A Realist Evaluation of Patients' Decisions to Deprescribe in the EMPOWER Trial

BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from

A realist evaluation of patients' decisions to deprescribe in the EMPOWER trial

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-015959

Article Type: Research

Date Submitted by the Author: 12-Jan-2017

Complete List of Authors: Martin, Philippe; Centre de recherche de l'Institut universitaire de gériatrie de Montréal, Tannenbaum, Cara; Université de Montréal,

Primary Subject Geriatric medicine Heading:

Secondary Subject Heading: Health services research, Patient-centred medicine, Addiction

Keywords: deprescribing, benzodiazepines, realist evaluation, mechanisms, EMPOWER

http://bmjopen.bmj.com/

on September 24, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 1 A realist evaluation of patients' decisions to deprescribe in the EMPOWER trial 4 5 a,b b,c 6 2 Philippe Martin, BSc , Cara Tannenbaum, MD MSc 7 8 9 3 10 11 a 12 4 Doctoral Student, Faculty of Pharmacy, Université de Montréal, Montreal, PQ Canada 13 14 15 b For peer review only 16 5 Institut Universitaire de Gériatrie de Montréal, Montreal, PQ Canada 17 18 c 19 6 Professor and Michel Saucier Endowed Chair, Université de Montréal, Faculties of Medicine 20 21 22 7 and Pharmacy, Montreal, PQ, Canada. 23 24 25 8 26 27 9 Corresponding Author: 28 29 10 Philippe Martin, B. Sc. 30 31 32 11 Institut Universitaire de Gériatrie de Montréal,

33 http://bmjopen.bmj.com/ 34 12 Faculties of Pharmacy and Medicine, 35 36 37 13 Université de Montréal, Montréal, Quebec, Canada 38 39 14 4545 Queen Mary Road 40

41 15 Montreal, QC H3W 1W5 on September 24, 2021 by guest. Protected copyright. 42 43 44 16 E-mail : [email protected] 45 46 17 Tel: 514-340-3540 ext: 4728 47 48 18 Fax: 514-340-2117 49 50 51 19 Alternate Corresponding Author: [email protected] 52 53 20 Word Count: Abstract (262 words), Main text (2877 words) 54 55 21 56 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 22 Funding statement: This work was supported by Operating Grant OTG-88591 from the 4 5 6 23 Canadian Institutes of Health Research (CIHR). Philippe Martin received a doctoral bursary 7 8 24 from the FRQS. Cara Tannenbaum was supported by a Senior Scientist Career Award from the 9 10 11 25 FRQS. The above funding organizations had no role in the design and conduct of the study; 12 13 26 collection, management, analysis, and interpretation of the data; preparation, review, or approval 14 15 27 of the manuscript;For or the decision peer to submit review the manuscript for publication.only 16 17 18 28 19 20 29 Data sharing statement: Dataset is available upon request to the corresponding author. 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 30 STRUCTURED ABSTRACT: 4 5 6 31 BACKGROUND AND OBJECTIVES: Successful mechanisms for engaging patients in the 7 8 32 deprescribing process remain unknown. We sought to determine in which contexts the 9 10 11 33 EMPOWER deprescribing intervention triggered older adults’: (1) motivation to deprescribe by 12 13 34 increasing knowledge and concern about benzodiazepines; (2) capacity to taper by augmenting 14 15 35 self-efficacy; andFor (3) opportunities peer to discuss review and receive support only from a healthcare provider to 16 17 18 36 deprescribe. 19 20 37 DESIGN: A realist evaluation using a sequential mixed methods approach, conducted alongside 21 22 38 the EMPOWER randomized clinical trial 23 24 25 39 SETTING: Community, Quebec, Canada. 26 27 40 PARTICIPANTS: 261 older chronic benzodiazepine consumers, who received the EMPOWER 28 29 41 intervention and had complete 6 month follow up data. 30 31 32 42 INTERVENTION: Mailed deprescribing brochure on benzodiazepines.

33 http://bmjopen.bmj.com/ 34 43 MEASUREMENTS: Motivation (change in knowledge test score; change in beliefs about the 35 36 37 44 risk-benefits of benzodiazepines, measured with the Beliefs about Medicines questionnaire; 38 39 45 intent to discuss deprescribing); Capacity (self-efficacy for tapering) and Opportunity (support 40

41 46 from a physician or pharmacist) on September 24, 2021 by guest. Protected copyright. 42 43 44 47 RESULTS: The intervention triggered an intent to deprescribe among 167 (n=64%) participants 45 46 48 (mean age 74.6 years + 6.3, 72% women). Participants intending to deprescribe had improved 47 48 49 knowledge (risk difference, 58.50% [95% CI, 46.98%-67.44%]), increased concern about taking 49 50 51 50 benzodiazepines (risk difference, 67.67% [95% CI, 57.36%-74.91%]), and higher self-efficacy 52 53 51 for tapering (risk difference, 56.90% [95% CI, 45.41%-65.77%]). Contexts where the 54 55 52 deprescribing mechanisms failed included lack of support from a health care provider, 56 57 58 59 60 3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 53 intolerance to withdrawal symptoms, perceived poor health, and psychological reliance on 4 5 6 54 benzodiazepines. 7 8 55 CONCLUSION: Deprescribing interventions that systematically target patient motivation and 9 10 11 56 capacity to deprescribe are successful to the extent that healthcare providers are supportive, and 12 13 57 that patients do not have internal competing motivations to remain on drug therapy. 14 15 58 For peer review only 16 17 18 59 Key words: deprescribing, benzodiazepines, realist evaluation, mechanisms, EMPOWER 19 20 60 ClinicalTrials.gov identifier is NCT01148186 21 61 22 62 23 24 25 63 26 64 ARTICLE SUMMARY: 27 65 28 66 STRENGTHS AND LIMITATIONS OF THE STUDY: 29 30 67 • Use of a mixed methods approach enabled us to explore the breadth, depth, and 31 68 complexity of the patient’s experience of deprescribing. 32 69 • Use of the realist evaluation allowed us to investigate how the mechanisms underlying 33 70 deprescribing interventions interact with specific contexts to yield positive or negative http://bmjopen.bmj.com/ 34 71 outcomes 35 36 72 • This study was conducted alongside a large cluster randomized clinical trial. 37 73 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 74 INTRODUCTION 4 5 6 75 Deprescribing refers to the collaborative process of tapering, discontinuing, stopping, or 7 8 76 withdrawing medications in order to reduce adverse drug events and improve outcomes.1-5 9 10 1 3 6 11 77 Deprescribing has many steps , with one key component being the engagement of patients in 12 1 7-15 13 78 shared decision-making. Research suggests that older adults have conflicted feelings about 14 15 79 medications4 14: For78% of older peer adults believe review that medications onlyare necessary to improve health, 16 17 18 80 but at the same time, 68% would like to reduce their current medication use, with 92% willing to 19 20 81 stop a regular medication if advised to do so by their physician14. 21 22 82 23 24 25 83 A better understanding of the mechanisms that trigger patient motivation and capacity to 26 27 84 engage in the deprescribing process could reduce the use of potentially inappropriate medications 28 29 85 in older adults. The aim of realist evaluation is to reveal how an intervention might generate 30 31 32 86 different outcomes in different circumstances, and how different mechanisms work in particular

33 http://bmjopen.bmj.com/ 34 87 contexts, by enabling or motivating participants to make different choices 16. Educational 35 36 37 88 strategies to increase patients’ knowledge, beliefs, and motivation are hypothesized to influence 38 10 39 89 deliberate action on the part of the patient to curtail the use of a drug . However, what works, 40

41 90 for whom, under which circumstances and why, are questions that have never been explored on September 24, 2021 by guest. Protected copyright. 42 43 44 91 systematically from the patient’s point of view. Recent reviews on deprescribing call for a 45 46 92 realistic evaluation of large deprescribing trials to investigate how the mechanisms underlying 47 48 93 deprescribing interventions interact with specific contexts to yield positive or negative 49 50 17 18 51 94 outcomes. The EMPOWER trial, which demonstrated a number-needed-to-treat of 4 for the 52 53 95 effectiveness of mailing a benzodiazepine deprescribing brochure on complete cessation of 54 55 56 57 58 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 96 benzodiazepines at 6 months, provides a timely opportunity to examine which deprescribing 4 5 12 6 97 mechanisms worked for different patients under different circumstances. 7 8 98 9 10 11 99 The initial theory underpinning the development of the EMPOWER intervention was 12 13 100 that most – if not all - older adults are unaware of the age-related harms of taking benzodiazepine 14 15 101 anti-anxiety drugsFor and sleeping peer pills. Side review effects of sedative-hypnotics only are well-documented in 16 17 18 102 the literature but rarely talked about in practice as being a potential cause of memory impairment, 19 20 103 falls and fractures19-24, feared by many older adults25 26. Not understanding why medications 21 22 104 should be discontinued is a patient barrier to deprescribing4 27. As most patients are uninformed 23 24 25 105 of the potential risks associated with the use of benzodiazepines, we hypothesized a linear 26 27 106 behavior change process whereby providing patients with an interactive educational brochure 28 29 107 detailing associated risks, safer alternatives, and steps for tapering, would trigger patients’ 30 31 32 108 motivation, capacity and opportunity to initiate the deprescribing process through discussion of

33 http://bmjopen.bmj.com/ 34 109 medication discontinuation with a healthcare provider. 35 36 37 110 38 39 111 This paper reports a realist evaluation of the deprescribing process from the patient’s 40

41 112 perspective. The realist evaluation answers the following three research questions: (1) Did the on September 24, 2021 by guest. Protected copyright. 42 43 44 113 EMPOWER intervention trigger patients’ motivation to deprescribe by increasing knowledge 45 46 114 and concern about benzodiazepines? (2) Did the intervention augment patients’ capacity and 47 48 115 self-efficacy to taper benzodiazepines? (3) Did the intervention create opportunities for the 49 50 51 116 patient to discuss and receive support from a healthcare provider to engage in the deprescribing 52 53 117 process? Our goal was to determine for whom the EMPOWER intervention worked, in what 54 55 118 context it worked best and in which situations it failed. 56 57 58 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 119 METHODS 4 5 6 120 Study design 7 8 121 A realist evaluation was conducted alongside the EMPOWER randomized controlled trial. 9 10 16 11 122 This report follows RAMESES II guidelines for realist evaluation. The approach was chosen to 12 12 13 123 expand on the results of the EMPOWER trial, in order to inform the implementation of future 14 15 124 deprescribing initiativesFor by peerexamining the reviewpossible causes and contextualonly factors associated with 16 17 28 18 125 change. Specifically, we chose to use a theory-based, sequential mixed methods approach in 19 20 126 order to gain a deeper understanding of the contexts, mechanisms and outcomes of the 21 22 127 deprescribing intervention as experienced by the patient. The study was approved by the Institut 23 24 25 128 Universitaire de Gériatrie de Montréal Ethics Committee in Montreal, Quebec, Canada. 26 27 129 28 29 130 Environment surrounding the evaluation 30 31 32 131 The EMPOWER trial “Eliminating Medications through Patient Ownership of End Results”

33 http://bmjopen.bmj.com/ 34 132 was a pragmatic randomized trial that examined the effectiveness of a direct-to-consumer, 35 36 37 133 written educational brochure mailed directly to patients on subsequent discontinuation of 38 29 39 134 sedative-hypnotic medication. The EMPOWER trial was rolled out between July 2010 and 40

41 135 November 2013, with community-dwelling participants randomly recruited via pharmacists on September 24, 2021 by guest. Protected copyright. 42 43 44 136 located within a 200 km radius of the Montreal urban area in Quebec, Canada. Participants were 45 46 137 303 older, community-dwelling, chronic users of benzodiazepine medication, and agreed to 47 48 138 home visits and telephone follow-up interviews by the research team. All benzodiazepine 49 50 51 139 prescriptions for seniors were covered under the publicly financed drug plan in the province of 52 53 140 Quebec, excluding the program’s deductible (if applicable). Provincial governments covered 54 55 141 physician reimbursements for patient visits, and drug dispensing fees for pharmacists, as part of 56 57 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 142 Canada’s universal health care program. 4 5 6 143 7 8 144 The EMPOWER intervention 9 10 11 145 The 8-page EMPOWER brochure, available at 12 30 13 146 http://www.criugm.qc.ca/fichier/pdf/BENZOeng.pdf, aims to promote active learning by 14 15 147 incorporating andFor using constructivist peer learning review principles31. The only brochure includes a self- 16 17 18 148 assessment component and presentation of the evidence-based risks associated with 19 20 149 benzodiazepine use in an effort to elicit cognitive dissonance.10 Elements of social comparison 21 22 150 theory32, through the use of peer champion stories, are also integrated in the intervention. The 23 24 25 151 brochure provides a self-guided tapering schedule, consisting of a visual tapering protocol 26 27 152 showing pictures of full pills, halved pills and quartered pills30. 28 29 153 30 31 32 154 The program theories embedded in the EMPOWER intervention are based on Mitchie et

33 http://bmjopen.bmj.com/ 34 155 al’s behavior change wheel33, targeting motivation, capacity and opportunity. Mitchie et al. 35 36 37 156 define motivation as the mental process that energizes and directs behaviors. Capability refers to 38 39 157 the psychological and physical capacity of the individual to engage in the behavior. Opportunity 40

41 158 refers to the external factors that permit or promote a behavior to happen, and include both the on September 24, 2021 by guest. Protected copyright. 42 43 44 159 physical and social environment of the individual. The specific mechanisms that we aimed to test 45 46 160 with the realist evaluation were whether the EMPOWER brochure: (1) triggered older adults’ 47 48 161 motivation to deprescribe by increasing knowledge and concern about benzodiazepines; (2) built 49 50 51 162 capacity to taper by augmenting self-efficacy; and (3) drove opportunities to receive support 52 53 163 from a healthcare provider to deprescribe. Table 1 links the programme theories and mechanisms 54 55 164 to the corresponding intervention components. 56 57 58 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 165 4 5 6 166 Evaluation of the programme theories 7 8 167 The evaluation of the programme theories consisted of quantitative data collection and 9 10 11 168 analysis, qualitative data collection and analysis, and triangulation of the quantitative and 12 34 13 169 qualitative results. Data collection was conducted between July 2010 and November 2013 as 14 15 170 part of the EMPOWERFor clinical peer trial. Analysis, review triangulation and only refinement of the Context- 16 17 18 171 Mechanism-Outcome configuration took place subsequent to completion of the trial. 19 20 172 21 22 173 Data collection methods 23 24 25 174 Quantitative data included pre- and 1-week post-intervention information on knowledge 26 27 175 about benzodiazepine-related harms, beliefs about the necessity of taking benzodiazepines versus 28 29 176 concern about harms, self-efficacy for tapering, and intent to discuss deprescribing with a health 30 31 32 177 care provider. We measured gains in knowledge with the four true or false questions listed in the

