronic Fatigue Syndrome Generalised Pain An illness without a known cause that is characterized by long-term exhaustion, muscle weakness, Deep Somatic depression, pain and sleep disturbance; widespread muscle and joint pain, along with frequent headaches, are commonly reported (Blatman Pain Clinic). Chronic Pancreatis Abdominal/Pelvic Pain Continuing, chronic, inflammatation of the pancreas, leading to chronic abdominal pain and/or impairment Visceral of endocrine and exocrine function of the pancreas. Chronic pancreatitis is a completely different inflammation process from acute pancreatitis. acute pancreatis is fully curable but chronic pancreatitis is a chronic, irreversible inflammation leading to fibrosis with calcification (American Chronic Pain Association, 2012). Cluster Headaches Headaches Often confused with migraines, these severe headaches are usually caused by enlarged blood vessels Deep somatic leading into the head (Blatman Pain Clinic). Complex Regional Pain Generalised Pain A chronic neurological syndrome characterized by: severe burning pain; pathological changes in bone and Neuropathic Syndrome skin; excessive sweating; tissue swelling; extreme sensitivity to light & touch (American Chronic Pain Association, 2012). Compression Fractures Back Pain Commonly associated with osteoporosis, compression fractures occur when brittle vertebral bones collapse Deep somatic (Jacques, 2009). Crohn’s Disease Abdominal/Pelvic Pain Inflammation of the digestive system, causing abdominal/pelvic pain. It is one of a group of diseases called Visceral inflammatory bowel disease (American Chronic Pain Association, 2012) Cystic Fibrosis Generalised Pain Cystic Fibrosis (CF) is a genetic disorder that affects the lungs and other organs; it is characterised by Combination scarring and cyst formations in the pancreas. Pain is a common symptom; about 60% of adults with Cystic Fibrosis report negative effects of pain on their daily life. Recurrent lung and sinus infections, surgery, pneumothorax, and intestinal secretions can lead to significant pain. Bone and joint pain is associated with osteomalacia and osteoporosis which result from malabsorption of essential minerals and vitamins (American Chronic Pain Association, 2012) (Broome & Subramaniam, 2003) (Filippo Festini, 2004). Dercum’s Disease Generalised Pain Extremely rare disorder characterized by multiple, severely painful growths consisting of fatty tissue Superficial (lipomas). These growths mainly occur on the trunk, the upper arms and upper legs and are found just Somatic below the skin. Pain may be caused by these growths pressing on nearby nerves.The exact cause of Dercum's disease is unknown (American Chronic Pain Association, 2012). 243

CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Diabetes (Diabetic Peripheral Pain Sensory nerve damage is a common side effect of diabetes. It can cause numbness or pain, most often in Neuropathic Neuropathy) the hands or feet (Jacques, 2009). Dystonia Generalised Pain A disorder where muscles involuntarily contract, causing an uncontrollable and often painful twisting and Deep Somatic cramping of the affected body part. Symptoms can be mild or severe and can affect the entire body or one area. Most cases of dystonia, however, tend to affect only one body part — often the neck, the face or an arm (American Chronic Pain Association, 2012) (Jankovic J. , 2006). Endrometriosis Abdominal/Pelvic Pain Tissue which lines the uterus (tissue called the endometrium) is found outside the uterus. This misplaced Visceral tissue develops into growths or lesions and may cause scarring. Less commonly they are found in the lung, arm, thigh, and other locations. Pelvic pain is a common symptom (Office of Communications and Public Liaison, 2012). Erythromelalgia (Mitchell's Peripheral Pain Episodes of pain, redness, and swelling in various parts of the body, particularly the hands and feet. These Neuropathic Disease) episodes are usually triggered by increased body temperature. The pain can be so debilitating that it impedes everyday activities such as wearing shoes and walking (American Chronic Pain Association, 2012). Eye Strain Headaches Headaches Ocular muscles become fatigued and cause head pain. This is usually caused by sitting at a computer for too Deep somatic long, or wearing the wrong eyeglass prescription (Jacques, 2009). Failed Back Surgery syndrome Back Pain A very generalized, non-medical term that is often used to describe the chronic pain condition of patients Combination (FBSS) who have not had a successful result with back surgery or spine surgery (American Chronic Pain Association, 2012). Fibromyalgia Generalised Pain Widespread muscle fatigue and pain, often accompanied by chronic fatigue, sleep disorders and irritable Neuropathic bowel syndrome. Fibromyalgia is medically unexplained but thought to be neuropathic in nature (Jacques, 2009). Frozen Shoulder (aka adhesive Joint Pain The connective tissue surrounding the shoulder joint becomes inflamed, causing chronic pain. Pain is Deep Somatic capsulitis) usually constant, worse at night, and when the weather is colder. The exact cause is unknown, and the condition can last for years (American Academy of Orthopedic Surgeons, 2011) (Ewald, 2011). Gout Joint Pain Caused by deposition of uric acid crystals in the joint (usually in the foot), causing inflammation and pain Deep Somatic (American Chronic Pain Association, 2012). Hidradenitis suppurativa (fox- Generalised Pain Chronic skin disease that results in extremely painful open wounds that do not heal. Treatment is difficult Superficial den disease) (New Zealand Dermatological Society Inc, 2012). Somatic Inflammatory Bowel Disease Abdominal/Pelvic Pain A chronic condition where the gastrointestinal tract, intestine or colon becomes inflammed, resulting in Visceral abdominal pain and cramping (see also Crohn's disease) (Office of Communications and Public Liaison, 2012). 244

CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Interstitial Cystisis (Painful Abdominal/Pelvic Pain A chronic condition characterized by a combination of uncomfortable bladder pressure, bladder pain and Visceral bladder syndrome) sometimes pain in the pelvis, which can range from mild burning or discomfort to severe pain. The severity of symptoms caused by interstitial cystitis often fluctuates, and some people may experience periods of remission (Office of Communications and Public Liaison, 2012) (American Chronic Pain Association, 2012). Irritable Bowel Syndrome Abdominal/Pelvic Pain IBS affects the colon, or large bowel. IBS is not a disease; it's a functional disorder, meaning that the bowel Visceral doesn't function correctly. The most common symptoms of IBS are abdominal pain or discomfort often reported as cramping, bloating, gas, diarrhoea, and/or constipation (American Chronic Pain Association, 2012). Knee Injury Knee Pain Knee Injury: Your knee joint is made up of bone, cartilage, ligaments and fluid. Muscles and tendons help Deep Somatic the knee joint move. When any of these structures is hurt or diseased, you have knee problems. Knee problems can cause pain and difficulty walking (American Chronic Pain Association, 2012). Loin Pain-Haematuria Abdominal/Pelvic Pain This syndrome is a combination of kidney pain and blood in the urine. The pain may be a continuous dull Visceral Syndrome (also known as ache, or intermittent, coming on only occasionally (American Chronic Pain Association, 2012). LPHS) Lupus Generalised Pain A chronic, autoimmune disease that can damage any part of the body (skin, joints, and/or organs inside the Combination body). The immune system malfunctions in lupus, creating antibodies that attack and destroy healthy tissue. These antibodies cause inflammation, pain, and damage in various parts of the body (American Chronic Pain Association, 2012). Meralgia Paresthetica Peripheral Pain A condition characterized by tingling, numbness and burning pain in the outer part of the thigh, caused by Neuropathic (Bernhardt-Roth Syndrome) either compression of the nerve that supplies sensation to the skin surface of the thigh or damage from a condition like diabetes or an injury (American Chronic Pain Association, 2012). Migraines Headaches Migraines can be caused by nervous system triggers or hormonal changes in the body. They often cause Deep somatic pain on one side of the head or face, and may be accompanied by nausea and sensitivities to light, sounds or smells (Jacques, 2009) (Blatman Pain Clinic). Multiple Sclerosis Generalised Pain Multiple Sclerosis (MS) is an autoimmune disease that affects the central nervous system, attacking and Neuropathic damaging the nerves in the body. Pain is a common symptom in those with Multiple Sclerosis. Chronic pain is experienced by 64-69% of people with MS (Jacques, 2009) (Office of Communications and Public Liaison, 2012) (MS Australia, 2009). Muscle Tension Headache Headaches Often caused by stress, fatigue or “sleeping wrong,” muscles of the neck, shoulders and scalp tighten. This Deep somatic causes pressure on the head, leading to pain (Jacques, 2009).

245

CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Myofascial Pain Syndromes Generalised Pain Myofascial pain syndromes cause weakness, restriction of motion, acute pain, and chronic pain; it comes Deep somatic from myofascial trigger points that form in muscle tissue and connective tissue. More active trigger points are tender and generate referral pain patterns when they are pushed on. They cause varying types of pain that include cramping, burning, aching, numbness and tingling. Less active myofascial trigger points can predispose a person to acute pain attacks and cause: muscle weakness, early fatigue and restriction of motion, with or without pain (i.e. frozen shoulder, tight hamstrings) (Blatman Pain Clinic). Occipital Neuralgia Headaches A distinct type of headache characterized by piercing, throbbing, or electric-shock-like chronic pain in the Neuropathic upper neck, back of the head, and behind the ears, usually on one side of the head. Typically, the pain of occipital neuralgia begins in the neck and then spreads upwards. Some individuals will also experience pain in the scalp, forehead, and behind the eyes. Their scalp may also be tender to the touch, and their eyes especially sensitive to light (American Chronic Pain Association, 2012). Osteoarthritis Joint Pain Wear and tear on joints over time. It is common in the elderly, and usually affects one or more of the larger Deep somatic joints in the body (Jacques, 2009). Osteoporosis Generalised Pain Osteoporosis is a disease characterized by low bone mass and deterioration of the bone tissue. This process Combination of bone thinning leads to an increase in bone fragility and the consequent risk of fracture. The hip, spine and wrists are the most common areas of osteoporosis-related bone fractures, though these can actually occur almost anywhere in the skeleton. Osteoporosis runs a major risk of passing by undiagnosed as it can progress painlessly until a bone actually breaks. In fact, due to this, osteoporosis is often referred to as the ‘silent thief’, lacking any obvious signs of onset. Broken bones can cause chronic pain While a hip fracture usually requires hospitalization and major surgery, spinal or vertebral fractures also have serious consequences such as loss of height, severe back pain and deformity (Office of Communications and Public Liaison, 2012). Ovarian cysts Abdominal/Pelvic Pain Fluid-filled growths on the ovaries. They can recur and cause chronic pelvic pain (Medicine Online, 2004). Visceral Paget’s disease of bone Generalised Pain Bones grow larger and weaker than normal, causing deep bone pain. They also might break easily. The Deep Somatic disease can lead to other health problems, too, such as arthritis and hearing loss. The disease is most common in the spine, pelvis, skull and legs. The disease might affect one or several bones. It is most common in older people (American Chronic Pain Association, 2012). Parkinson’s disease Generalised Pain Degenerative disorder of the central nervous system. Joint pain is commonly associated with this disease; Combination the disease can also impair the senses, including an increased or decreased sensation of pain (Jankovic J. , 2008).