33 http://bmjopen.bmj.com/ 34 178 ‘‘Test Your Knowledge’’ section of the questionnaire.29 30 Correct answers were summed to a 35 36 37 179 maximum of 4 points, and answers were compared prior to and after receiving the intervention. 38 39 180 Participants’ beliefs about consuming benzodiazepines were measured with the Beliefs about 40

41 181 Medicines questionnaire (BMQ-Specific) at both time points. The BMQ-Specific consists of two on September 24, 2021 by guest. Protected copyright. 42 43 44 182 validated 5-item sub-scales assessing the respondents’ perceptions about the necessity and 45 46 183 concerns associated with taking benzodiazepines.35 Participants indicate their degree of 47 48 184 agreement with each statement on a 5 point Likert scale (1=strongly disagree, 5=strongly agree). 49 50 51 185 Scores are summed into their respective sub-category (5-25 point scale) with higher scores 52 53 186 indicating stronger beliefs. Risk perception was assessed using a single question 1-week post 54 55 187 intervention in which participants were asked whether they perceived the same, increased, or no 56 57 58 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 188 risk from consumption of their benzodiazepine following the intervention. In order to determine 4 5 6 189 whether the EMPOWER brochure increased capacity to taper by augmenting self-efficacy, we 7 8 190 measured self-efficacy for tapering on the Medication Reduction Self-efficacy scale, which 9 10 11 191 allows the respondent to rate on a scale of 0 to 100 their degree of confidence for tapering 12 36 13 192 benzodiazepines. Participants were also asked to indicate (yes/no) post intervention if they had 14 15 193 spoken to or intendedFor to discuss peer medication review discontinuation with only their doctor and/or pharmacist. 16 17 18 194 Health status was assessed at baseline using the first item of the Short-Form-12 Health Survey 19 20 195 and dichotomized by categorizing poor to fair responses as poor health.37 21 22 196 23 24 25 197 Qualitative data were collected after the 6-month follow-up, using semi-structured 26 27 198 interviews conducted at participants’ homes. Twenty-one participants were strategically sampled 28 29 199 for in-depth qualitative interviews using a contrast sample design, based on cessation of 30 31 38 32 200 benzodiazepines (yes or no) combined with intent to discuss tapering (yes or no). Interviews

33 http://bmjopen.bmj.com/ 34 201 lasted approximately one hour, were recorded with consent and professionally transcribed 35 36 37 202 verbatim. The interviews were based on a pre-established discussion guide, the major themes of 38 39 203 which included initial reactions to the intervention, reasons underlying the decision to taper, 40

41 204 experience with the tapering process, and personal interactions with health care providers. on September 24, 2021 by guest. Protected copyright. 42 43 44 205 45 46 206 Analysis 47 48 207 Participants with complete follow-up data were included in the quantitative analysis 49 50 51 208 (n=261, mean age 74.6 + 6.3, 72% women). Participant characteristics were described and 52 53 209 compared using means with standard deviations and independent t-tests for continuous data, and 54 55 210 percentages and Chi-square tests for categorical data, according to each of three outcomes: intent 56 57 58 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 211 to deprescribe with successful discontinuation, intent to deprescribe with failed discontinuation, 4 5 6 212 and no intent to deprescribe. Individual changes were computed from baseline to post- 7 8 213 intervention in knowledge, in the BMQ necessity and concerns subscales, and in self-efficacy 9 10 11 214 scores for tapering. Risk differences with 95% confidence intervals were calculated for the 12 13 215 proportion of participants in each group who demonstrated increased knowledge, heightened 14 15 216 concern about benzodiazepineFor peer use, and augmented review self-efficacy onlyfor tapering. The statistical 16 17 39 18 217 significance for all analyses was set at p<0.05 (two-sided). SPSS Version 21.0 (SPSS Inc. 19 20 218 Chicago, IL, USA) was used for all analyses. 21 22 219 Qualitative data from the semi-structured interviews were analyzed using thematic content 23 24 40 25 220 analysis. Discourses were contrasted according to whether participants discontinued 26 27 221 benzodiazepines and/or expressed the intent to discuss discontinuation. Interviews were coded 28 29 222 using Dedoose software. Themes were derived from the data and supported by quotes. Initially, 30 31 32 223 two researchers independently read the transcripts and field notes, then collaboratively developed

33 http://bmjopen.bmj.com/ 34 224 first order codes, which were subsequently verified by double coding. Second order thematic 35 36 37 225 coding was performed for the purpose of building concepts. 38 39 226 40

41 227 Quantitative and qualitative results were combined and analyzed in an iterative fashion on September 24, 2021 by guest. Protected copyright. 42 43 41 44 228 through use of a triangulation protocol using a convergence coding matrix, as described by 45 46 229 Farmer et al.42 The convergence matrix served to inform which theories should be retained or 47 48 230 discarded, as well as which contexts favorably or unfavorably influenced a patient’s trajectory 49 50 51 231 based on agreement, partial agreement, or dissonance between the quantitative and qualitative 52 53 232 data.41 42 Differences were adjudicated via discussion and consensus. 42 The convergence-coding 54 55 233 matrix is available from the authors upon request. 56 57 58 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 234 RESULTS 4 5 6 235 Contexts in which the EMPOWER intervention succeeded in triggering motivation and 7 8 236 capacity to deprescribe 9 10 11 237 The EMPOWER intervention triggered motivation and capacity to deprescribe in 167 of 12 13 238 261 participants (64%). Compared to individuals who reported no intent to deprescribe, 14 15 239 participants whoFor intended topeer deprescribe werereview significantly more only likely to demonstrate improved 16 17 18 240 knowledge (risk difference, 58.50% [95% CI, 46.98%-67.44 %]), lower perceived necessity (risk 19 20 241 difference, 56.03% [95% CI, 44.63%-64.81%]), increased concern (risk difference, 67.67% 21 22 242 [95% CI, 57.36%-74.91%]), enhanced self-efficacy for tapering (risk difference, 56.90% [95% 23 24 25 243 CI, 45.41%-65.77%]), and a greater perception of risk (risk difference, 35.14% [95% CI, 26 27 244 23.06%-45.39%]) (Table 2). No differences were observed in the age, sex, educational level, or 28 29 245 indication for benzodiazepine use between the two groups. Individuals who did not intend to 30 31 32 246 deprescribe were more likely to report poor health (40% vs 28%, 12.28% [95% CI, 0.44 %-

33 http://bmjopen.bmj.com/ 34 247 24.18 %] ). Favorable contexts that enabled the EMPOWER mechanisms to work included 35 36 37 248 previous support and encouragement from a health care provider, stable health status, and a 38 39 249 positive outlook on aging. Table 3 shows the theme frequencies and supporting citations that 40

41 250 emerged from the qualitative interviews, describing the contexts that enabled deprescribing. on September 24, 2021 by guest. Protected copyright. 42 43 44 251 45 46 252 Contexts in which the EMPOWER intervention failed 47 48 253 Two points were identified along the deprescribing process where the program theories 49 50 51 254 failed. The first was at the point where the intervention aimed to trigger motivation to 52 53 255 deprescribe. This situation occurred in 36% of participants, and was characterized by a lack of 54 55 256 new knowledge and no increase in perceived risk after receipt of the EMPOWER brochure. The 56 57 58 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 257 contexts included previous reassurance by a physician that benzodiazepines were safe or 4 5 6 258 necessary, the belief that the benefits of benzodiazepines outweighed the risks for immediate 7 8 259 symptom relief, and a reliance on benzodiazepines for everyday coping because of poor health 9 10 11 260 (Table 4). The second failure point occurred during participants’ efforts to initiate or complete 12 13 261 the deprescribing process. Lack of support from a healthcare provider, intolerance to withdrawal 14 15 262 symptoms, and aFor sudden loss peer of confidence review to live without sleeping only pills were contexts that led 16 17 18 263 participants to abort the deprescribing process once they showed initial motivation and capacity 19 20 264 to deprescribe (Table 4). 21 22 265 23 24 25 266 Refining the context-mechanism-outcome configuration for deprescribing interventions 26 27 267 The initial context-mechanism-outcome configuration that drove the development of the 28 29 268 EMPOWER intervention was a simple, linear progression along different stages of readiness to 30 31 32 269 deprescribe, similar to Prochaska & DiClemente’s transtheoretical model of change (Figure

33 http://bmjopen.bmj.com/ 34 270 1a).43 We believed that the EMPOWER brochure would move patients from pre-contemplation 35 36 37 271 about deprecribing to action and maintenance, by increasing knowledge about the harms of 38 39 272 benzodiazepines, enhancing self-efficacy, and triggering motivation and capacity to create 40

41 273 opportunities to discuss deprescribing with a healthcare professional. We assumed that the on September 24, 2021 by guest. Protected copyright. 42 43 44 274 healthcare provider would be supportive and accompany the patient along the deprescribing 45 46 275 process. This initial configuration oversimplified the stages through which individuals 47 48 276 transitioned after receiving the deprescribing intervention. Figure 1b depicts a revised, non-linear 49 50 51 277 context-mechanism-outcome configuration that takes into account the complexity of internal and 52 53 278 external influences on initiating and completing the deprescribing process from the consumer’s 54 55 279 perspective. The revised model recognizes that new information influences beliefs and actions 56 57 58 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 280 only if the information generates a desire strong enough not to be overwhelmed by competing 4 5 6 281 factors arising from other sources. In many instances, the desire for risk reduction, which was the 7 8 282 prime motivator behind the development of the EMPOWER intervention, did not supersede 9 10 11 283 concerns about symptom recurrence, and other psychological and health factors, as well as 12 13 284 interpersonal relationships with healthcare providers, which played critical contextual roles in the 14 15 285 outcome of the intervention.For peer review only 16 17 18 286 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 287 DISCUSSION 4 5 6 288 7 8 289 This realist evaluation tested the mechanisms embedded in the EMPOWER intervention, 9 10 11 290 and showed that motivation and capacity to deprescribe were triggered in 64% of older chronic 12 13 291 benzodiazepines consumers, supporting the theory that provision of new knowledge about 14 15 292 medication harmsFor can raise peerconcern and augment review patients’ self-efficacy only to deprescribe. However, 16 17 18 293 the analysis also indicated that human motivation to deprescribe is complex and unstable. A 19 20 294 variety of internal and external contexts can interfere along the pathway of deprescribing 21 22 295 interventions. These influences include perceptions about one’s long-term health goals, fear of 23 24 25 296 symptom recurrence, confidence about tapering, psychological attachment to the drug, current 26 27 297 health state and lack of support from health care providers. 28 29 298 Use of a mixed methods approach enabled us to explore the breadth, depth, and 30 31 32 299 complexity of the patient’s experience of deprescribing from a social, behavioural and health

33 http://bmjopen.bmj.com/ 34 300 perspective, allowing stronger inferences about the various contexts affecting patients’ decisions 35 36 44 37 301 than could be achieved through a quantitative or qualitative lens alone. However, other 38 39 302 mechanisms and contexts may trigger motivation to deprescribe beyond what is described in this 40

41 303 realist evaluation. One untested mechanism is provision of information about the lack of drug on September 24, 2021 by guest. Protected copyright. 42 43 44 304 benefits in certain populations, such as statins to reduce cholesterol levels in palliative care 45 46 305 patients with limited life expectancy. 45 46 47 48 306 49 50 51 307 Our findings contribute to the literature by illustrating that linear progression along 52 53 308 different stages of readiness to deprescribe does not fully explain successful deprescribing from 54 55 309 the patient’s perspective. This conclusion is consistent with other critiques of the transtheoretical 56 57 58 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 310 model, which claim that the stages of readiness are arbitrary, that human beings do not make 4 5 6 311 logical and stable plans to change their behavior, and that setbacks can occur along the trajectory 7 8 312 of change47. Education appears to be necessary but insufficient for many individuals, and new 9 10 11 313 strategies will be needed to trigger deprescribing in prohibitive contexts where the EMPOWER 12 13 314 mechanisms failed. As capacity and motivations change over time, reminders and ongoing 14 15 315 discussions aboutFor the risks ofpeer inappropriate review medications may progressively only trigger and sustain 16 17 18 316 patients’ commitments to engage in the deprescribing process. Some competing factors may 19 20 317 wane, such as poor health. Offering cognitive behavioural therapy to patients during the most 21 22 318 difficult last quarter period of the tapering protocol may augment self-efficacy for overcoming 23 24 36 25 319 withdrawal symptoms. Interventions can be directed at health care providers who discourage 26 27 320 deprescribing efforts. Continuing medical education to inform health providers about the 28 29 321 mounting evidence on the harms of benzodiazepine use may curtail the phenomenon of 30 31 20 48 32 322 physicians who continue to promote the use of inappropriate medication. Future research

33 http://bmjopen.bmj.com/ 34 323 directions should also include measurement of cognitive dissonance, which lies at the heart of 35 36 49 37 324 constructivist learning. Methods to measure cognitive dissonance, defined as a feeling of 38 39 325 tension between two sets of competing beliefs and motivations, may shed light on the way in 40

41 326 which tensions about deprescribing are played out and drive behavior change49 50. As we did not on September 24, 2021 by guest. Protected copyright. 42 43 44 327 directly ask patients if they felt internal tension, we were unable to record feelings or processes 45 46 328 of cognitive dissonance. 47 48 329 In conclusion, this realist evaluation of deprescribing from the patient’s perspective 49 50 51 330 increases current understanding about the specific mechanisms and contexts that can be 52 53 331 successful in triggering patients to engage in the deprescribing process. 54 55 332 56 57 58 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 333 ACKNOWLEDGMENTS 4 5 6 334 We wish to acknowledge the work of Anne-Sophie Michaud and Anastasia Soboleva, who 7 8 335 conducted the in-home interviews and helped in identifying first order coding for the qualitative 9 10 11 336 portion of the manuscript. Aditionally, we would like to thank Johanne Collin for her advice in 12 13 337 devising the general strategy for the qualitative interviews. We express gratitude to all the 14 15 338 participants and Forpharmacists peer who took part review in this trial. Particular only thanks are offered to the 16 17 18 339 Pharmacy Services Department of the Jean Coutu Group (PJC) Inc. for their collaboration and 19 20 340 support. 21 22 341 23 24 25 342 Conflict of Interest Checklist: 26 27 Elements of Martin Tannenbaum343 28 Financial/Personal 29 Conflicts 30 Yes No Yes No 31 32 Employment or X X

33 Affiliation http://bmjopen.bmj.com/ 34 35 Grants/Funds X X 36 37 Honoraria X X 38 39 Speaker Forum X X 40

41 on September 24, 2021 by guest. Protected copyright. 42 Consultant X X 43 44 Stocks X X 45 46 Royalties X X 47 48 Expert Testimony X X 49 50 51 Board Member X X 52 53 Patents X X 54 55 Personal Relationship X X 56 344 57 58 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 345 4 5 6 346 7 8 347 Author Contributions: 9 10 11 348 Martin and Tannenbaum contributed to the Study concept and design, Analysis and 12 13 349 Interpretation of data, Drafting of the manuscript, critical revision of the manuscript for 14 15 350 important intellectualFor content, peer and final reviewapproval of the version only to be published. 16 17 18 351 19 20 352 Sponsor: 21 22 353 The study was funded by the Canadian Institutes of Health Research. Philippe Martin received a 23 24 25 354 bursary from the Fonds de Recherche en Santé de Quebec & the Michel Saucier Endowed Chair 26 27 355 in Pharmacology, Health and Aging of the Faculty of Pharmacy of the Université de Montréal. 28 29 356 Cara Tannenbaum is a clinician scientist funded by the Fonds de Recherche en Santé de Quebec. 30 31 32 357 The authors retained full independence from the study sponsors in the design and conduct of the

33 http://bmjopen.bmj.com/ 34 358 study; collection, management, analysis, and interpretation of the data; preparation, review, and 35 36 37 359 approval of the manuscript; and decision to submit the manuscript for publication. 38 39 360 40

41 361 Sponsor’s Role: None on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 19 BMJ Open http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only

Page 19 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 20 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 Figures and Legends: 4 5 6 Figure 1 a: Initial Deprescribing Context-Mechanism-Outcome configuration 7 8 Figure 1b: Refined Deprescribing Context-Mechanism-Outcome configuration 9 10 11 Legend: 12 C = Context (grey circles); 13 M = Mechanism (purple diamonds); 14 O = Outcome (blue rectangles) 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 20

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 REFERENCES 4 5 6 1. Jansen J, Naganathan V, Carter SM, et al. Too much medicine in older people? 7 8 Deprescribing through shared decision making. 2016 9 10 11 2. Scott IA, Hilmer SN, Reeve E, et al. Reducing inappropriate polypharmacy: the 12 13 process of deprescribing. JAMA internal medicine 2015;175(5):827-34. 14 15 3. Frank C, WeirFor E. Deprescribing peer for olderreview patients. CMAJ only: Canadian Medical 16 17 18 Association journal = journal de l'Association medicale canadienne 2014;186(18):1369- 19 20 76. doi: 10.1503/cmaj.131873 [published Online First: 2014/09/04] 21 22 4. Reeve E, To J, Hendrix I, et al. Patient barriers to and enablers of deprescribing: a 23 24 25 systematic review. Drugs & aging 2013;30(10):793-807. doi: 10.1007/s40266-013-0106- 26 27 8 28 29 5. Woodward MC. Deprescribing: achieving better health outcomes for older people 30 31 32 through reducing medications. Journal of Pharmacy Practice and Research

33 http://bmjopen.bmj.com/ 34 2003;33(4):323-28. 35 36 37 6. Scott IA, Gray LC, Martin JH, et al. Deciding when to stop: towards evidence-based 38 39 deprescribing of drugs in older populations. Evidence Based Medicine 2013;18(4):121-24. 40

41 7. Iyer S, Naganathan V, McLachlan AJ, et al. Medication withdrawal trials in people on September 24, 2021 by guest. Protected copyright. 42 43 44 aged 65 years and older: a systematic review. Drugs & aging 2008;25(12):1021-31. doi: 45 46 10.2165/0002512-200825120-00004 [published Online First: 2008/11/22] 47 48 8. van der Cammen TJ, Rajkumar C, Onder G, et al. Drug cessation in complex older 49 50 51 adults: time for action. Age and ageing 2014;43(1):20-5. doi: 10.1093/ageing/aft166 52 53 [published Online First: 2013/11/14] 54 55 56 57 58 59 60 21

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 9. Declercq T, Petrovic M, Azermai M, et al. Withdrawal versus continuation of chronic 4 5 6 antipsychotic drugs for behavioural and psychological symptoms in older people with 7 8 dementia. Cochrane database of systematic reviews (Online) 2013(3):CD007726. doi: 9 10 11 10.1002/14651858.CD007726.pub2 [published Online First: 2013/04/02] 12 13 10. Martin P, Tamblyn R, Ahmed S, et al. A drug education tool developed for older 14 15 adults changesFor knowledge, peer beliefs and riskreview perceptions about only inappropriate 16 17 18 benzodiazepine prescriptions in the elderly. Patient education and counseling 2013 doi: 19 20 10.1016/j.pec.2013.02.016 [published Online First: 2013/04/02] 21 22 11. Ostini R, Hegney D, Jackson C, et al. Systematic review of interventions to improve 23 24 25 prescribing. The Annals of pharmacotherapy 2009;43(3):502-13. doi: 10.1345/aph.1L488 26 27 [published Online First: 2009/03/06] 28 29 12. Tannenbaum C, Martin P, Tamblyn R, et al. Reduction of inappropriate 30 31 32 benzodiazepine prescriptions among older adults through direct patient education: the

33 http://bmjopen.bmj.com/ 34 EMPOWER cluster randomized trial. JAMA internal medicine 2014;174(6):890-8. doi: 35 36 37 10.1001/jamainternmed.2014.949 38 39 13. Cook JM, Biyanova T, Masci C, et al. Older patient perspectives on long-term 40

41 anxiolytic benzodiazepine use and discontinuation: a qualitative study. Journal of general on September 24, 2021 by guest. Protected copyright. 42 43 44 internal medicine 2007;22(8):1094-100. doi: 10.1007/s11606-007-0205-5 [published 45 46 Online First: 2007/05/12] 47 48 14. Reeve E, Wiese MD, Hendrix I, et al. People's attitudes, beliefs, and experiences 49 50 51 regarding polypharmacy and willingness to Deprescribe. Journal of the American 52 53 Geriatrics Society 2013;61(9):1508-14. doi: 10.1111/jgs.12418 54 55 56 57 58 59 60 22

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 15. Pollmann AS, Murphy AL, Bergman JC, et al. Deprescribing benzodiazepines and Z- 4 5 6 drugs in community-dwelling adults: a scoping review. BMC pharmacology & toxicology 7 8 2015;16:19. doi: 10.1186/s40360-015-0019-8 [published Online First: 2015/07/05] 9 10 11 16. Wong G, Westhorp G, Manzano A, et al. RAMESES II reporting standards for realist 12 13 evaluations. BMC Med 2016;14(1):96. doi: 10.1186/s12916-016-0643-1 [published 14 15 Online First:For 2016/06/28] peer review only 16 17 18 17. Johansson T, Abuzahra ME, Keller S, et al. Impact of strategies to reduce 19 20 polypharmacy on clinically relevant endpoints: a systematic review and meta-analysis. 21 22 British journal of clinical pharmacology 2016;82(2):532-48. doi: 10.1111/bcp.12959 23 24 25 18. Page AT, Clifford RM, Potter K, et al. The feasibility and effect of deprescribing in 26 27 older adults on mortality and health: a systematic review and meta-analysis. British 28 29 journal of clinical pharmacology 2016;82(3):583-623. doi: 10.1111/bcp.12975 30 31 32 19. Billioti de Gage S, Moride Y, Ducruet T, et al. Benzodiazepine use and risk of

33 http://bmjopen.bmj.com/ 34 Alzheimer's disease: case-control study. Bmj 2014;349:g5205. doi: 10.1136/bmj.g5205 35 36 37 [published Online First: 2014/09/12] 38 39 20. AGS 2015 Beers Criteria Update Expert Panel. American Geriatrics Society 2015 40

41 Updated Beers Criteria for Potentially Inappropriate Medication Use in Older Adults. on September 24, 2021 by guest. Protected copyright. 42 43 44 Journal of the American Geriatrics Society 2015;63(11):2227-46. doi: 10.1111/jgs.13702 45 46 [published Online First: 2015/10/09] 47 48 21. Finkle WD, Der JS, Greenland S, et al. Risk of fractures requiring hospitalization 49 50 51 after an initial prescription for zolpidem, alprazolam, lorazepam, or diazepam in older 52 53 adults. Journal of the American Geriatrics Society 2011;59(10):1883-90. doi: 54 55 10.1111/j.1532-5415.2011.03591.x 56 57 58 59 60 23

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 22. Paterniti S, Dufouil C, Alperovitch A. Long-term benzodiazepine use and cognitive 4 5 6 decline in the elderly: the Epidemiology of Vascular Aging Study. Journal of clinical 7 8 psychopharmacology 2002;22(3):285-93. [published Online First: 2002/05/15] 9 10 11 23. Allain H, Bentue-Ferrer D, Polard E, et al. Postural instability and consequent falls 12 13 and hip fractures associated with use of hypnotics in the elderly: a comparative review. 14 15 Drugs & agingFor 2005;22(9):749-65. peer [published review Online First: only2005/09/15] 16 17 18 24. Cotroneo A, Gareri P, Nicoletti N, et al. Effectiveness and safety of hypnotic drugs in 19 20 the treatment of insomnia in over 70-year old people. Archives of gerontology and 21 22 geriatrics 2007;44 Suppl 1:121-4. doi: 10.1016/j.archger.2007.01.018 [published Online 23 24 25 First: 2007/02/24] 26 27 25. Tannenbaum C. Effect of age, education and health status on community dwelling 28 29 older men's health concerns. The aging male : the official journal of the International 30 31 32 Society for the Study of the Aging Male 2012;15(2):103-8. doi:

33 http://bmjopen.bmj.com/ 34 10.3109/13685538.2011.626819 [published Online First: 2011/11/23] 35 36 37 26. Tannenbaum C, Mayo N, Ducharme F. Older women's health priorities and 38 39 perceptions of care delivery: results of the WOW health survey. CMAJ : Canadian 40

41 Medical Association journal = journal de l'Association medicale canadienne on September 24, 2021 by guest. Protected copyright. 42 43 44 2005;173(2):153-9. doi: 10.1503/cmaj.050059 45 46 27. Anderson K, Stowasser D, Freeman C, et al. Prescriber barriers and enablers to 47 48 minimising potentially inappropriate medications in adults: a systematic review and 49 50 51 thematic synthesis. BMJ open 2014;4(12):e006544. 52 53 28. Pawson R. The science of evaluation: a realist manifesto: Sage 2013. 54 55 56 57 58 59 60 24

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 29. Martin P, Tamblyn R, Ahmed S, et al. An educational intervention to reduce the use 4 5 6 of potentially inappropriate medications among older adults (EMPOWER study): 7 8 protocol for a cluster randomized trial. Trials 2013;14:80. doi: 10.1186/1745-6215-14-80 9 10 11 [published Online First: 2013/03/22] 12 13 30. Tannenbaum C, Martin P. Sedative-Hypnotic Medication De-Prescribing Brochure, 14 15 Available at:For http://www.criugm.qc.ca/fichier/pdf/BENZOeng.pdf peer review only. In: Montréal 16 17 18 Iudgd, ed., 2014. 19 20 31. Hein G. Constructivist learning theory. Institute for Inquiry Available 21 22 at:/http://www/ exploratorium edu/ifi/resources/constructivistlearning htmlS 1991 23 24 25 32. Festinger L. A theory of social comparison processes. Human relations 26 27 1954;7(2):117-40. 28 29 30 33. Michie S, van Stralen MM, West R. The behaviour change wheel: a new method for 31 32 characterising and designing behaviour change interventions. Implementation science : IS

33 http://bmjopen.bmj.com/ 34 2011;6:42. doi: 10.1186/1748-5908-6-42 [published Online First: 2011/04/26] 35 36 37 34. JW C, V P-C. Designing and Con- ducting Mixed Methods Research. Thousand Oaks, 38 39 CA: Sage Publications 2007. 40

41 35. Horne R, Weinman J, Hankins M. The Beliefs about Medicines Questionnaire: The on September 24, 2021 by guest. Protected copyright. 42 43 44 development and evaluation of a new method for assessing the cognitive representation 45 46 of medication. Psychology & Health 1999;14(1):1-24. doi: 47 48 49 http://dx.doi.org/10.1080/08870449908407311 50 51 36. Belanger L, Morin CM, Bastien C, et al. Self-efficacy and compliance with 52 53 benzodiazepine taper in older adults with chronic insomnia. Health psychology : official 54 55 56 57 58 59 60 25

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 journal of the Division of Health Psychology, American Psychological Association 4 5 6 2005;24(3):281-7. doi: 10.1037/0278-6133.24.3.281 [published Online First: 2005/05/19] 7 8 37. McDowell I. Measuring health: A guide to rating scales and questionnaires (3rd ed.). 9 10 11 New York, NY: Oxford University Press; US 2006:xvi, 748. 12 13 38. Sandelowski M. Combining qualitative and quantitative sampling, data collection, 14 15 and analysis Fortechniques peerin mixed-method review studies. Research inonly nursing & health 16 17 18 2000;23(3):246-55. [published Online First: 2000/06/28] 19 20 39. Ellis P. The Essential Guide to Effect Sizes: Statistical Power, Meta-Analysis, 21 22 and the Interpretation of Research Results. Cambridge: Cambridge University Press 2010. 23 24 25 40. Braun V, Clarke V. Using thematic analysis in psychology. Qualitative Research in 26 27 Psychology 2006;3(2):77-101. 28 29 41. O'Cathain A, Murphy E, Nicholl J. Three techniques for integrating data in mixed 30 31 32 methods studies. British Medical Journal 2010;341:c4587. doi: 10.1136/bmj.c4587