246

CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Phantom Limb Pain Peripheral Pain Pain that is felt in an amputated limb. . Phantom limb pain can be mild to extremely painful. In some cases, Neuropathic phantom limb pain can be disabling and can lead to a lifelong struggle with chronic pain. When phantom limb pain continues for more than six months, the prognosis for spontaneous improvement is poor (Blatman Pain Clinic) (American Chronic Pain Association, 2012). Polymyalgia rheumatica Generalised Pain An inflammatory disorder that causes muscle pain and stiffness, primarily in the neck, shoulders, upper Deep Somatic arms, hips and thighs. Symptoms usually begin quickly over a few days. More common in older people. Polymyalgia rheumatica is related to and may coexist with another inflammatory disorder called giant cell arteritis, which can cause headaches, visual impairment, jaw pain and other symptoms (American Chronic Pain Association, 2012). Post-Mastectomy Pain Peripheral Pain A type of chronic postoperative pain that persists after a mastectomy or other type of breast surgery is Neuropathic Syndrome performed. Post-mastectomy pain is often neuropathic; may be caused by damage to the nerves in the breast and underarm area, or the development of a neuroma (American Chronic Pain Association, 2012). Refractory angina or Generalised Pain Refractory angina or Intractable angina is a condition in which patients of heart disease continue to suffer Combination Intractable angina from recurrent restricting angina, even though they are following the right medication plan. Refractory angina is a chronic and incapacitating condition and often responds poorly to treatment (American Chronic Pain Association, 2012). Repetitive Strain Injury Peripheral Pain A painful condition affecting people who overuse muscles as a result of, for example, regularly operating a Combination computer keyboard and mouse or playing the piano. It is usually caused by some combination of inflammation and myofascial trigger points in the painful areas (Blatman Pain Clinic). Repetitive Strain Injury Joint Pain Common in athletes, frequent injuries over time can result in chronic pain. Typically these involve larger Deep somatic joints like the knee or the shoulder (Jacques, 2009). Restless Leg Syndrome Peripheral Pain A powerful urge to move the legs. Legs become uncomfortable when lying down or sitting. Some people Neuropathic describe it as a creeping, crawling, tingling or burning sensation. Moving relieves the discomfort, but not for long (American Chronic Pain Association, 2012). Rheumatoid Arthritis Joint Pain Often present in early adulthood, causes swelling in the joint spaces. Eventually it also damages bones, Deep somatic ligaments and tendons (Jacques, 2009). Sacroiliac Joint Dysfunction Back Pain Dysfunction in the sacroiliac joint, or SI joint, is thought to cause low back and/or leg pain. The leg pain can Combination be particularly difficult, and may feel similar to sciatica or pain caused by a lumbar disc herniation (American Chronic Pain Association, 2012). Sciatica Back/Leg Pain The sciatic nerve runs from your back to your feet. Compression or damage of this nerve can cause pain to Neuropathic shoot down the leg on one side of the body (Jacques, 2009). 247

CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Shingles (Postherpetic Generalised Pain A painful condition affecting your nerve fibers and skin; symptoms can become chronic. Postherpetic Neuropathic Neuralgia) neuralgia is a complication of shingles. During an initial infection of chickenpox, some of the virus remains in your body, lying dormant inside nerve cells. Years later, the virus may reactivate, causing shingles (a viral infection of the nerve roots). In Postherpetic neuralgia, shingles damaged skin nerves, causing pain signals to be sent to the brain (American Chronic Pain Association, 2012) (Helgason, Petursson, Gudmundsson, & Sigurdsson, 2000). Slipped or bulging disc Back Pain Often the result of twisting or lifting injuries. Damaged discs protrude into the spinal canal, pressing against Deep somatic (herniated disc) nerves as they exit the spinal cord (Jacques, 2009). Soft Tissue Damage Back Pain Heavy lifting or trauma can cause damage to back muscles, ligaments and tendons (Jacques, 2009). Deep somatic Somatization disorder Generalised Pain Recurring, multiple, clinically significant complaints about pain in different body sites that are not fully Psychogenic explained by a known medical condition (Allen, Woolfolk, Escobar, Gara, & Hamer, 2006). Sphincter of Oddi dysfunction Abdominal/Pelvic Pain Sphincter of Oddi dysfunction refers to structural disorders where there is an obstruction of bile and Visceral pancreatic juice flow, causing pain (American Chronic Pain Association, 2012). Spinal Stenosis Back Pain Narrowing of the spinal canal, which can compress nerves (Blatman Pain Clinic). Deep somatic/Neur opathic Spinal stenosis Back/Buttock Spinal stenosis is a condition where the narrowing of the spinal canal can cause spinal cord or nerve Neuropathic pinching which leads to chronic pain in the buttocks, limping, lack of feeling in the lower extremities, and decreased physical activity (American Chronic Pain Association, 2012). Trigeminal Neuralgia Headaches A neuropathic disorder of trigeminal nerve in the face. It causes episodes of intense pain in any or all of the Neuropathic (Fothergill’s disease) following: the ear, eye, lips, nose, scalp, forehead, teeth or jaw on one side of the face. Can be caused by Multiple Sclerosis (American Chronic Pain Association, 2012) (Office of Communications and Public Liaison, 2012). Stroke (Post-Stroke Pain Generalised Pain Pain that follows a stroke (a rapid loss of brain function due to a disturbance in the brain's blood supply) is Neuropathic Syndrome) termed post-stroke pain. Most strokes do not cause pain, only numbness. However, sometimes people experience ongoing deep burning, pins and needles sensations and often muscle contractions in various parts of the body, thought to be caused by damage to the neurons in the brain (American Chronic Pain Association, 2012). Structural Deformities Back Pain Spinal abnormalities such as scoliosis put strain on the muscles that control posture, causing pain and Deep somatic fatigue (Jacques, 2009).

248

CONDITION PAIN LOCATION DESCRIPTION TYPE OF PAIN Syringomyelia Generalised Pain A rare disorder that causes a cyst to form in the spinal cord which gets bigger and destroys part of the Neuropathic spinal cord. Damage to the spinal cord from the syrinx can cause symptoms such as Pain and weakness in the back, shoulders, arms or legs, headaches , inability to feel hot or cold (6). Temporomandibular joint Headaches/Joint Pain A painful condition involving the temporomandibular joint and the muscles used for chewing, sometimes Deep somatic dysfunction (3) causing clicking sounds and restricted jaw movement. It is a common and sometimes overlooked cause of chronic headaches (2). Thorocotomy (cut in the chest Generalised Pain Post Thorocotomy Pain Syndrome Pain after thoracotomy (cutting the chest open for surgery) is very severe Combination for surgery) and chronic pain, probably the most severe pain experienced after surgery, most likely caused by trauma to - increasing the intercostal nerve during surgery (6) (Karmaker & Ho, 2004). evidence to show it is neuropathic Transverse Myelitis Generalised Pain A neurological disorder caused by inflammation across both sides of one segment of the spinal cord. Neuropathic Attacks of inflammation can damage or destroy myelin, the fatty insulating substance that covers nerve cell fibers. This damage causes nervous system scars that interrupt communications between the nerves in the spinal cord and the rest of the body (6). Traumatic Brain Injury Generalised Pain A traumatic brain injury can cause damage to the way the nerves transmit signals to the brain; can be a Neuropathic cause of Central Pain Syndrome Traumatic Fracture Back Pain Falls from elevation, car accidents or crush injuries can cause painful vertebral fractures (1) Deep somatic Vulvodynia Abdominal/Pelvic Pain Chronic vulvar pain without an identifiable cause. The location, constancy and severity of the pain vary Visceral among sufferers. Some women experience pain in only one area of the vulva, while others experience pain in multiple areas. The most commonly reported symptom is burning, but women’s descriptions of the pain vary (6) (London Pain Clinic, 2006). Whiplash (Whiplash Generalised Pain Whiplash a non-medical term. It is caused by the neck and head being thrown suddenly backward then Combination syndrome) forward upon impact. The impact forces the neck and head beyond their normal range of movement, causing tissue damage and pain (6). In some cases, patients of whiplash continue experiencing pain even after a whiplash trauma is over: headaches and pain, reduced movement at the back of the neck, tingling in arms, lumbar pains, fatigue, sleep disturbances and reduced libido. Up to 60 percent of all whiplash injury patients suffer from lower back pain afterwards. This is commonly caused due to the injury to the discs, facet joints of the low back or the sacroiliac joints (London Pain Clinic, 2006).

249

Appendix F: General pain categories

Category Description Common Medications Joint Pain OsteoArthritis – inflammation of joints – osteoarthritis is commonly associated • Analgesics (pain relievers), from Paracetamol to opiates. (Arthritis/ with older populations. • Non-steroidal anti-inflammatory drugs (NSAIDs) such as naproxen Rheumatism) & ibuprofen (Nurofen), ketoprofen (Orudis) • Corticosteroids (taken by tablet or injection) (prednisolone, Rheumatoid Arthritis – has a younger age of onset than Osteoarthritis; the cortisone injection) body’s immune system starts to attack tissues, including joint tissues, which • Disease-modifying anti-rheumatic drugs (DMARDS) – suppress the causes inflammation and pain. immune system (biological DMARDS block substances that cause inflammation). Halting the inflammation reduces pain (Orencia, Humira) (Arthritis Information Sheet: Medicines and arthritis, 2012) Chronic back or Commonly pain in the lower back but can be in the neck. • Analgesics, including opioids neck problems • NSAIDs • Muscle Relaxants (benzodiazepines such as Valium or Xanax) • Antidepressants (Bogduk, 2004) Frequent or Most common form of chronic headache pain is Migraine – chronic, moderate- • NSAIDS, including Naproxen or ketorolac injections severe headaches severe headache & nausea. • Triptan (Seratonin antagonist) such as Sumatriptan • Antidepressants (Tricyclic or SSRI’s) may help prevent migraines • Corticosteroid injection Nerve Pain (Neuralgias) can also occur as a headache. • Ergot alkaloid - Ergotamine (vasoconstrictor alkaloid often combined with caffeine for treatment of migraines, i.e. Cafergot) • Combination of paracetamol, codeine and antihistamine doxylamine (Dolased or Mersyndol) (Duckro, 1999) (Consumer Medicine Information: Mersyndol, 2007) (Consumer Medicine Informaiton: Dolased Analgesic, 2004) Generalised Pain Common chronic generalised pain conditions (pain throughout the body) • Analgesics (including opioids) include: • NSAIDS • Antidepressants such as Duloxetine used for control of neuropathic • Complex Regional Pain Syndrome (CRPS) – intense, burning chronic pain caused by nerve damage pain conditions like Fibromyalgia. • 1. Reflex Sympathetic Dystrophy (CRPS1) – chronic nerve disorder occurs Anticonvulsants such as Lyrica are also used for neuropathic pain most often in arms or legs after a minor injury conditions Fibromyalgia (Consumer Medicine Information: Lyrica, 250

2. Causalgia (CRPS2) - caused by injury to the nerve 2011) (Eustice, 2008) • Fibromyalgia – chronic widespread pain, fatigue and heightened and painful response to pressure.