33 http://bmjopen.bmj.com/ 34 42. Farmer T, Robinson K, Elliott SJ, et al. Developing and implementing a triangulation 35 36 37 protocol for qualitative health research. Qualitative health research 2006;16(3):377-94. 38 39 doi: 10.1177/1049732305285708 40

41 43. Prochaska J. Transtheoretical Model of Behavior Change. In: Gellman M, Turner JR, on September 24, 2021 by guest. Protected copyright. 42 43 44 eds. Encyclopedia of Behavioral Medicine: Springer New York 2013:1997-2000. 45 46 44. Johnson B, Christensen LB. Educational research : quantitative, qualitative, and 47 48 mixed approaches. Fifth ed2014. 49 50 51 45. Hilmer SN, Gnjidic D, Le Couteur DG. Thinking through the medication list: 52 53 Appropriate prescribing and deprescribing in robust and frail older patients. Australian 54 55 family physician 2012;41(12):924. 56 57 58 59 60 26

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 46. Holmes HM, Todd A. Evidence-based deprescribing of statins in patients with 4 5 6 advanced illness. JAMA internal medicine 2015;175(5):701-02. 7 8 47. West R. Time for a change: putting the Transtheoretical (Stages of Change) Model to 9 10 11 rest. Addiction 2005;100(8):1036-9. doi: 10.1111/j.1360-0443.2005.01139.x [published 12 13 Online First: 2005/07/27] 14 15 48. Cassel CK,For Guest JA. peer Choosing wisely: review helping physicians only and patients make smart 16 17 18 decisions about their care. JAMA : the journal of the American Medical Association 19 20 2012;307(17):1801-2. doi: 10.1001/jama.2012.476 [published Online First: 2012/04/12] 21 22 49. Theory-driven design strategies for technologies that support behavior change in 23 24 25 everyday life. Proceedings of the SIGCHI Conference on Human Factors in Computing 26 27 Systems; 2009. ACM. 28 29 50. Bem DJ. Self-Perception: An Alternative Interpretation of Cognitive Dissonance 30 31 32 Phenomena. Psychological Review 1967;74(3):183-200.

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 27

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from Page 28 of 37 28 The printed format of the 8- the of format printed The knowledge effective and take to piece transfer ahealthcare to show Provision of an easy-to-use easy-to-use an of Provision tapering week 16-20 visual a take to when showing tool pill, quarter or half whole, dose the toskip when and completely Information about changes changes about Information with metabolism indrug a higher to lead can that age meant effects, side of risk elicit and beliefs tochange of safety the about concern older in themedication adults page brochure makes it an it makes brochure page provider

Logos on the brochure brochure the on Logos the for credibility conversations An inspirational story story inspirational An social using peer and comparison to championing self-efficacy increase tapering for Interactive knowledge knowledge Interactive true/false 4 with test answers and questions of harms the about increasing at aimed knowledge provide source source provide initiate to patient benzodiazepines, benzodiazepines, BMJ Open http://bmjopen.bmj.com/ Components of the EMPOWER brochure brochure EMPOWER the of Components on September 24, 2021 by guest. Protected copyright. Instruction to “Please consult your your consult “Please to Instruction before pharmacist or doctor a in medication” any stopping box red large A list of alternative non- alternative list of A can patients that anxiety and sleep substitutes useas Messaging on the front page page front the on Messaging raise to Risk” beat May “You of theharms of awareness pharmacological approaches to approaches pharmacological benzodiazepines benzodiazepines For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only Linking the components of the EMPOWER intervention to the underlying program theory and mechanisms and theory mechanisms underlying program the to intervention EMPOWER of the components Linkingthe Program Theory/Mechanism Theory/Mechanism Program TABLES TABLES Table1:

Drive opportunities to discuss and and to discuss opportunities Drive healthcare a with deprescribing initiate provider Increase capacity to taper by by taper to capacity Increase self-efficacy augmenting Increase motivation to deprescribe by by to deprescribe motivation Increase beliefs and knowledge changing

Tab 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 29

* *

* * *

2) 2) .54 .02 .01 .01 (95% CI) CI) (95% P-value/ Risk difference difference Risk P-value/ Successful completion of completion Successful deprescribing (Groups 1 vs 1 vs (Groups deprescribing -4.6 [-19.1-10.4] * -5.0 [-19.8-10.0] 0 [-12.6-13.3] -2.6 [-15.0-10.4] 18.9[5.3-32.0] 13.6[1.6-25.9] -2.2 [-15.9-11.3]

* * * * * * *

* *

.23 .10 .00 .03 (95% CI) CI) (95% (Groups 1+2 vs 3) vs 1+2 (Groups Initiating deprescribing deprescribing Initiating P-value/ Risk difference difference Risk P-value/ (n=94) (n=94) Group 3- 3- Group deprescribe deprescribe No attempt to attempt No Failed Failed g (n=75) g (n=75) Group 2- 2- Group deprescribin BMJ Open http://bmjopen.bmj.com/

(n=92) (n=92) Group 1- 1- Group Successful Successful deprescribing

on September 24, 2021 by guest. Protected copyright. .85 (.99) .97 (1.08) .87 (.97) .71 (.90) 56 (20.1) 25 (27.1) 22 (28.9) 9 (9.7) 18.6[8.7-27.1] All (n=261) All (n=261) 13.8 (3.4) 13.0(.3) 14.3(.4) 14.1(.4) 13.4 (2.7) 13.4(.3) 14.1(.3) 12.9(.3) 103 (39.5) 42 (45.6) 38 (50.6) 23 (25.0) 23.4[11.3-34.1] 138 (52.8) 75 (81.5) 47 (62.7) 16 (17.4) 56.0[44.6-64.8] 118 (44.8) 51 (55.4) 45 (60.0) 21 (22.3) 35.1[23.1%-45.4] 145 (55.5) 70 (76.1) 57 (76.0) 18(19.1) 56.9[45.4-65.8] 138 (52.8) 70 (76.1) 59 (78.7) 9 (9.7) 67.7[57.3-74.9] 156 (59.8) 80 (86.9) 55 (73.3) 21 (22.3) 58.5[47.0-67.4] 37.8 (35.7) 47.3(34.6) 35.0(37.4) 31.0(33.6) For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml For peer review only

, (%) n a (/25), (SD)Mean a , (SD)Mean a Changes in knowledge, beliefs, and self-efficacy to taper among those with successful versus failed deprescribing efforts deprescribing failed versus successful with those among taper to self-efficacy and beliefs, inknowledge, Changes

Characteristic and Outcome

Table 2:Table

Discussed Discussed pharmacist, with n (%) Discussed Discussed physician, with (%) n Outreach to aprofessional: health Perception (yes, increased of risk no),(%) n Risk Perception:Risk Increased post-intervention Self-Efficacy, (%) n Baseline self-efficacyBaseline (/100), (SD)Mean Self-Efficacy for Tapering Increased (%)concern, n Decreased necessity Baseline concernBaseline Baseline necessityBaseline Beliefs about benzodiazepines Increased knowledge Post-Intervention, (%) n Baseline KnowledgeBaseline (/4), (SD)Mean Knowledge: As some participantsAs selected than more one condition, total does not equal 100%. Level of significance, p < 0.05.

Independent sample t-test for continuous variables, chi square for categorical variables. * ‡ Page 29 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from Page 30 of 37 30

[84 ” [84

[79 y.o. [79successful y.o. man, [68 y.o. [68 y.o. successful woman, [72 y.o. [72successful y.o. taper]man,

[72 y.o. woman, successful taper][72 y.o. woman, [84 y.o. man, [84successful taper] y.o. man,

Supporting Supporting citation

[87 y.o. successful woman, taper] “When I told my doctor I towanted hestop, said, problem, let’s “no do it.” “In the past I pill tried toallthe atstop once.But using the tapering tool, “At “At my age I believe suchdon’tmiracles inas being able sleep for to 8, 9 or 10 each hours night. It bewould impossible for me, so I content myself thewith hours of sleep I get.” “I understood could stopI taking it [after I the read brochure], that it notwas an obligationit].”[to take “My physician“My told me it [thedrugs] could cost me my memory. My memory has become very important to me.” taper] y.o. man, successfuly.o. taper]man, “I persuaded that myself neededget I to rid of this, no matter what. “I don’thave Ias much pain usedas It’s to.now under socontrol it was for easier me to stop. Beforeno - way.” taper] “He [my doctor] notdrug was me told the good for me and Ithat could experience side while takingeffects it.” y.o. man, successfuly.o. taper]man, I understood that it need to a be gradual drastic and process.” anot

BMJ Open http://bmjopen.bmj.com/

(n=7) (n=7) 1 (14%) deprescribe No attempt toNo attempt

(n=7) (n=7) Failed Failed on September 24, 2021 by guest. Protected copyright. 4 (57%) deprescribing deprescribing For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

(n=7) (n=7) 5 (71%) 5 (71%) 3 (43%) 0 3 (43%) 1 (14%) 0 5 (71%) 3 (43%) 1 (14%) 6 (86%) 5 (71%) 1 (14%) 6 (86%) 4 (57%) 1 (14%) 5 (71%) 4 (57%) 2 (29%) Successful

deprescribing deprescribing For peer review only Key theme

Table 3: Contexts that enabled deprescribing mechanisms to work to enabled that Table Contexts 3: deprescribing mechanisms Supportive health care provider 3 (43%) 2 (29%) 0 Tapering Tapering tool provides support Preventive perspectivePreventive on active aging and healthy Lack of psychologicalLack of attachment Perception increased of risk Certainty andCertainty confidence about (post-intervention)tapering Stable health status Previous support physician/ from discontinuation

positive attitudepositive towards 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 31

[85 y.o. [85failed y.o. man,

[68 toy.o. no intent woman, taper] Supporting Supporting citation .” [85 y.o. woman, y.o..” [85 tapering]failed woman, [70 tono intent taper] y.o. woman, [72 y.o. man, [72 tono intent y.o. taper]man, [85 y.o. man, [85failed man, tapering] y.o. [85 y.o. [85 y.o. failedwoman, tapering] [72 tono intent y.o. taper]man,

“My doctor“My told me: age, “At your don’t about worry it. You’ve been taking this pill for a while and taking you Youare fine. aren’t a dangerousat dose all

“If “If you take all pills asyou’ll your prescribed, never haveproblems in your […]life something prescribes my When doctor for me, I know notit’s junk, I goodit’s know me. And Ifor don’t question it”. “If “If anyonepills, poof, stops Imy would for becausesure die of my “ I“ asked him doctor],[my there“Are any ofI my medications could stop?” He me,told “No, notwe’re takinganything you away, are doing I well”. then told my medication him getting very was expensive to which he “Youreplied, know priceless”. is Mr., life [75 tono intent y.o. taper]man,

“At “At my age I don’tabout care Ithe don’trisks. careto Iif 100 live or not.” “I recall that he doctor][my me told that the in long-term my benzodiazepineaffect mycould memory. But my memory is

“I knew that I’dbewithout in my pills. trouble It’s been a long time now. How canput I If it in I words? ran ofout I’d pills be in trouble.” tapering] “When I decreased started thegetting I dose headaches. I felt miserable not being atable to night.” sleep “Without this medication, I that know my wouldlife be plagued by anxiety, ofam this Icertain.”” poor health.” ”. fantastic.” BMJ Open http://bmjopen.bmj.com/

to (n=7) (n=7) 5 (71%) No attempt No attempt deprescribe

(n=7) (n=7) Failed Failed on September 24, 2021 by guest. Protected copyright. 1 (14%) deprescribing deprescribing For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

0 1 (14%) 4 (57%) 0 4 (57%) 4 (57%) 0 5 (71%) _ (n=7) (n=7) 1 (14%) 1 (14%) 1 (14%) 4 (57%) Successful

deprescribing deprescribing For peer review only

Key theme

Loss of Loss confidence to complete the processtapering (post-intervention) Intolerance to recurrence of symptoms/withdrawal effects Discouragement a physicianfrom 1 (14%) 3 (43%) 5 (71%) Quality of lifeof Quality during focus end-of life 0 2 (29%) 3 (43%) Reliance on for medication coping/everyday function Lack of perceptionLack of personalof risk 1 (14%) 2 (29%) 5 (71%) Unquestioning inbelief their physician 1 (14%) 1 (14%) 3 (43%) Poor health Poorstatus health Previous discouragement from physician / Negative attitude physicianof towards deprescribing Table 4: Contexts where deprescribing mechanisms failed mechanisms deprescribing where Contexts Table 4: Page 31 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from

BMJ Open Page 32 of 37

1 2 Figure 1a: Initial Deprescribing Context-Mechanism-Outcome configuration 3 4 5 6 7 8 9 M2/3 = 10 M1 = Increase Increase 11 ForMotivation peer review only 12 Capacity + 13 Create 14 Opportunity 15 16 17 http://bmjopen.bmj.com/ 18 19 20 Stage 3: Outcome= Successful 21 Stage 2: Action - Maintenance Stage 1: Pre- deprescribing over a 6-month 22 Initiate (Complete Contemplation 23 Deprescribing deprescribing period 24 protocol) 25 26 on September 24, 2021 by guest. Protected copyright. 27 28 29 30 31 C1 = Lack of 32 C2 = 33 knowledge of 34 the harms Supportive 35 associated with environment 36 medication use 37 38 39 40 41 42 43 Legend: C = Context (grey circles); M = Mechanism ( purple diamonds); O = Outcome (Blue rectangles) 44 45 46 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from

Page 33 of 37 BMJ Open Figure 1b: Refined Deprescribing Context-Mechanism-Outcome configuration