• Chronic fatigue syndrome – chronic muscle pain, joint pain, or headaches are some of the symptoms of this disorder. Visceral Pain Pain that is felt inside the body. Pain from the organs. Some common conditions Opioids are a common treatment for visceral pain (Giamberardino, involving visceral pain include: 2005). • Endometriosis – a condition affecting women caused by an excess of cells in the uterine cavity. Major symptom of this condition is recurring pelvic pain • Chronic Pancreatis - inflammation of the pancreas Other Neuropathic pain is caused by damaged or malfunctioning nerves. It is often Neuropathic pain is difficult to treat compared to somatic pain. A Neuropathic Pain described as different from other forms of pain and is often described as: combination of treatments can be used: Conditions electric shocks, burning, or “heavy” sensations. It can occur by itself or along • Antidepressants of the Tricyclic type with other forms of pain in many conditions. Some examples include: • Anticonvulsants (Gabapentin) • Alcohol dependence and Diabetes can contribute to peripheral neuropathy • Analgesics like opioids (may be ineffective but pain that does not (pain at extremities such as hands and feet) respond to other treatments may respond to opioids) • Phantom pain – pain from a part of the body that has been lost, common • Local anaesthetics experience of amputees and can become chronic. • Ketamine (Helme, 2006) • Operations such as mastectomies or thorocotomy or amputations can leave ongoing chronic neuropathic pain at the affected site.

251

Appendix G: Common medications used for pain conditions Brand of Medication Name of Drug Classification Type Comment Form Actiq (Consumer Medicine Fentanyl Opioid - Synthetic Analgesic Used for management of Lozenge or "lollipop" - fast- Information: Actiq, 2010) breakthrough pain in chronic release conditions Antenex (Consumer Medicine Diazepam Benzodiazepine Sedative/Muscle Relaxant Pill - (2mg & 5mg) Information: Antenex, 2008) Arcoxia (Consumer Medicine Etoricoxib NSAID (selective) Anti-Inflammatory Prescription medication used for Pill - 30mg, 60mg & 120mg Information: Arcoxia, 2011) arthritis or gout Biodone (Consumer Medicine Methadone Opioid - Synthetic Analgesic Low cost, long-acting Same strength as methadone information: Biodone, 1999) syrup but has fewer additives. It puts less strain on the liver to process and is better for those who have Hepatitis B or Hepatitis C (Frequently Asked Questions: Opioid Pharmacotherapy Treatments). Cafergot (Consumer Medicine Ergotamine Ergot Alkaloid Vasoconstrictor prescribed for migrane Pill (1mg ergotamine & 100mg Information: Cafergot, 2008) caffeine) & suppositories (2mg ergotamien & 100mg caffeine) Celebrex (Consumer Medicine Celecoxib NSAID (selective) Anti-Inflammatory Prescription medication used for Pill - 100mg & 200mg Information: Celebrex, 2010) arthritis or joint injuries Celestone Chronodose Cortisone Corticosteroid Steroid - Anti- Immunosuppressant that reduces Injection (Epidural) (Consumer Medicine Inflammatory inflammation Information: Celestone Chronodose, 2011)

252

Brand of Medication Name of Drug Classification Type Comment Form

Cicloral (Consumer Medicine Cyclosporin Disease Modifying Immunosuppressant Pill - (25mg, 50mg or 100mg) Information: Cicloral, 2010) Anti-Rheumatic Drug (DMARD) Codalgin Plus (Consumer Codeine Opioid Analgesic Over the counter pain-relief Pill - contains 500mg Medicine Information: Codalgin medication that contains an Paracetemol, 9.6 mg Codeine Plus, 2005) antihistamine with sedative effects. & 9.6 mg Doxylamine (antihistamine) Cortisone (Fadale & Wiggins, Cortisone Corticosteroid Steroid - Anti- Immunosuppressant that reduces Pill or Injection 1994) Inflammatory inflammation Cymbalta (Consumer Medicine Duloxetine Serotonin & Anti-Depressant Also prescribed for nerve pain Pill - 30mg, 60mg Information: Cymbalta, 2012) Hydrochloride Noradrenaline Reuptake Inhibitors (SNRI) Di-Gesic (Consumer Medicine Dextrapro- Opioid Analgesic 2012 - Therapeutic Goods Information: Di-Gesic, 2012) poxyphene Administration has ordered the de- registration of this drug Dilaudid (Consumer Medicine Hydromorphone Opioid -Semi- Analgesic Better speed of onset and lower instant-release pill taken every Informtation: Dilaudid, 2010) Synthetic dependence liability than morphine, 1-4 hours but dependence and tolerance can build quickly Dolased (Consumer Medicine Codeine, Analgesic Over the counter pain-relief Pill (500mg paracetemol, 10mg Informaiton: Dolased Analgesic, Doxylamine, & medication Codeine, 5.1mg Doxylamine 2004) Paracetemol Succinate) Doloxene (Consumer Medicine Dextrapro- Opioid Analgesic 2012 - Therapeutic Goods Information: Doloxene, 2012) poxyphene Administration has ordered the de- registration of this drug