1 2 3 4 5 C2 = Lack of 6 psychological 7 attachment; 8 few/weak C3 = Health 9 competing care provider 10 factors 11 For peer review onlyencourages 12 deprescribing 13 C1 = Lack of 14 knowledge of M2 = Patient 15 the harms 16 empowerment O3= Successful 17 associated with and increase in M1 = http://bmjopen.bmj.com/ completion of 18 medication use Motivation self-efficacy deprescribing 19 20 process 21 22 23 C4 = Health 24 care provider 6-months 25 Pre-intervention post-intervention discourages on September 24, 2021 by guest. Protected copyright. 26 deprescribing 27 28 29 O1 = No initiation O2 = Initiation of 30 Intervention of 31 deprescribing 32 deprescribing 33 process 34 35 O4= Failure of 36 37 deprescribing 38 C4 = Competing factors process 39 and desires preclude C5 = Consumer 40 41 assimilation of the new fearful or information and/or do cannot tolerate 42 Legend: 43 not generate a recurrence C = Context (grey circles); 44 perception of increased of symptoms M = Mechanism (purple diamonds); 45 personal risk O = Outcome (Blue rectangles) 46 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 34 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 In the title, identify the document as a 9 1 Y 1 10 realist evaluation 11 SUMMARY OR ABSTRACT 12 13 Journal articles will usually require an 14 abstract, while reports and other forms of 15 For peer review only 16 publication will usually benefit from a 17 short summary. The abstract or summary 18 should include brief details on: the policy, 19 programme or initiative under evaluation; 20 programme setting; purpose of the 21 evaluation; evaluation question(s) and/or 22 23 objective(s); evaluation strategy; data 24 collection, documentation and analysis 25 methods; key findings and conclusions 26 2 Y 3-4 27 Where journals require it and the nature of 28 29 the study is appropriate, brief details of 30 respondents to the evaluation and 31 recruitment and sampling processes may 32 also be included 33 http://bmjopen.bmj.com/ 34 35 Sufficient detail should be provided to 36 identify that a realist approach was used 37 and that realist programme theory was 38 developed and/or refined 39 INTRODUCTION 40

41 Rationale for Explain the purpose of the evaluation and on September 24, 2021 by guest. Protected copyright. 3 Y 5 42 evaluation the implications for its focus and design 43 44 Describe the initial programme theory (or 45 4 Programme theory theories) that underpin the programme, Y 6 46 policy or initiative 47 48 State the evaluation question(s) and specify 49 the objectives for the evaluation. Describe Evaluation questions, 50 5 whether and how the programme theory Y 6 objectives and focus 51 was used to define the scope and focus of 52 53 the evaluation 54 State whether the realist evaluation 55 6 Ethical approval required and has gained ethical approval Y 7 56 57 from the relevant authorities, providing 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 9 details as appropriate. If ethical approval 10 was deemed unnecessary, explain why 11 METHODS 12 13 Rationale for using Explain why a realist evaluation approach 7 Y 7 14 realist evaluation was chosen and (if relevant) adapted 15 For peer review only 16 Environment Describe the environment in which the 17 8 surrounding the Y 7 evaluation took place 18 evaluation 19 20 Describe the 21 programme policy, Provide relevant details on the programme, 9 Y 8 22 initiative or product policy or initiative evaluated 23 evaluated 24 25 A description and justification of the 26 evaluation design (i.e. the account of what 27 was planned, done and why) should be 28 29 included, at least in summary form or as an 30 appendix, in the document which presents Describe and justify the main findings. If this is not done, the 31 10 Y 8-9 32 the evaluation design omission should be justified and a 33 reference or link to the evaluation design http://bmjopen.bmj.com/ 34 given. It may also be useful to publish or 35 36 make freely available (e.g. online on a 37 website) any original evaluation design 38 document or protocol, where they exist 39 40 Describe and justify the data collection

41 methods – which ones were used, why and on September 24, 2021 by guest. Protected copyright. 42 how they fed into developing, supporting, 43 Data collection refuting or refining programme theory 11 Y 9-10 44 methods 45 46 Provide details of the steps taken to 47 enhance the trustworthiness of data 48 collection and documentation 49 Describe how respondents to the evaluation 50 51 Recruitment process were recruited or engaged and how the 52 12 and sampling sample contributed to the development, Y 7, 9-10 53 strategy support, refutation or refinement of 54 programme theory 55 Describe in detail how data were analysed. 56 13 Data analysis Y 10-11 57 This section should include information on 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 37 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 9 the constructs that were identified, the 10 process of analysis, how the programme 11 theory was further developed, supported, 12 refuted and refined, and (where relevant) 13 14 how analysis changed as the evaluation 15 Forunfolded peer review only 16 RESULTS 17 18 Report (if applicable) who took part in the 19 evaluation, the details of the data they Details of 20 14 provided and how the data was used to Y 12-13 participants 21 develop, support, refute or refine 22 23 programme theory 24 Present the key findings, linking them to 25 contexts, mechanisms and outcome 26 27 15 Main findings configurations. Show how they were used Y 12-14 28 to further develop, test or refine the 29 programme theory 30 DISCUSSION 31 32 Summarise the main findings with attention 33 to the evaluation questions, purpose of the http://bmjopen.bmj.com/ 16 Summary of findings Y 15-16 34 evaluation, programme theory and intended 35 36 audience 37 Discuss both the strengths of the evaluation 38 and its limitations. These should include 39 40 (but need not be limited to): (1)

41 consideration of all the steps in the on September 24, 2021 by guest. Protected copyright. 42 evaluation processes; and (2) comment on 43 the adequacy, trustworthiness and value of 44 the explanatory insights which emerged Strengths, limitations 45 17 Y 15-16 46 and future directions 47 In many evaluations, there will be an 48 expectation to provide guidance on future 49 directions for the programme, policy or 50 initiative, its implementation and/or design. 51 The particular implications arising from the 52 53 realist nature of the findings should be 54 reflected in these discussions 55 Comparison with Where appropriate, compare and contrast 56 18 Y 16 57 existing literature the evaluation’s findings with the existing 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 37 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 9 literature on similar programmes, policies 10 or initiatives 11 12 List the main conclusions that are justified Conclusion and by the analyses of the data. If appropriate, 13 19 Y 16 14 recommendations offer recommendations consistent with a 15 Forrealist peer approach review only 16 17 State the funding source (if any) for the Funding and conflict evaluation, the role played by the funder (if 18 20 Y 2,17 19 of interest any) and any conflicts of interests of the 20 evaluators 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

A realist evaluation of patients' decisions to deprescribe in the EMPOWER trial

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-015959.R1

Article Type: Research

Date Submitted by the Author: 21-Feb-2017

Complete List of Authors: Martin, Philippe; Centre de recherche de l'Institut universitaire de gériatrie de Montréal, Tannenbaum, Cara; Université de Montréal,

Primary Subject Geriatric medicine Heading:

Secondary Subject Heading: Health services research, Patient-centred medicine, Addiction

Keywords: deprescribing, benzodiazepines, realist evaluation, mechanisms, EMPOWER

http://bmjopen.bmj.com/

on September 24, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 1 A realist evaluation of patients' decisions to deprescribe in the EMPOWER trial 4 5 a,b b,c 6 2 Philippe Martin, BSc , Cara Tannenbaum, MD MSc 7 8 9 3 10 11 a 12 4 Doctoral Student, Faculty of Pharmacy, Université de Montréal, Montreal, PQ Canada 13 14 15 b For peer review only 16 5 Institut Universitaire de Gériatrie de Montréal, Montreal, PQ Canada 17 18 c 19 6 Professor and Michel Saucier Endowed Chair, Université de Montréal, Faculties of Medicine 20 21 22 7 and Pharmacy, Montreal, PQ, Canada. 23 24 25 8 26 27 9 Corresponding Author: 28 29 10 Philippe Martin, B. Sc. 30 31 32 11 Institut Universitaire de Gériatrie de Montréal,

33 http://bmjopen.bmj.com/ 34 12 Faculties of Pharmacy and Medicine, 35 36 37 13 Université de Montréal, Montréal, Quebec, Canada 38 39 14 4545 Queen Mary Road 40

41 15 Montreal, QC H3W 1W5 on September 24, 2021 by guest. Protected copyright. 42 43 44 16 E-mail : [email protected] 45 46 17 Tel: 514-340-3540 ext: 4728 47 48 18 Fax: 514-340-2117 49 50 51 19 Alternate Corresponding Author: [email protected] 52 53 20 Word Count: Abstract (262 words), Main text (2877 words) 54 55 21 56 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 22 Funding statement: This work was supported by Operating Grant OTG-88591 from the 4 5 6 23 Canadian Institutes of Health Research (CIHR). Philippe Martin received a doctoral bursary 7 8 24 from the FRQS. Cara Tannenbaum was supported by a Senior Scientist Career Award from the 9 10 11 25 FRQS. The above funding organizations had no role in the design and conduct of the study; 12 13 26 collection, management, analysis, and interpretation of the data; preparation, review, or approval 14 15 27 of the manuscript;For or the decision peer to submit review the manuscript for publication.only 16 17 18 28 19 20 29 Data sharing statement: Dataset is available upon request to the corresponding author. 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 30 STRUCTURED ABSTRACT: 4 5 6 31 BACKGROUND AND OBJECTIVES: Successful mechanisms for engaging patients in the 7 8 32 deprescribing process remain unknown. We sought to determine in which contexts the 9 10 11 33 EMPOWER deprescribing intervention triggered older adults’: (1) motivation to deprescribe by 12 13 34 increasing knowledge and concern about benzodiazepines; (2) capacity to taper by augmenting 14 15 35 self-efficacy; andFor (3) opportunities peer to discuss review and receive support only from a healthcare provider to 16 17 18 36 deprescribe. 19 20 37 DESIGN: A realist evaluation using a sequential mixed methods approach, conducted alongside 21 22 38 the EMPOWER randomized clinical trial 23 24 25 39 SETTING: Community, Quebec, Canada. 26 27 40 PARTICIPANTS: 261 older chronic benzodiazepine consumers, who received the EMPOWER 28 29 41 intervention and had complete 6 month follow up data. 30 31 32 42 INTERVENTION: Mailed deprescribing brochure on benzodiazepines.

41 46 from a physician or pharmacist) on September 24, 2021 by guest. Protected copyright. 42 43 44 47 RESULTS: The intervention triggered an intent to deprescribe among 167 (n=64%) participants 45 46 48 (mean age 74.6 years + 6.3, 72% women). Participants intending to deprescribe had improved 47 48 49 knowledge (risk difference, 58.50% [95% CI, 46.98%-67.44%]), increased concern about taking 49 50 51 50 benzodiazepines (risk difference, 67.67% [95% CI, 57.36%-74.91%]), and higher self-efficacy 52 53 51 for tapering (risk difference, 56.90% [95% CI, 45.41%-65.77%]). Contexts where the 54 55 52 deprescribing mechanisms failed included lack of support from a health care provider, 56 57 58 59 60 3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 53 intolerance to withdrawal symptoms, perceived poor health, and psychological reliance on 4 5 6 54 benzodiazepines. 7 8 55 CONCLUSION: Deprescribing interventions that systematically target patient motivation and 9 10 11 56 capacity to deprescribe are successful to the extent that healthcare providers are supportive, and 12 13 57 that patients do not have internal competing motivations to remain on drug therapy. 14 15 58 For peer review only 16 17 18 59 Key words: deprescribing, benzodiazepines, realist evaluation, mechanisms, EMPOWER 19 20 60 ClinicalTrials.gov identifier is NCT01148186 21 61 22 62 23 24 25 63 26 64 ARTICLE SUMMARY: 27 65 28 66 STRENGTHS AND LIMITATIONS OF THE STUDY: 29 30 67 • Use of a mixed methods approach enabled us to explore the breadth, depth, and 31 68 complexity of the patient’s experience of deprescribing. 32 69 • Use of the realist evaluation allowed us to investigate how the mechanisms underlying 33 70 deprescribing interventions interact with specific contexts to yield positive or negative http://bmjopen.bmj.com/ 34 71 outcomes 35 36 72 • This study was conducted alongside a large cluster randomized clinical trial. 37 73 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 74 INTRODUCTION 4 5 6 75 Deprescribing refers to the collaborative process of tapering, discontinuing, stopping, or 7 8 76 withdrawing medications in order to reduce adverse drug events and improve outcomes.1-5 9 10 1 3 6 11 77 Deprescribing has many steps , with one key component being the engagement of patients in 12 1 7-15 13 78 shared decision-making. Research suggests that older adults have conflicted feelings about 14 15 79 medications4 14: For78% of older peer adults believe review that medications onlyare necessary to improve health, 16 17 18 80 but at the same time, 68% would like to reduce their current medication use, with 92% willing to 19 20 81 stop a regular medication if advised to do so by their physician14. 21 22 82 23 24 25 83 A better understanding of the mechanisms that trigger patient motivation and capacity to 26 27 84 engage in the deprescribing process could reduce the use of potentially inappropriate medications 28 29 85 in older adults. The aim of realist evaluation is to reveal how an intervention might generate 30 31 32 86 different outcomes in different circumstances, and how different mechanisms work in particular

41 90 for whom, under which circumstances and why, are questions that have never been explored on September 24, 2021 by guest. Protected copyright. 42 43 44 91 systematically from the patient’s point of view. Recent reviews on deprescribing call for a 45 46 92 realistic evaluation of large deprescribing trials to investigate how the mechanisms underlying 47 48 93 deprescribing interventions interact with specific contexts to yield positive or negative 49 50 17 18 51 94 outcomes. The EMPOWER trial, which demonstrated a number-needed-to-treat of 4 for the 52 53 95 effectiveness of mailing a benzodiazepine deprescribing brochure on complete cessation of 54 55 56 57 58 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 96 benzodiazepines at 6 months, provides a timely opportunity to examine which deprescribing 4 5 12 6 97 mechanisms worked for different patients under different circumstances. 7 8 98 9 10 11 99 The initial theory underpinning the development of the EMPOWER intervention was 12 13 100 that most – if not all - older adults are unaware of the age-related harms of taking benzodiazepine 14 15 101 anti-anxiety drugsFor and sleeping peer pills. Side review effects of sedative-hypnotics only are well-documented in 16 17 18 102 the literature but rarely talked about in practice as being a potential cause of memory impairment, 19 20 103 falls and fractures19-24, feared by many older adults25 26. Not understanding why medications 21 22 104 should be discontinued is a patient barrier to deprescribing4 27. As most patients are uninformed 23 24 25 105 of the potential risks associated with the use of benzodiazepines, we hypothesized a linear 26 27 106 behavior change process whereby providing patients with an interactive educational brochure 28 29 107 detailing associated risks, safer alternatives, and steps for tapering, would trigger patients’ 30 31 32 108 motivation, capacity and opportunity to initiate the deprescribing process through discussion of