253

Brand of Medication Name of Drug Classification Type Comment Form

Duragesic (Consumer Fentanyl Opioid - Synthetic Analgesic Used for management of chronic pain Patches - slow-release (48-72 Information: Duragesic, 2011) hrs) Duratram (Consumer Medicine Tramadol Opioid - Synthetic Analgesic Acts differently from other opioids - Information: Duratram, 2008) effective against nerve pain Endone (Consumer Medicine Oxycodone Opioid - Semi- Analgesic Can be used to manage chronic pain; Immediate-release pill taken Information: Endone, 2010) Synthetic can be better tolerated than every 4-6 hrs to manage morphine by some breakthrough pain (Analgesia (Pain-Relievers)). Humira (Consumer Medicine Adalimumab Antibody Antibody Intended for treatment of arthritis Injection Information: Humira, 20089) and other inflammatory diseases such as Crohn's disease Imigran (Consumer Medicine Sumatriptan Triptan Serotonin Receptor prescribed for migraine headaches Pill - 50mg & 100mg Information: Imigran, 2007) Agonist Jurnista (Consumer Medicine Hydromorphone Opioid -Semi- Analgesic Better speed of onset and lower 24 hour extended release pill Information: Jurnista prolonged Synthetic dependence liability than morphine, release tablet , 2010) but dependence and tolerance can build quickly Kapanol (Consumer Medicine Morphine Opioid Analgesic Often prescribed for chronic severe Available in both 12 hr or 24 Information: Kapanol, 2006) pain hour extended-release capsules Ledertrexate (Patient Methotrexate Disease Modifying Immunosuppressant Intended for treatment of Pill - once a week (2.5mg or Information on Methotrexate, Anti-Rheumatic Drug rheumatoid arthritis and lupus 10mg) 2008) (DMARD) Lyrica (Consumer Medicine Pregabalin Anti-Convulsant Anti-Convulsant Used to treat neuropathic pain Pill - 25mg, 75mg, 150mg, Information: Lyrica, 2011) 300mg

254

Brand of Medication Name of Drug Classification Type Comment Form Maxalt (Consumer Medicine Rizatriptan Triptan Serotonin Receptor prescribed for migraine headaches Wafer that is dissolved on Information: Maxalt Wafers, Agonist tongue - 5mg or 10mg 2012) Mersyndol (Consumer Medicine Codeine Opioid Analgesic Over the counter pain-relief Pill - contains 500mg Information: Mersyndol, 2007) medication Paracetmol & 9.6 mg Codeine Methadone syrup (Consumer Methadone Opioid - Synthetic Analgesic Low cost, long-acting liquid, Medicine Information: considered more effective against Methadone Syrup, 2007) neuropathic pain Methoblastin (Patient Methotrexate Disease Modifying Immunosuppressant Intended for treatment of Pill - once a week (2.5mg or Information on Methotrexate, Anti-Rheumatic Drug rheumatoid arthritis and lupus 10mg) 2008) (DMARD) Mobic (Consumer Medicine Meloxicam NSAID Anti-Inflammatory Prescription medication used to treat Pill - 7.5mg & 15mg Information: Mobic, 2009) arthritis MSContin (Consumer Medicine Morphine Opioid Analgesic Often prescribed for chronic severe Pills – extended-release taken Information: MS Contin, 2010) pain every 12 hours. Available in seven strengths: 5, 10, 15, 30, 60, 100, 200 mg. Naprosyn (Consumer Medicine Naproxen NSAID Anti-Inflammatory Prescription only pain relief often Pill - 250mg & 500mg, also in Information: Naprosyn, 2008) recommended for migraine or 24hr slow release tablets arthritis Naramig (Consumer Medicine Naratriptan Triptan Serotonin Receptor Can be prescribed for relief of Pill - 2.5mg Information: Naramig, 2007) Agonist migraine headaches Neoral (Consumer Medicine Cyclosporin Disease Modifying Immunosuppressant Prescription medication used to treat Either solution or pill (25mg, Information: Neoral, 2010) Anti-Rheumatic Drug rheumatoid arthritis 50mg, & 100mg) (DMARD) Normison (Consumer Medicine Temazepam Benzodiazepine Sedative/Hypnotic/Muscle Can be prescribed for chronic back Pill - 10 mg Information: Normison, 2009) Relaxant pain 255

Brand of Medication Name of Drug Classification Type Comment Form

Norspan (Consumer Medicine Bupenorphine Opioid - Semi- Analgesic Controlled release of medication to Patch - slow release (7 days) Information: Norspan, 2009) Synthetic client Numorphan (Oxymorphone) Oxymorphone Opioid - Semi- Analgesic Not available in Australia - can be Synthetic imported Numorphone (Oxymorphone) Oxymorphone Opioid - Semi- Analgesic Not available in Australia - can be Synthetic imported Nurofen (Pain relievers Ibuprofen NSAID Anti-Inflammatory Over the counter pain-relief Pill, Gel, Heat Patches explained, 2012) medication Nurofen Plus (Consumer Codeine Opioid Analgesic Over the counter pain-relief Pill - contains 200mg Ibuprofen Medicine Information: Nurofen medication & 12.8 mg Codeine Plus, 2009) Opana (Opana Consumer Oxymorphone Opioid - Semi- Analgesic Not available in Australia - can be Information, 2012) Synthetic imported Ordine (Consumer Medicine Morphine Opioid Analgesic Often prescribed for chronic severe Fast-acting Oral liquid solution: Information: Ordine, 2010) pain when other treatments are not 1,2,5,10 mg (every four hours) effective. Orencia (Consumer Medicine Abatacept Disease Modifying Immunosuppressant Used for rheumatoid arthritis when Either slow infusion into vein Information: Orencia, 2012) Anti-Rheumatic Drug there is an inadequate response to (30 minutes) or injection (DMARD) other drugs. Orudis (Consumer Medicine Ketoprofen NSAID Anti-Inflammatory Prescription only pain relief often Topical Gel (750mg/30g or Information: Orudis, 2011) recommended for arthritis or other 1.5g/60g) or capsules (200 mg) musculo-skeletal inflammation Oruvail (Consumer Medicine Ketoprofen NSAID Anti-Inflammatory Prescription only pain reliever often Pill - 200mg Information: Oruvail, 2011) used to reduce pain & inflamation with arthritis

256

Brand of Medication Name of Drug Classification Type Comment Form

OxyContin (Consumer Medicine Oxycodone Opioid - Semi- Analgesic Can be used to manage chronic pain; Extended-Release Pill (every 12 Information: OxyContin Tablets, Synthetic can be better tolerated than hours) – 5, 10, 15, 20, 30, 40, 2010) morphine 80 mg OxyNorm (Consumer Medicine Oxycodone Opioid - Semi- Analgesic Can be used to manage chronic pain; Short-acting pill (every 4-6 hrs) Information: OxyNorm, 2010) Synthetic can be better tolerated than - 5, 10, 20 mg morphine Panadeine (Consumer Medicine Codeine Opioid Analgesic Over the counter pain-relief Pill, rapid soluble, contains Information: Panadeine, 2009) medication 500mg Paracetemol & 8mg Codeine Panadeine Extra (Consumer Codeine Opioid Analgesic Strongest over-the counter Pill - contains 500mg Medicine Information: medication containing codeine; Paracetemol & 15 mg Codeine Panadeine Extra, 2009) contains paracetamol & codeine Panadeine Forte (Consumer Codeine Opioid Analgesic Only available with prescription Pill - contains 500mg Medicine Information: Paracetemol & 30 mg Codeine Panadeine Forte, 2007) Panadol Extra (Panadol Extra Paracetemol NSAID Anti-Inflammatory Over the counter pain-relief Pill - contains 500mg Product Information, 2009) medication recommended for Paracetemol & 65mg Caffeine headaches, arthritis, and other pain. Caffeine helps with the absorption of the paracetamol. Panafcort (Consumer Medicine Prednisone Corticosteroid Steroid - Anti- Reduce swelling and pain in Pill - available in 1mg, 5mg, Information: Panafcort, 2010) Inflammatory joints/other organs and 25mg doses

Panafcortelone (Consumer Prednisolone Corticosteroid Steroid - Anti- Reduce swelling and pain in Pill - available in 1mg, 5mg, Medicine Information: Inflammatory joints/other organs and 25mg doses Panafcortelone, 2009)

257

Brand of Medication Name of Drug Classification Type Comment Form

Paradex (Consumer Medicine Dextropro- Opioid Analgesic 2012 - Therapeutic Goods Pill – 32.5mg Information: Paradex, 2010) poxyphene Administration has ordered the de- dextropropoxyphene & 325mg registration of this drug in Australia Paracetamol (Update on TGA decision to cancel prescription pain-killers, 2012) Pethidine (Consumer Medicine Pethidine Opioid - Synthetic Analgesic High Toxicity and Fast release has Available by white tablet or Information: Pethidine, 2008) reduced popularity for prescription medically-supervised injection for chronic pain. Higher reported euphoric sensations than other opiates. Physeptone (Consumer Medicine Methadone Opioid - Synthetic Analgesic Methadone in tablet form - more Information: Physeptone likely to be subscribed for relief of Tablets, 2011) chronic pain Plaquenil (Consumer Medicine Hydroxychloroquine Anti-Malarial Anti-Inflammatory Prescription only - often prescribed Pill - 200mg Information: Plaquenil, 2006) Sulfate for arthritis or lupus Predsolone (Consumer Medicine Prednisolone Corticosteroid Steroid - Anti- Used to reduce symptoms of Pill - available in 1mg dose Information: Predsolone, 2009) Inflammatory inflammation, including pain in arthritis. Predsone (Consumer Medicine Prednisone Corticosteroid Steroid - Anti- Used to reduce symptoms of Pill - available in 1mg dose Information: Predsone, 2009) Inflammatory inflammation, including pain in arthritis. Rapifen (Consumer Medicine Alfentanil Opioid - Synthetic Analgesic Potent, short-acting, often used for Injection Information: Rapifen, 2011) anaesthesia Relpax (Consumer Medicine Eletriptan Triptan Serotonin Receptor prescribed for relief of migraine Pill - 40mg & 80 mg Information: Relpax, 2012) Agonist headaches

258

Brand of Medication Name of Drug Classification Type Comment Form

Sandimmun (Consumer Medicine Cyclosporin Disease Modifying Immunosuppressant Can be prescribed to reduce Either solution or pill (25mg, Information: Sandimmun, 2011) Anti-Rheumatic Drug inflammation in rheumatoid arthritis 50mg, & 100mg) (DMARD) Solone (Consumer Medicine Prednisolone Corticosteroid Steroid - Anti- Pill - available in 5mg & 25mg Information: Solone, 1997) Inflammatory doses Sone (Consumer Medicine Prednisone Corticosteroid Steroid - Anti- Pill - available in 5mg & 25mg Information: Sone, 2001) Inflammatory doses Suboxone (Consumer Medicine Bupenorphine Opioid - Semi- Analgesic Used as both analgesic and opioid Pill or Film form - contains Information: Suboxone, 2010) Synthetic substitution treatment opioid antagonist Naloxone to discourage intravenous drug use Subutex (PBS - Subutex Bupenorphine Opioid - Semi- Analgesic Used as both analgesic and opioid Pill - administered sublingually (buprenorphine), 2003) Synthetic substitution treatment Sumagran (Consumer Medicine Sumatriptan Triptan Serotonin Receptor prescribed for relief of migraine Pill - 50mg & 100mg Information: Sumagran, 2007) Agonist headaches Sumatab (Consumer Medicine Sumatriptan Triptan Serotonin Receptor prescribed for migraine headaches Pill - 50mg & 100mg Information: Sumatab, 2009) Agonist

Targin (Consumer Medicine Oxycodone Opioid - Semi- Analgesic Can be used to manage chronic pain; Pill – 2.5, 5, 10, 20 mg Information: Targin, 2011) Synthetic contains naloxone which can block some of the effects of opioids in the gut, such as constipation. Temgesic (Consumer Medicine Bupenorphine Opioid - Semi- Analgesic Used as both analgesic and opioid Pill - administered sublingually, Information: Temgesic, 2008) Synthetic substitution treatment or injection

259

Brand of Medication Name of Drug Classification Type Comment Form

Tramal (Consumer Medicine Tramadol Opioid - Synthetic Analgesic Acts differently from other opioids - Information: Tramal, 2006) effective against nerve pain Valium (Consumer Medicine Diazepam Benzodiazepine Sedative/Muscle Relaxant Can be used as a muscle relaxant for Pill - (2mg & 5mg) Information: Valium, 2010) back pain Vicodin Hydrocodone Opioid - Semi- Analgesic Not available in Australia - can be Hydrocodone is similar to Synthetic imported codeine. Combined with paracetamol. Xanax (Consumer Medicine Alprazolam Benzodiazepine Sedative/Muscle Relaxant Can be used as a muscle relaxant for Information: Xanax, 2011) back pain Xylocaine (Consumer Medicine Lignocaine Local Anaesthetic Anaesthetic can be used to soothe peripheral Injection or Jelly 20mg/g Information: Xylocaine, March) Hydrochloride neuropathic pain Zomig (Consumer Medicine Zolmitriptan Triptan Serotonin Receptor prescribed for relief of migraine Pill (2.5mg or 5mg) or orally information: Zomig, 2006) Agonist headaches dispersible tablets Zydol (Consumer Medicine Tramadol Opioid - Synthetic Analgesic Acts differently from other opioids - Sustained release pills (1-2 per Information: Zydol, 2011) effective against nerve pain day) – 50, 100, 150 & 200 mg

260

Chapter eleven: Appendices

Appendix H: Initial fax sent to pharmacies

261

Chapter eleven: Appendices

262

Chapter eleven: Appendices

Appendix I: Participant invitation to research

263

Chapter eleven: Appendices

264

Chapter eleven: Appendices

Appendix J: Participant information

265

Chapter eleven: Appendices

266

Chapter eleven: Appendices

267

Chapter eleven: Appendices

Appendix K: POINT participant consent form

268

Chapter eleven: Appendices

269

Chapter eleven: Appendices

Appendix L: POINT locator form

270

Chapter eleven: Appendices

271

Chapter eleven: Appendices

272

Chapter eleven: Appendices

Appendix M: POINT baseline interview

273

Chapter eleven: Appendices

274

Chapter eleven: Appendices

275

Chapter eleven: Appendices

276

Chapter eleven: Appendices

277

Chapter eleven: Appendices

278

Chapter eleven: Appendices

279

Chapter eleven: Appendices

280

Chapter eleven: Appendices

281

Chapter eleven: Appendices

282

Chapter eleven: Appendices

283

Chapter eleven: Appendices

284

Chapter eleven: Appendices

285

Chapter eleven: Appendices

286

Chapter eleven: Appendices

287

Chapter eleven: Appendices

288

Chapter eleven: Appendices

289

Chapter eleven: Appendices

290

Chapter eleven: Appendices

291

Chapter eleven: Appendices

292

Chapter eleven: Appendices

293

Chapter eleven: Appendices

Appendix N: POINT baseline self-complete questionnaire

294

Chapter eleven: Appendices

295

Chapter eleven: Appendices

296

Chapter eleven: Appendices

297

Chapter eleven: Appendices

298

Chapter eleven: Appendices

299

Chapter eleven: Appendices

300

Chapter eleven: Appendices

301

Chapter eleven: Appendices

302

Chapter eleven: Appendices

303

Chapter eleven: Appendices

304

Chapter eleven: Appendices

305

Chapter eleven: Appendices

306

Appendix O: POINT 7-day medication diary

307