41 112 perspective. The realist evaluation answers the following three research questions: (1) Did the on September 24, 2021 by guest. Protected copyright. 42 43 44 113 EMPOWER intervention trigger patients’ motivation to deprescribe by increasing knowledge 45 46 114 and concern about benzodiazepines? (2) Did the intervention augment patients’ capacity and 47 48 115 self-efficacy to taper benzodiazepines? (3) Did the intervention create opportunities for the 49 50 51 116 patient to discuss and receive support from a healthcare provider to engage in the deprescribing 52 53 117 process? Our goal was to determine for whom the EMPOWER intervention worked, in what 54 55 118 context it worked best and in which situations it failed. 56 57 58 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 119 METHODS 4 5 6 120 Study design 7 8 121 A realist evaluation was conducted alongside the EMPOWER randomized controlled trial. 9 10 16 11 122 This report follows RAMESES II guidelines for realist evaluation. The approach was chosen to 12 12 13 123 expand on the results of the EMPOWER trial, in order to inform the implementation of future 14 15 124 deprescribing initiativesFor by peerexamining the reviewpossible causes and contextualonly factors associated with 16 17 28 18 125 change. Specifically, we chose to use a theory-based, sequential mixed methods approach in 19 20 126 order to gain a deeper understanding of the contexts, mechanisms and outcomes of the 21 22 127 deprescribing intervention as experienced by the patient. The study was approved by the Institut 23 24 25 128 Universitaire de Gériatrie de Montréal Ethics Committee in Montreal, Quebec, Canada. 26 27 129 28 29 130 Environment surrounding the evaluation 30 31 32 131 The EMPOWER trial “Eliminating Medications through Patient Ownership of End Results”

41 135 November 2013, with community-dwelling participants randomly recruited via pharmacists on September 24, 2021 by guest. Protected copyright. 42 43 44 136 located within a 200 km radius of the Montreal urban area in Quebec, Canada. Participants were 45 46 137 303 older, community-dwelling, chronic users of benzodiazepine medication, and agreed to 47 48 138 home visits and telephone follow-up interviews by the research team. All benzodiazepine 49 50 51 139 prescriptions for seniors were covered under the publicly financed drug plan in the province of 52 53 140 Quebec, excluding the program’s deductible (if applicable). Provincial governments covered 54 55 141 physician reimbursements for patient visits, and drug dispensing fees for pharmacists, as part of 56 57 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 142 Canada’s universal health care program. 4 5 6 143 7 8 144 The EMPOWER intervention 9 10 11 145 The 8-page EMPOWER brochure, available at 12 30 13 146 http://www.criugm.qc.ca/fichier/pdf/BENZOeng.pdf, aims to promote active learning by 14 15 147 incorporating andFor using constructivist peer learning review principles31. The only brochure includes a self- 16 17 18 148 assessment component and presentation of the evidence-based risks associated with 19 20 149 benzodiazepine use in an effort to elicit cognitive dissonance.10 Elements of social comparison 21 22 150 theory32, through the use of peer champion stories, are also integrated in the intervention. The 23 24 25 151 brochure provides a self-guided tapering schedule, consisting of a visual tapering protocol 26 27 152 showing pictures of full pills, halved pills and quartered pills30. 28 29 153 30 31 32 154 The program theories embedded in the EMPOWER intervention are based on Mitchie et

41 158 refers to the external factors that permit or promote a behavior to happen, and include both the on September 24, 2021 by guest. Protected copyright. 42 43 44 159 physical and social environment of the individual. The specific mechanisms that we aimed to test 45 46 160 with the realist evaluation were whether the EMPOWER brochure: (1) triggered older adults’ 47 48 161 motivation to deprescribe by increasing knowledge and concern about benzodiazepines; (2) built 49 50 51 162 capacity to taper by augmenting self-efficacy; and (3) drove opportunities to receive support 52 53 163 from a healthcare provider to deprescribe. Table 1 links the programme theories and mechanisms 54 55 164 to the corresponding intervention components. 56 57 58 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 165 4 5 6 166 Evaluation of the programme theories 7 8 167 The evaluation of the programme theories consisted of quantitative data collection and 9 10 11 168 analysis, qualitative data collection and analysis, and triangulation of the quantitative and 12 34 13 169 qualitative results. Data collection was conducted between July 2010 and November 2013 as 14 15 170 part of the EMPOWERFor clinical peer trial. Analysis, review triangulation and only refinement of the Context- 16 17 18 171 Mechanism-Outcome configuration took place subsequent to completion of the trial. 19 20 172 21 22 173 Data collection methods 23 24 25 174 Quantitative data included pre- and 1-week post-intervention information on knowledge 26 27 175 about benzodiazepine-related harms, beliefs about the necessity of taking benzodiazepines versus 28 29 176 concern about harms, self-efficacy for tapering, and intent to discuss deprescribing with a health 30 31 32 177 care provider. We measured gains in knowledge with the four true or false questions listed in the

41 181 Medicines questionnaire (BMQ-Specific) at both time points. The BMQ-Specific consists of two on September 24, 2021 by guest. Protected copyright. 42 43 44 182 validated 5-item sub-scales assessing the respondents’ perceptions about the necessity and 45 46 183 concerns associated with taking benzodiazepines.35 Participants indicate their degree of 47 48 184 agreement with each statement on a 5 point Likert scale (1=strongly disagree, 5=strongly agree). 49 50 51 185 Scores are summed into their respective sub-category (5-25 point scale) with higher scores 52 53 186 indicating stronger beliefs. Risk perception was assessed using a single question 1-week post 54 55 187 intervention in which participants were asked whether they perceived the same, increased, or no 56 57 58 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 188 risk from consumption of their benzodiazepine following the intervention. In order to determine 4 5 6 189 whether the EMPOWER brochure increased capacity to taper by augmenting self-efficacy, we 7 8 190 measured self-efficacy for tapering on the Medication Reduction Self-efficacy scale, which 9 10 11 191 allows the respondent to rate on a scale of 0 to 100 their degree of confidence for tapering 12 36 13 192 benzodiazepines. Participants were also asked to indicate (yes/no) post intervention if they had 14 15 193 spoken to or intendedFor to discuss peer medication review discontinuation with only their doctor and/or pharmacist. 16 17 18 194 Health status was assessed at baseline using the first item of the Short-Form-12 Health Survey 19 20 195 and dichotomized by categorizing poor to fair responses as poor health.37 21 22 196 23 24 25 197 Qualitative data were collected after the 6-month follow-up, using semi-structured 26 27 198 interviews conducted at participants’ homes. Twenty-one participants were strategically sampled 28 29 199 for in-depth qualitative interviews using a contrast sample design, based on cessation of 30 31 38 32 200 benzodiazepines (yes or no) combined with intent to discuss tapering (yes or no). Interviews

41 204 experience with the tapering process, and personal interactions with health care providers. on September 24, 2021 by guest. Protected copyright. 42 43 44 205 45 46 206 Analysis 47 48 207 Participants with complete follow-up data were included in the quantitative analysis 49 50 51 208 (n=261, mean age 74.6 + 6.3, 72% women). Participant characteristics were described and 52 53 209 compared using means with standard deviations and independent t-tests for continuous data, and 54 55 210 percentages and Chi-square tests for categorical data, according to each of three outcomes: intent 56 57 58 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 211 to deprescribe with successful discontinuation, intent to deprescribe with failed discontinuation, 4 5 6 212 and no intent to deprescribe. Individual changes were computed from baseline to post- 7 8 213 intervention in knowledge, in the BMQ necessity and concerns subscales, and in self-efficacy 9 10 11 214 scores for tapering. Risk differences with 95% confidence intervals were calculated for the 12 13 215 proportion of participants in each group who demonstrated increased knowledge, heightened 14 15 216 concern about benzodiazepineFor peer use, and augmented review self-efficacy onlyfor tapering. The statistical 16 17 39 18 217 significance for all analyses was set at p<0.05 (two-sided). SPSS Version 21.0 (SPSS Inc. 19 20 218 Chicago, IL, USA) was used for all analyses. 21 22 219 Qualitative data from the semi-structured interviews were analyzed using thematic content 23 24 40 25 220 analysis. Discourses were contrasted according to whether participants discontinued 26 27 221 benzodiazepines and/or expressed the intent to discuss discontinuation. Interviews were coded 28 29 222 using Dedoose software. Themes were derived from the data and supported by quotes. Initially, 30 31 32 223 two researchers independently read the transcripts and field notes, then collaboratively developed

41 227 Quantitative and qualitative results were combined and analyzed in an iterative fashion on September 24, 2021 by guest. Protected copyright. 42 43 41 44 228 through use of a triangulation protocol using a convergence coding matrix, as described by 45 46 229 Farmer et al.42 The convergence matrix served to inform which theories should be retained or 47 48 230 discarded, as well as which contexts favorably or unfavorably influenced a patient’s trajectory 49 50 51 231 based on agreement, partial agreement, or dissonance between the quantitative and qualitative 52 53 232 data.41 42 Differences were adjudicated via discussion and consensus. 42 The convergence-coding 54 55 233 matrix is available from the authors upon request. 56 57 58 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 234 RESULTS 4 5 6 235 Contexts in which the EMPOWER intervention succeeded in triggering motivation and 7 8 236 capacity to deprescribe 9 10 11 237 The EMPOWER intervention triggered motivation and capacity to deprescribe in 167 of 12 13 238 261 participants (64%). Compared to individuals who reported no intent to deprescribe, 14 15 239 participants whoFor intended topeer deprescribe werereview significantly more only likely to demonstrate improved 16 17 18 240 knowledge (risk difference, 58.50% [95% CI, 46.98%-67.44 %]), lower perceived necessity (risk 19 20 241 difference, 56.03% [95% CI, 44.63%-64.81%]), increased concern (risk difference, 67.67% 21 22 242 [95% CI, 57.36%-74.91%]), enhanced self-efficacy for tapering (risk difference, 56.90% [95% 23 24 25 243 CI, 45.41%-65.77%]), and a greater perception of risk (risk difference, 35.14% [95% CI, 26 27 244 23.06%-45.39%]) (Table 2). No differences were observed in the age, sex, educational level, or 28 29 245 indication for benzodiazepine use between the two groups. Individuals who did not intend to 30 31 32 246 deprescribe were more likely to report poor health (40% vs 28%, 12.28% [95% CI, 0.44 %-

41 250 emerged from the qualitative interviews, describing the contexts that enabled deprescribing. on September 24, 2021 by guest. Protected copyright. 42 43 44 251 45 46 252 Contexts in which the EMPOWER intervention failed 47 48 253 Two points were identified along the deprescribing process where the program theories 49 50 51 254 failed. The first was at the point where the intervention aimed to trigger motivation to 52 53 255 deprescribe. This situation occurred in 36% of participants, and was characterized by a lack of 54 55 256 new knowledge and no increase in perceived risk after receipt of the EMPOWER brochure. The 56 57 58 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 257 contexts included previous reassurance by a physician that benzodiazepines were safe or 4 5 6 258 necessary, the belief that the benefits of benzodiazepines outweighed the risks for immediate 7 8 259 symptom relief, and a reliance on benzodiazepines for everyday coping because of poor health 9 10 11 260 (Table 4). The second failure point occurred during participants’ efforts to initiate or complete 12 13 261 the deprescribing process. Lack of support from a healthcare provider, intolerance to withdrawal 14 15 262 symptoms, and aFor sudden loss peer of confidence review to live without sleeping only pills were contexts that led 16 17 18 263 participants to abort the deprescribing process once they showed initial motivation and capacity 19 20 264 to deprescribe (Table 4). 21 22 265 23 24 25 266 Refining the context-mechanism-outcome configuration for deprescribing interventions 26 27 267 The initial context-mechanism-outcome configuration that drove the development of the 28 29 268 EMPOWER intervention was a simple, linear progression along different stages of readiness to 30 31 32 269 deprescribe, similar to Prochaska & DiClemente’s transtheoretical model of change (Figure

41 273 opportunities to discuss deprescribing with a healthcare professional. We assumed that the on September 24, 2021 by guest. Protected copyright. 42 43 44 274 healthcare provider would be supportive and accompany the patient along the deprescribing 45 46 275 process. This initial configuration oversimplified the stages through which individuals 47 48 276 transitioned after receiving the deprescribing intervention. Figure 1b depicts a revised, non-linear 49 50 51 277 context-mechanism-outcome configuration that takes into account the complexity of internal and 52 53 278 external influences on initiating and completing the deprescribing process from the consumer’s 54 55 279 perspective. The revised model recognizes that new information influences beliefs and actions 56 57 58 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 280 only if the information generates a desire strong enough not to be overwhelmed by competing 4 5 6 281 factors arising from other sources. In many instances, the desire for risk reduction, which was the 7 8 282 prime motivator behind the development of the EMPOWER intervention, did not supersede 9 10 11 283 concerns about symptom recurrence, and other psychological and health factors, as well as 12 13 284 interpersonal relationships with healthcare providers, which played critical contextual roles in the 14 15 285 outcome of the intervention.For peer review only 16 17 18 286 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 287 DISCUSSION 4 5 6 288 7 8 289 This realist evaluation tested the mechanisms embedded in the EMPOWER intervention, 9 10 11 290 and showed that motivation and capacity to deprescribe were triggered in 64% of older chronic 12 13 291 benzodiazepines consumers, supporting the theory that provision of new knowledge about 14 15 292 medication harmsFor can raise peerconcern and augment review patients’ self-efficacy only to deprescribe. However, 16 17 18 293 the analysis also indicated that human motivation to deprescribe is complex and unstable. A 19 20 294 variety of internal and external contexts can interfere along the pathway of deprescribing 21 22 295 interventions. These influences include perceptions about one’s long-term health goals, fear of 23 24 25 296 symptom recurrence, confidence about tapering, psychological attachment to the drug, current 26 27 297 health state and lack of support from health care providers. 28 29 298 Use of a mixed methods approach enabled us to explore the breadth, depth, and 30 31 32 299 complexity of the patient’s experience of deprescribing from a social, behavioural and health

41 303 realist evaluation. One untested mechanism is provision of information about the lack of drug on September 24, 2021 by guest. Protected copyright. 42 43 44 304 benefits in certain populations, such as statins to reduce cholesterol levels in palliative care 45 46 305 patients with limited life expectancy. 45 46 47 48 306 49 50 51 307 Our findings contribute to the literature by illustrating that linear progression along 52 53 308 different stages of readiness to deprescribe does not fully explain successful deprescribing from 54 55 309 the patient’s perspective. This conclusion is consistent with other critiques of the transtheoretical 56 57 58 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 310 model, which claim that the stages of readiness are arbitrary, that human beings do not make 4 5 6 311 logical and stable plans to change their behavior, and that setbacks can occur along the trajectory 7 8 312 of change47. Education appears to be necessary but insufficient for many individuals, and new 9 10 11 313 strategies will be needed to trigger deprescribing in prohibitive contexts where the EMPOWER 12 13 314 mechanisms failed. As capacity and motivations change over time, reminders and ongoing 14 15 315 discussions aboutFor the risks ofpeer inappropriate review medications may progressively only trigger and sustain 16 17 18 316 patients’ commitments to engage in the deprescribing process. Some competing factors may 19 20 317 wane, such as poor health. Offering cognitive behavioural therapy to patients during the most 21 22 318 difficult last quarter period of the tapering protocol may augment self-efficacy for overcoming 23 24 36 25 319 withdrawal symptoms. Interventions can be directed at health care providers who discourage 26 27 320 deprescribing efforts. Continuing medical education to inform health providers about the 28 29 321 mounting evidence on the harms of benzodiazepine use may curtail the phenomenon of 30 31 20 48 32 322 physicians who continue to promote the use of inappropriate medication. Future research

41 326 which tensions about deprescribing are played out and drive behavior change49 50. As we did not on September 24, 2021 by guest. Protected copyright. 42 43 44 327 directly ask patients if they felt internal tension, we were unable to record feelings or processes 45 46 328 of cognitive dissonance. 47 48 329 In conclusion, this realist evaluation of deprescribing from the patient’s perspective 49 50 51 330 increases current understanding about the specific mechanisms and contexts that can be 52 53 331 successful in triggering patients to engage in the deprescribing process. 54 55 332 56 57 58 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 333 ACKNOWLEDGMENTS 4 5 6 334 We wish to acknowledge the work of Anne-Sophie Michaud and Anastasia Soboleva, who 7 8 335 conducted the in-home interviews and helped in identifying first order coding for the qualitative 9 10 11 336 portion of the manuscript. Aditionally, we would like to thank Johanne Collin for her advice in 12 13 337 devising the general strategy for the qualitative interviews. We express gratitude to all the 14 15 338 participants and Forpharmacists peer who took part review in this trial. Particular only thanks are offered to the 16 17 18 339 Pharmacy Services Department of the Jean Coutu Group (PJC) Inc. for their collaboration and 19 20 340 support. 21 22 341 23 24 25 342 Conflict of Interest Checklist: 26 27 Elements of Martin Tannenbaum343 28 Financial/Personal 29 Conflicts 30 Yes No Yes No 31 32 Employment or X X

33 Affiliation http://bmjopen.bmj.com/ 34 35 Grants/Funds X X 36 37 Honoraria X X 38 39 Speaker Forum X X 40

41 on September 24, 2021 by guest. Protected copyright. 42 Consultant X X 43 44 Stocks X X 45 46 Royalties X X 47 48 Expert Testimony X X 49 50 51 Board Member X X 52 53 Patents X X 54 55 Personal Relationship X X 56 344 57 58 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 345 4 5 6 346 7 8 347 Author Contributions: 9 10 11 348 Martin and Tannenbaum contributed to the Study concept and design, Analysis and 12 13 349 Interpretation of data, Drafting of the manuscript, critical revision of the manuscript for 14 15 350 important intellectualFor content, peer and final reviewapproval of the version only to be published. 16 17 18 351 19 20 352 Sponsor: 21 22 353 The study was funded by the Canadian Institutes of Health Research. Philippe Martin received a 23 24 25 354 bursary from the Fonds de Recherche en Santé de Quebec & the Michel Saucier Endowed Chair 26 27 355 in Pharmacology, Health and Aging of the Faculty of Pharmacy of the Université de Montréal. 28 29 356 Cara Tannenbaum is a clinician scientist funded by the Fonds de Recherche en Santé de Quebec. 30 31 32 357 The authors retained full independence from the study sponsors in the design and conduct of the

Page 19 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 20 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 Figures and Legends: 4 5 6 Figure 1 a: Initial Deprescribing Context-Mechanism-Outcome configuration 7 8 Figure 1b: Refined Deprescribing Context-Mechanism-Outcome configuration 9 10 11 Legend: 12 C = Context (grey circles); 13 M = Mechanism (purple diamonds); 14 O = Outcome (blue rectangles) 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 20

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 REFERENCES 4 5 6 1. Jansen J, Naganathan V, Carter SM, et al. Too much medicine in older people? 7 8 Deprescribing through shared decision making. 2016 9 10 11 2. Scott IA, Hilmer SN, Reeve E, et al. Reducing inappropriate polypharmacy: the 12 13 process of deprescribing. JAMA internal medicine 2015;175(5):827-34. 14 15 3. Frank C, WeirFor E. Deprescribing peer for olderreview patients. CMAJ only: Canadian Medical 16 17 18 Association journal = journal de l'Association medicale canadienne 2014;186(18):1369- 19 20 76. doi: 10.1503/cmaj.131873 [published Online First: 2014/09/04] 21 22 4. Reeve E, To J, Hendrix I, et al. Patient barriers to and enablers of deprescribing: a 23 24 25 systematic review. Drugs & aging 2013;30(10):793-807. doi: 10.1007/s40266-013-0106- 26 27 8 28 29 5. Woodward MC. Deprescribing: achieving better health outcomes for older people 30 31 32 through reducing medications. Journal of Pharmacy Practice and Research

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 9. Declercq T, Petrovic M, Azermai M, et al. Withdrawal versus continuation of chronic 4 5 6 antipsychotic drugs for behavioural and psychological symptoms in older people with 7 8 dementia. Cochrane database of systematic reviews (Online) 2013(3):CD007726. doi: 9 10 11 10.1002/14651858.CD007726.pub2 [published Online First: 2013/04/02] 12 13 10. Martin P, Tamblyn R, Ahmed S, et al. A drug education tool developed for older 14 15 adults changesFor knowledge, peer beliefs and riskreview perceptions about only inappropriate 16 17 18 benzodiazepine prescriptions in the elderly. Patient education and counseling 2013 doi: 19 20 10.1016/j.pec.2013.02.016 [published Online First: 2013/04/02] 21 22 11. Ostini R, Hegney D, Jackson C, et al. Systematic review of interventions to improve 23 24 25 prescribing. The Annals of pharmacotherapy 2009;43(3):502-13. doi: 10.1345/aph.1L488 26 27 [published Online First: 2009/03/06] 28 29 12. Tannenbaum C, Martin P, Tamblyn R, et al. Reduction of inappropriate 30 31 32 benzodiazepine prescriptions among older adults through direct patient education: the

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 15. Pollmann AS, Murphy AL, Bergman JC, et al. Deprescribing benzodiazepines and Z- 4 5 6 drugs in community-dwelling adults: a scoping review. BMC pharmacology & toxicology 7 8 2015;16:19. doi: 10.1186/s40360-015-0019-8 [published Online First: 2015/07/05] 9 10 11 16. Wong G, Westhorp G, Manzano A, et al. RAMESES II reporting standards for realist 12 13 evaluations. BMC Med 2016;14(1):96. doi: 10.1186/s12916-016-0643-1 [published 14 15 Online First:For 2016/06/28] peer review only 16 17 18 17. Johansson T, Abuzahra ME, Keller S, et al. Impact of strategies to reduce 19 20 polypharmacy on clinically relevant endpoints: a systematic review and meta-analysis. 21 22 British journal of clinical pharmacology 2016;82(2):532-48. doi: 10.1111/bcp.12959 23 24 25 18. Page AT, Clifford RM, Potter K, et al. The feasibility and effect of deprescribing in 26 27 older adults on mortality and health: a systematic review and meta-analysis. British 28 29 journal of clinical pharmacology 2016;82(3):583-623. doi: 10.1111/bcp.12975 30 31 32 19. Billioti de Gage S, Moride Y, Ducruet T, et al. Benzodiazepine use and risk of

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 22. Paterniti S, Dufouil C, Alperovitch A. Long-term benzodiazepine use and cognitive 4 5 6 decline in the elderly: the Epidemiology of Vascular Aging Study. Journal of clinical 7 8 psychopharmacology 2002;22(3):285-93. [published Online First: 2002/05/15] 9 10 11 23. Allain H, Bentue-Ferrer D, Polard E, et al. Postural instability and consequent falls 12 13 and hip fractures associated with use of hypnotics in the elderly: a comparative review. 14 15 Drugs & agingFor 2005;22(9):749-65. peer [published review Online First: only2005/09/15] 16 17 18 24. Cotroneo A, Gareri P, Nicoletti N, et al. Effectiveness and safety of hypnotic drugs in 19 20 the treatment of insomnia in over 70-year old people. Archives of gerontology and 21 22 geriatrics 2007;44 Suppl 1:121-4. doi: 10.1016/j.archger.2007.01.018 [published Online 23 24 25 First: 2007/02/24] 26 27 25. Tannenbaum C. Effect of age, education and health status on community dwelling 28 29 older men's health concerns. The aging male : the official journal of the International 30 31 32 Society for the Study of the Aging Male 2012;15(2):103-8. doi:

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 29. Martin P, Tamblyn R, Ahmed S, et al. An educational intervention to reduce the use 4 5 6 of potentially inappropriate medications among older adults (EMPOWER study): 7 8 protocol for a cluster randomized trial. Trials 2013;14:80. doi: 10.1186/1745-6215-14-80 9 10 11 [published Online First: 2013/03/22] 12 13 30. Tannenbaum C, Martin P. Sedative-Hypnotic Medication De-Prescribing Brochure, 14 15 Available at:For http://www.criugm.qc.ca/fichier/pdf/BENZOeng.pdf peer review only. In: Montréal 16 17 18 Iudgd, ed., 2014. 19 20 31. Hein G. Constructivist learning theory. Institute for Inquiry Available 21 22 at:/http://www/ exploratorium edu/ifi/resources/constructivistlearning htmlS 1991 23 24 25 32. Festinger L. A theory of social comparison processes. Human relations 26 27 1954;7(2):117-40. 28 29 30 33. Michie S, van Stralen MM, West R. The behaviour change wheel: a new method for 31 32 characterising and designing behaviour change interventions. Implementation science : IS

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 journal of the Division of Health Psychology, American Psychological Association 4 5 6 2005;24(3):281-7. doi: 10.1037/0278-6133.24.3.281 [published Online First: 2005/05/19] 7 8 37. McDowell I. Measuring health: A guide to rating scales and questionnaires (3rd ed.). 9 10 11 New York, NY: Oxford University Press; US 2006:xvi, 748. 12 13 38. Sandelowski M. Combining qualitative and quantitative sampling, data collection, 14 15 and analysis Fortechniques peerin mixed-method review studies. Research inonly nursing & health 16 17 18 2000;23(3):246-55. [published Online First: 2000/06/28] 19 20 39. Ellis P. The Essential Guide to Effect Sizes: Statistical Power, Meta-Analysis, 21 22 and the Interpretation of Research Results. Cambridge: Cambridge University Press 2010. 23 24 25 40. Braun V, Clarke V. Using thematic analysis in psychology. Qualitative Research in 26 27 Psychology 2006;3(2):77-101. 28 29 41. O'Cathain A, Murphy E, Nicholl J. Three techniques for integrating data in mixed 30 31 32 methods studies. British Medical Journal 2010;341:c4587. doi: 10.1136/bmj.c4587

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 46. Holmes HM, Todd A. Evidence-based deprescribing of statins in patients with 4 5 6 advanced illness. JAMA internal medicine 2015;175(5):701-02. 7 8 47. West R. Time for a change: putting the Transtheoretical (Stages of Change) Model to 9 10 11 rest. Addiction 2005;100(8):1036-9. doi: 10.1111/j.1360-0443.2005.01139.x [published 12 13 Online First: 2005/07/27] 14 15 48. Cassel CK,For Guest JA. peer Choosing wisely: review helping physicians only and patients make smart 16 17 18 decisions about their care. JAMA : the journal of the American Medical Association 19 20 2012;307(17):1801-2. doi: 10.1001/jama.2012.476 [published Online First: 2012/04/12] 21 22 49. Theory-driven design strategies for technologies that support behavior change in 23 24 25 everyday life. Proceedings of the SIGCHI Conference on Human Factors in Computing 26 27 Systems; 2009. ACM. 28 29 50. Bem DJ. Self-Perception: An Alternative Interpretation of Cognitive Dissonance 30 31 32 Phenomena. Psychological Review 1967;74(3):183-200.

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 27

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from Page 28 of 41 28 The printed format of the 8- the of format printed The knowledge effective and take to piece transfer ahealthcare to show Provision of an easy-to-use easy-to-use an of Provision tapering week 16-20 visual a take to when showing tool pill, quarter or half whole, dose the toskip when and completely Information about changes changes about Information with metabolism indrug a higher to lead can that age meant effects, side of risk elicit and beliefs tochange of safety the about concern older in themedication adults page brochure makes it an it makes brochure page provider

* *

* * *

* * * * * * *

* *

(n=92) (n=92) Group 1- 1- Group Successful Successful deprescribing

Characteristic and Outcome

Table 2:Table

Independent sample t-test for continuous variables, chi square for categorical variables. * ‡ Page 29 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from Page 30 of 41 30

[84 ” [84

[79 y.o. [79successful y.o. man, [68 y.o. [68 y.o. successful woman, [72 y.o. [72successful y.o. taper]man,

[72 y.o. woman, successful taper][72 y.o. woman, [84 y.o. man, [84successful taper] y.o. man,

Supporting Supporting citation

BMJ Open http://bmjopen.bmj.com/

(n=7) (n=7) 1 (14%) deprescribe No attempt toNo attempt

(n=7) (n=7) Failed Failed on September 24, 2021 by guest. Protected copyright. 4 (57%) deprescribing deprescribing For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

(n=7) (n=7) 5 (71%) 5 (71%) 3 (43%) 0 3 (43%) 1 (14%) 0 5 (71%) 3 (43%) 1 (14%) 6 (86%) 5 (71%) 1 (14%) 6 (86%) 4 (57%) 1 (14%) 5 (71%) 4 (57%) 2 (29%) Successful

deprescribing deprescribing For peer review only Key theme

[85 y.o. [85failed y.o. man,

“My doctor“My told me: age, “At your don’t about worry it. You’ve been taking this pill for a while and taking you Youare fine. aren’t a dangerousat dose all

to (n=7) (n=7) 5 (71%) No attempt No attempt deprescribe

(n=7) (n=7) Failed Failed on September 24, 2021 by guest. Protected copyright. 1 (14%) deprescribing deprescribing For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

0 1 (14%) 4 (57%) 0 4 (57%) 4 (57%) 0 5 (71%) _ (n=7) (n=7) 1 (14%) 1 (14%) 1 (14%) 4 (57%) Successful

deprescribing deprescribing For peer review only

Key theme

Loss of Loss confidence to complete the processtapering (post-intervention) Intolerance to recurrence of symptoms/withdrawal effects Discouragement a physicianfrom 1 (14%) 3 (43%) 5 (71%) Quality of lifeof Quality during focus end-of life 0 2 (29%) 3 (43%) Reliance on for medication coping/everyday function Lack of perceptionLack of personalof risk 1 (14%) 2 (29%) 5 (71%) Unquestioning inbelief their physician 1 (14%) 1 (14%) 3 (43%) Poor health Poorstatus health Previous discouragement from physician / Negative attitude physicianof towards deprescribing Table 4: Contexts where deprescribing mechanisms failed mechanisms deprescribing where Contexts Table 4: Page 31 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 32 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 Figure 1a: Initial Deprescribing Context-Mechanism-Outcome configuration 27 28 Legend: 29 C = Context (grey circles); 30 M = Mechanism (purple diamonds); 31 O = Outcome (Blue rectangles) 32 Figure 1a 98x55mm (300 x 300 DPI) 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

33 Figure 1b: Refined Deprescribing Context-Mechanism-Outcome configuration http://bmjopen.bmj.com/ 34 Legend: 35 C = Context (grey circles); 36 M = Mechanism (purple diamonds); 37 O = Outcome (Blue rectangles) 38 Figure 1b 39 133x102mm (300 x 300 DPI) 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Interview Guide 5 6 7 EMPOWER Study - Qualitative section 8 9 6 months post-intervention 10 11 I. INTRODUCTION 12 13 Mr. X / Mrs. Y, Hi, 14 15 For peer review only 16 First off, we would like to thank you for taking the time to participate in all the steps of the EMPOWER 17 study. During the course of this research project you received an educational brochure which allowed 18 you to come to your own conclusions on your use of medication XY. The objective of the interview 19 20 today is to collect your opinion of the whole intervention process in order to better evaluate what 21 happened with the intervention. As you already know, we wish to discuss with you your experience 22 during this process and to collect your opinion on various aspects of your experience. 23 24 25 This interview is conducted for a University study and all information shared today will be 26 confidential and anonymized. There are now right or wrong answers here, all that is important to us 27 is to capture your honest opinion and experiences. 28 29 30 If you have no objections, we would like to record this conversation in order to facilitate the full 31 collection of this interview. 32 33 Lets start with a small introduction, my name is…. And here is …. http://bmjopen.bmj.com/ 34 35 36 Our colleague XX, the person who previously contacted you has indicated that you would be a 37 good candidate to be interviewed. Although you have already discussed some of your experiences 38 39 with her, we would like to start over from the start in order to capture all the details of your 40 experience.

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 II. Aging, disease and medication 45 46 1. How old are you? Are you still active? Do you still work? Do you volunteer? Are you 47 48 close to your family? 49 50 Relaunch on: 51 52 Perception of aging 53 54 Physical psychological and social difficulties associated with aging. 55 56 2. How would you describe your current health status? 57 58 3. I would like to discuss your attitude towards medications in general. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 What do medications represent at your age? Tell me about your current prescriptions. 4 5 How do you manage taking your daily medications? 6 7 Relaunch: 8 9 Do they take/manage them themselves? 10 11 Do they use a Dispill? 12 13 Do you get any help managing them? 14 15 Do you ever forget them?For peer review only 16 17 4. Now, let’s talk about the medication for which you received the educational 18 brochure. Tell me, how do you take your benzodiazepine? For which reasons do 19 20 you take them and under which circumstances did you start taking this 21 medication? 22 23 Relaunch: 24 25 - Sleeping pill? 26 27 - Anxiolytic? 28 29 5. How long have you been taking this medication for? How has the use of this medication 30 evolved over time? 31 32 6. Do you still find this medication effective? Why? Do you have any side effects from this 33 http://bmjopen.bmj.com/ 34 medication? 35 36 Relaunch: 37 38 Falls? 39 40 Dizziness?

41 Etc. on September 24, 2021 by guest. Protected copyright. 42 43 (As a whole, what are the advantages and disadvantages associated with taking this 44 45 medication?) 46 47 7. Have you ever intended on ceasing your benzodiazepine (or actually attempted to) before 48 receiving the educational brochure? 49 50 a. If yes, how did it go? 51 52 b. If no, why? 53 54 8. Question only for those who intended to but did not try ceasing medication: 55 56 a. How important is ceasing your medication to you 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 b. How confident if your ability to stop using this medication? 4 5 c. How do you think you would feel if you ceased the medication? 6 7 For the interviewer: If the participant mentions having attempted of ceased the medication 8 9 AFTER the brochure, just mention that this will be discussed later on in the interview. 10 11 12 13 III. Prescription and the patient-physician relation 14 15 9. Who prescribed your benzodiazepine? Tell me about your experience (first For peer review only interaction? 16 and after?). What did your doctor tell you about taking this medication? 17 18 10. Do you see your doctor often? Tell me about your relation with him/her? 19 20 11. Is it always the same physician who prescribes you your medications? (Multiple 21 22 physicians? Family doctor?) 23 24 12. Before receiving the brochure, had you previously discussed or been approached by your 25 physician about the possibility to switch/stop your benzodiazepine? 26 27 a. If so, what were your expectations? What was their reaction? How did it go? 28 29 b. If not, why? 30 31 32

33 IV. Acquisition and relation with the pharmacist http://bmjopen.bmj.com/ 34 35 13. Now let’s talk about your pharmacist. 36 37 How do you manage buying your benzodiazepine? Tell me about your experiences. 38 39 àDo you go see your pharmacist yourself? Do you have your medications delivered? At what 40 frequency?

41 on September 24, 2021 by guest. Protected copyright. 42 What has your pharmacist told you about this medication? How does he interact with you? 43 44 14. Before receiving the brochure, had you previously discussed or been approached by your 45 46 pharmacist about the possibility to switch/stop your benzodiazepine? 47

48 a. If so, what were your expectations? What was their reaction? How did it go? 49 b. If not, why? 50 51 52 V. Reaction to the intervention 53 54 15. What was your first reaction when you read the information contained in the brochure 55 that we sent you? 56 57 a. What did you learn about the potential alternative treatments? 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 b. What did you learn about benzodiazepine cessation and withdrawal? 4 5 c. How do you now perceive the potential risks associated with your benzodiazepine 6 7 prescription? 8 9 16. Please explain to me what happened once you read the brochure? (Did you read it more 10 than once, discuss it with others?) 11 12 Relaunch: 13 14 Did you intend on initiating the tapering protocol after reading the brochure? Why? 15 For peer review only 16 17. Did you try implementing the tapering protocol suggested at the last page? 17 18 a. If yes: For what reasons did you decide to initiate the tapering program? 19 20 How was the process? (What were your withdrawal symptoms?) What helped you 21 succeed your tapering? What were the obstacles? How do you feel since tapering 22 23 off the medication? 24 25 b. If no: For what reasons did you decide not to initiate the tapering program? 26 27 What were the barriers? Do you have any questions or preoccupations regarding 28 the tapering process? Do you think you could change your mind in the future? 29 30 What would be the required criteria for you to stop your benzodiazepine? 31 32 18. For you, what are the important criteria/results that are most important to determine the 33 success of a stopping a medication? http://bmjopen.bmj.com/ 34 35 19. To conclude, what is your appraisal of the intervention? In what measure was the 36 37 information provided useful to you? Would you recommend this intervention to someone 38 else? Why? 39 40 20. To what degree do you value the importance of your implication in the management of

41 your medication? on September 24, 2021 by guest. Protected copyright. 42 43 21. Are there any other subjects that were not discussed in the context of this interview but 44 45 that you feel are important and that you would like to discuss? 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 In the title, identify the document as a 9 1 Y 1 10 realist evaluation 11 SUMMARY OR ABSTRACT 12 13 Journal articles will usually require an 14 abstract, while reports and other forms of 15 For peer review only 16 publication will usually benefit from a 17 short summary. The abstract or summary 18 should include brief details on: the policy, 19 programme or initiative under evaluation; 20 programme setting; purpose of the 21 evaluation; evaluation question(s) and/or 22 23 objective(s); evaluation strategy; data 24 collection, documentation and analysis 25 methods; key findings and conclusions 26 2 Y 3-4 27 Where journals require it and the nature of 28 29 the study is appropriate, brief details of 30 respondents to the evaluation and 31 recruitment and sampling processes may 32 also be included 33 http://bmjopen.bmj.com/ 34 35 Sufficient detail should be provided to 36 identify that a realist approach was used 37 and that realist programme theory was 38 developed and/or refined 39 INTRODUCTION 40

41 Rationale for Explain the purpose of the evaluation and on September 24, 2021 by guest. Protected copyright. 3 Y 5 42 evaluation the implications for its focus and design 43 44 Describe the initial programme theory (or 45 4 Programme theory theories) that underpin the programme, Y 6 46 policy or initiative 47 48 State the evaluation question(s) and specify 49 the objectives for the evaluation. Describe Evaluation questions, 50 5 whether and how the programme theory Y 6 objectives and focus 51 was used to define the scope and focus of 52 53 the evaluation 54 State whether the realist evaluation 55 6 Ethical approval required and has gained ethical approval Y 7 56 57 from the relevant authorities, providing 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 9 details as appropriate. If ethical approval 10 was deemed unnecessary, explain why 11 METHODS 12 13 Rationale for using Explain why a realist evaluation approach 7 Y 7 14 realist evaluation was chosen and (if relevant) adapted 15 For peer review only 16 Environment Describe the environment in which the 17 8 surrounding the Y 7 evaluation took place 18 evaluation 19 20 Describe the 21 programme policy, Provide relevant details on the programme, 9 Y 8 22 initiative or product policy or initiative evaluated 23 evaluated 24 25 A description and justification of the 26 evaluation design (i.e. the account of what 27 was planned, done and why) should be 28 29 included, at least in summary form or as an 30 appendix, in the document which presents Describe and justify the main findings. If this is not done, the 31 10 Y 8-9 32 the evaluation design omission should be justified and a 33 reference or link to the evaluation design http://bmjopen.bmj.com/ 34 given. It may also be useful to publish or 35 36 make freely available (e.g. online on a 37 website) any original evaluation design 38 document or protocol, where they exist 39 40 Describe and justify the data collection

41 methods – which ones were used, why and on September 24, 2021 by guest. Protected copyright. 42 how they fed into developing, supporting, 43 Data collection refuting or refining programme theory 11 Y 9-10 44 methods 45 46 Provide details of the steps taken to 47 enhance the trustworthiness of data 48 collection and documentation 49 Describe how respondents to the evaluation 50 51 Recruitment process were recruited or engaged and how the 52 12 and sampling sample contributed to the development, Y 7, 9-10 53 strategy support, refutation or refinement of 54 programme theory 55 Describe in detail how data were analysed. 56 13 Data analysis Y 10-11 57 This section should include information on 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 41 BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 9 the constructs that were identified, the 10 process of analysis, how the programme 11 theory was further developed, supported, 12 refuted and refined, and (where relevant) 13 14 how analysis changed as the evaluation 15 Forunfolded peer review only 16 RESULTS 17 18 Report (if applicable) who took part in the 19 evaluation, the details of the data they Details of 20 14 provided and how the data was used to Y 12-13 participants 21 develop, support, refute or refine 22 23 programme theory 24 Present the key findings, linking them to 25 contexts, mechanisms and outcome 26 27 15 Main findings configurations. Show how they were used Y 12-14 28 to further develop, test or refine the 29 programme theory 30 DISCUSSION 31 32 Summarise the main findings with attention 33 to the evaluation questions, purpose of the http://bmjopen.bmj.com/ 16 Summary of findings Y 15-16 34 evaluation, programme theory and intended 35 36 audience 37 Discuss both the strengths of the evaluation 38 and its limitations. These should include 39 40 (but need not be limited to): (1)

41 consideration of all the steps in the on September 24, 2021 by guest. Protected copyright. 42 evaluation processes; and (2) comment on 43 the adequacy, trustworthiness and value of 44 the explanatory insights which emerged Strengths, limitations 45 17 Y 15-16 46 and future directions 47 In many evaluations, there will be an 48 expectation to provide guidance on future 49 directions for the programme, policy or 50 initiative, its implementation and/or design. 51 The particular implications arising from the 52 53 realist nature of the findings should be 54 reflected in these discussions 55 Comparison with Where appropriate, compare and contrast 56 18 Y 16 57 existing literature the evaluation’s findings with the existing 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-015959 on 4 May 2017. Downloaded from 1 2 3 4 Reported in 5 document Page(s) in RAMESES II reporting standards for realist evaluations 6 document 7 Y/N/Unclear 8 9 literature on similar programmes, policies 10 or initiatives 11 12 List the main conclusions that are justified Conclusion and by the analyses of the data. If appropriate, 13 19 Y 16 14 recommendations offer recommendations consistent with a 15 Forrealist peer approach review only 16 17 State the funding source (if any) for the Funding and conflict evaluation, the role played by the funder (if 18 20 Y 2,17 19 of interest any) and any conflicts of interests of the 20 evaluators 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml