Interventions to Improve Pharmacological Adherence Among Adults With Psychotic Spectrum Disorders and Bipolar Disorder: A Systematic Review

Author’s Accepted Manuscript Interventions to Improve Pharmacological Adherence among Adults with Psychotic Spectrum Disorders and Bipolar Disorder: A Systematic Review Daniel Hartung, Allison Low, Kazuaki Jindai, David Mansoor, Matthew Judge, Aaron Mendelson, Devan Kansagara, Makalapua Motu′apuaka, Michele Freeman, Karli Kondo

PII: DOI: Reference:

www.elsevier.com/locate/psym

S0033-3182(16)30100-1 http://dx.doi.org/10.1016/j.psym.2016.09.009 PSYM679

To appear in: Psychosomatics Received date: 14 June 2016 Revised date: 16 September 2016 Accepted date: 16 September 2016 Cite this article as: Daniel Hartung, Allison Low, Kazuaki Jindai, David Mansoor, Matthew Judge, Aaron Mendelson, Devan Kansagara, Makalapua Motu′apuaka, Michele Freeman and Karli Kondo, Interventions to Improve Pharmacological Adherence among Adults with Psychotic Spectrum Disorders and Bipolar Disorder: A Systematic Review, Psychosomatics, http://dx.doi.org/10.1016/j.psym.2016.09.009 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Interventions to Improve Pharmacological Adherence among Adults with Psychotic Spectrum Disorders and Bipolar Disorder: A Systematic Review Daniel Hartung, PharmD, MPH (1), [email protected] Allison Low, BA (2), [email protected] Kazuaki Jindai, MD, MPH (3, 4), [email protected] David Mansoor, MD (3, 4), [email protected] Matthew Judge, MD (3, 4), [email protected] Aaron Mendelson, BA (4), [email protected] Devan Kansagara, MD, MCR (2, 3, 4), [email protected] Makalapua Motu'apuaka, BA (2), makalapua.motu'[email protected] Michele Freeman, MPH (2, 4), [email protected] Karli Kondo, PhD (2, 4), [email protected] Author Affiliations: Oregon State University/Oregon Health & Science University College of Pharmacy (1), Portland, OR, USA; VA Evidence-based Synthesis Program (2), VA Portland Health Care System (3), Portland, OR, USA; Oregon Health & Science University (4), Portland, OR, USA. Author Contact: Karli Kondo, PhD; [email protected]; (503) 490-1907. VA Portland Health Care System, Mail Code R&D 71 3710 SW US Veterans Hospital Road Portland, OR 97239-2999 Conflict of Interest: The authors have no conflicts of interest to report.

2 ABSTRACT Background: It is unclear which interventions are effective at improving medication adherence in individuals with serious and persistent mental illness. The goal of this systematic review is to synthesize evidence examining the effectiveness, harms, and costs of interventions to improve medication adherence in patients with psychotic spectrum disorders and bipolar disorder. Methods: We conducted a systematic search of several electronic databases and clinical trial registries through January 2015 using a structured search strategy. Studies were included if they involved adult patients in general mental health settings, reported both measures of medication adherence and clinical outcomes, and were of sufficient methodological rigor. Studies were quality assessed and synthesized using established methods. Results: We identified 24 studies that met inclusion criteria. Twenty studies addressed interventions in patients with psychotic spectrum disorders. These interventions varied widely, with generally mixed findings contributing to low or insufficient strength of evidence; studies involving family members and technology interventions were the most consistently associated with a positive effect, however the strength of the evidence was low because of intervention heterogeneity. The evidence was insufficient to determine the effectiveness of interventions in patients with bipolar disorder. Conclusions: In individuals with psychotic spectrum disorders, interventions with family members or technology had the most consistent positive effect on adherence, although replication with objective adherence measures and evaluation of harms and costs is needed. There was insufficient evidence to draw conclusions about interventions in individuals with bipolar disorder. Keywords: medication adherence, pharmacotherapy, bipolar disorder, schizophrenia.

3 INTRODUCTION Among individuals with serious and persistent mental illness, medications are a cornerstone of therapy and nonadherence is a powerful predictor of poor outcomes.1-3 Poor adherence is strongly associated with reduced functional status and increased relapse rates among individuals with schizophrenia or bipolar disorder.4 Despite this, antipsychotic medication adherence in patients with schizophrenia is consistently estimated to be between 25% and 50%.5-7 Similarly, medications are the first line of treatment for persons with bipolar disorder,8 and reported rates of adherence are between 30% and 57%.9-11 Mental illness is commonly comorbid with other chronic physical health disorders. The prevalence of cardiovascular disease, diabetes, hypertension, and dyslipidemia are several fold higher among individuals with schizophrenia than the general population,12 and studies suggest that adherence to medications for these comorbid conditions is also poor.13, 14 Medication adherence is a complex behavior comprised of a series interrelated steps involving patients, their providers, and the healthcare system.15 As others have noted, the causes of medication nonadherence are multifactorial.9 Barriers to improved medication adherence include lack insight into the illness, cognitive function, side effects, regimen complexity, comorbid substance use, and poor therapeutic alliance.16, 17 While financial barriers are widely cited, even when completely eliminated, adherence to certain therapies remains sub-optimal.18 Consequently, a wide range of interventions designed to improve medication adherence have been developed. A recent systematic review of interventions for medication adherence in patients with chronic illness found that interventions involving education, case management, clinical reminders, reduction of patient expenses, and multicomponent approaches consistently improved

4 adherence across different clinical conditions.19 This review examined interventions for medication adherence in patients with depression but did not include other serious mental illnesses due to fundamental differences in both the underlying reasons for and the consequences of nonadherence in these patients, resulting in interventions with characteristics that may not be applicable to other populations. The goal of this systematic review was to summarize current evidence examining both the effectiveness of interventions to improve medication adherence in patients with serious mental illness and their effect on patient outcomes, as well as the related costs and interventionspecific harms.

METHODS This review was conducted through the Department of Veterans Affairs Evidence-based Synthesis Program using PRISMA guidelines and focused on medication adherence in patients with psychotic spectrum and bipolar disorders.20 Given the differences between the conditions regarding symptoms, treatment, and reasons for nonadherence, we examined each population separately to facilitate identification of variation of treatment effect by disorder. For each condition, we sought to answer the following key questions: a) What are the effects of medication adherence interventions on psychopharmacological adherence (including long-acting injectable [depot] formulations)? b) What are the effects of medication adherence interventions on adherence to treatment plans for comorbid medical conditions (e.g. pharmacotherapy for conditions such as diabetes)? c) What are the effects of these interventions on patient outcomes? d) What are the harms and costs related to these interventions?

5 A protocol summarizing our methods was published a priori in the PROSPERO trial registry (CRD42015017190). Search Strategy To identify relevant articles, we conducted a systematic search of MEDLINE, PubMed, PsychINFO, Embase, CINAHL, and the Cochrane Library from database inception to January 27, 2015 using keywords and MESH terms related to adherence and the patient populations of interest (detailed search strategies available in the eMethods section of the online supplement). We also evaluated the bibliographies of included primary studies and relevant reviews. Study Selection We included only studies with adult populations examining interventions designed to improve medication adherence in general mental health settings (inpatient and outpatient) that reported both a patient outcome measure and an objective or validated subjective measure of medication adherence. Objective metrics of adherence included pharmacy claims, pill counts, or blood plasma concentration levels. Subjective clinician-rated or self-reported measures quantifying medication adherence using standardized and validated assessments were also eligible (e.g., Morisky Medication Adherence Scales [MMAS]; Medication Adherence Questionnaire [MAQ]; the Drug Attitude Inventory (DAI), which correlates with other measures of medication adherence).21 Studies set in institutional or forensic settings with incarcerated participants were excluded due to limited applicability (including increased supervision and medication distribution). Eligible study designs included trials and methodologically rigorous observational studies, including before/after studies with at least 3 time points and that completed analyses controlling for time. Two independent reviewers evaluated titles and abstracts for a random 10% sample of the search yield in order to ensure reliability between

6 reviewers, with the remaining 90% decided by a single reviewer. At the full-text screening stage, 2 independent reviewers assessed all articles for inclusion. Discordant results were resolved through discussion or consultation with a third reviewer. Data Abstraction and Quality Assessment Data from published reports were abstracted into evidence tables by one investigator and confirmed by the Principal Investigator. Two reviewers independently assessed the quality of each study using a revised version of the risk of bias (ROB) assessment tool22, 23 developed for a recent high-quality effectiveness review examining medication adherence interventions in populations not included in this report.19 The tool included assessments of traditional domains regarding selection (e.g., randomization or enrollment methods), performance (e.g., accounting for concurrent interventions or exposures), and attrition biases, as well as several domains assessing the outcome measurement validity and handling of potential confounders (eTables 3 and 4 of the supplement). Disagreements were resolved through discussion with a third reviewer. Each study was given an overall summary assessment of low, moderate, or high ROB. Data Synthesis We summarized the primary literature by abstracting relevant data and qualitatively synthesizing the literature for each clinical population. Due to substantial heterogeneity in the literature, a meta-analysis was not performed. We assessed the overall strength of evidence (SOE) using a method developed by the Agency for Healthcare Research and Quality (AHRQ) which considers study limitations, directness, consistency, precision, and reporting bias to classify the SOE for individual outcomes independently for randomized controlled trials and observational studies.24 We supplemented our SOE determination with the domains of dose-

7 response association, presence of plausible confounders that would decrease the observed effect, strength of association, and applicability.

RESULTS Our search strategy identified 8,470 references, of which 152 citations were pulled for full-text review (Figure 1). After full-text review, we identified 21 articles (20 individual studies) for psychotic spectrum disorder and 4 studies for bipolar disorder. Psychotic Spectrum Disorders Summary of Results. We identified 20 distinct studies involving individuals with psychotic spectrum disorder (eTable 1 of the supplement). One study25 had a separate article reporting a cost analysis of the intervention.26 Studies were conducted in outpatient community mental health settings and/or inpatient hospital settings, mostly in the U.S. (6 studies) or Europe (12 studies). We found limited evidence suggesting improved adherence associated with interventions involving family members, technology, and those combining depot antipsychotics with another intervention (Table 1). Several studies compared multiple types of interventions and therefore are considered under more than one intervention category. Findings were mixed regarding the effectiveness of multicomponent behavioral interventions, with no conclusive data to support Compliance Therapy (CT)1 or Adherence Therapy (AT).2 In addition, no clear evidence exists to support conclusions regarding the effect of medication adherence interventions on patient 1

Compliance Therapy (CT) is a multicomponent intervention first studied by Hayward in 199527 and described fully in a manual by Kemp in 199728 that combines motivational interviewing (MI), cognitive, and psychoeducation approaches targeting psychotic symptoms, focusing on illness and treatment history, beliefs and understanding of the illness, and ambivalence towards treatment and stigma. 2 Adherence Therapy29 (AT) is a manualized intervention that builds on CT, and is a patient-centered cognitive behavioral approach incorporating MI and emphasizing joint decision-making, medication problem solving, exploring ambivalence, and the discussion of beliefs and concerns about medication.

8 outcomes. Only one randomized controlled trial (RCT) with moderate ROB (supplementary eTables 3 and 4) examined adherence to non-psychiatric medications in patients with psychotic spectrum disorders and found no differences between interventions involving short message service (SMS), telephone support, or their combination.30 No studies reported interventionspecific harms. Family Interventions (SOE: Low). We identified 3 studies (moderate ROB) examining interventions involving family members.31-33 The interventions varied in nature and content, with one study examining a culturally-modified version of multifamily group therapy that included both individual family sessions and a patient-family group,31 and another that included relative and patient psychoeducation groups.32 The third study examined family psychoeducation plus psychosocial skills training administered in weekly group sessions to improve independent and community functioning.33 The 3 studies examining family interventions show a generally positive effect on medication adherence.31-33 Two of the 3 studies found family interventions to be significantly more effective than usual care at improving clinician-, family-, or pharmacy-assessed measures of adherence.32, 33 The third study found that although patients in the family intervention arm had slightly higher adherence, the effect was non-significant after controlling for time.31 All of the family intervention studies reported positive effects on patient outcomes, including a significant decrease in positive and negative symptoms,33 a greater reduction in symptom severity,32, 33 larger decreases in functional impairment,32, 33 fewer hospital admissions or a longer time to readmission or relapse.31-33 Technology Interventions (SOE: Low). Four included studies provide evidence of a positive effect on medication adherence associated with technology interventions.30, 34-36 Two

9 RCTs evaluated electronic monitoring (e-monitoring) using smart pill containers/dispensers versus a variety of comparators. One low ROB study found significantly better adherence when assessed using an e-monitoring system as compared with usual care, but not as assessed by pill counts.34 The second study (high ROB) examined how the method of measuring adherence may influence adherence.36 An e-monitoring program, which required participants to transmit adherence data daily and to discuss any lapse in adherence with study staff, was compared to pill counts and self-reported adherence at study visits using a validated scale. Results indicated that adherence in the e-monitoring group was significantly higher than the groups assigned to both pill counts by a pharmacist and self-report. Two studies with moderate ROB examined SMS and/or telephone interventions and found no difference between the technology interventions but observed significantly better adherence associated with SMS compared with usual care.30, 35 The e-monitoring studies were mixed with respect to patient outcomes. One study with high ROB found consistently positive effects on patient outcomes, with significant effects on the severity of schizophrenia symptoms and psychiatric and emergency department visits.36 However, the other study (low ROB) found no differences between groups on psychiatric symptoms, social and occupational functioning, hospitalizations or emergency department use.34 The 2 studies30, 35 of SMS interventions generally indicated that they had a positive impact on symptom severity and quality of life, although one study found the effects were not sustained 3 months following completion of the intervention.35 Adherence Interventions Combined with a Depot Antipsychotic (SOE: Insufficient). Two studies (moderate ROB) examining interventions for medication adherence in patients prescribed a depot antipsychotic found limited evidence of a positive effect.37, 38 One study compared a depot antipsychotic with treatment as usual to depot plus an intervention for family members,

10 psychoeducation, and early warning sign detection. Results indicated better adherence associated with the intervention as compared to treatment as usual for up to 24 months.38 The second study included depot antipsychotics in combination with a customized multicomponent behavioral intervention, and found that as compared to baseline, adherence was significantly better through 25 weeks; however, this study lacked a control group for comparison.37 The effects of these interventions on patient outcomes were mixed. One study was associated with a reduced risk of relapse and significant improvement on symptom severity and patient functioning; however, there were no significant differences between groups when controlling for time.38 The single-arm study found improvements on several of the same outcomes.37 Pharmacist-led Interventions (SOE: Insufficient). One study (high ROB) examined pharmacist-led informational group sessions and found no impact on nurse assessments of adherence or patient outcomes.39 System-level Interventions (SOE: Insufficient). One RCT (moderate ROB) on a systemlevel intervention found no significant differences in medication adherence or patient outcomes as compared to CT.40 Multicomponent Behavioral Interventions (SOE: Insufficient). Findings of the 7 studies examining multicomponent interventions (i.e., CT, AT, and Cognitive Behavior Therapy [CBT] with Motivational Interviewing [MI]) were mixed. Of the 4 studies examining CT,25, 40-42 one high ROB study found improved adherence at one month but no differences by 6 months followup,42 and another study (high ROB) found significantly greater adherence as compared to baseline for up to 18 months.25 The other 2 studies (moderate ROB) found no significant differences in medication adherence between CT and comparators, and no improvement from

11 baseline adherence for either group.40, 41 Only the study showing a sustained improvement on adherence also found improved patient functioning and insight, and reduced risk of readmission.25 Two low ROB trials of AT found no benefit over comparators on adherence.43, 44 Patient outcomes were mixed, with one study showing a significant improvement in symptom severity and patient functioning at 12-weeks post-discharge,44 and the other showing no effect on psychiatric symptoms or quality of life at 12 months.43 One other RCT (moderate ROB) of inpatients compared CBT plus MI to group psychoeducation plus MI and found no difference in medication adherence or patient outcomes at 24-months follow-up.45 Other Interventions (SOE: Insufficient). Finally, there was insufficient evidence to determine the effectiveness of other interventions such as MI, Cognitive Adaption Therapy (CAT), and shared decision making on improving medication adherence for people with psychotic spectrum disorders. Two studies (one low ROB,34 one moderate ROB46) found that medication adherence-focused CAT (Pharm-CAT), an intervention focused on individualized strategies and environmental supports, was better than usual care over 15 months as evaluated by pill counts.34, 46 Studies involving MI (low ROB)47 and shared decision making (high ROB)48 failed to demonstrate significantly improved medication adherence compared to control. No intervention significantly improved clinical outcomes. Bipolar Disorder Summary of Results. Four moderate ROB studies in patients with bipolar I or bipolar II disorders met inclusion criteria (eTable 2 of the supplement).49-52 Two studies were conducted in Iran,51, 52 while the other 2 were conducted in community health centers in the U.S.49, 50 Two

12 studies examined psychoeducation,51, 52 and another examined the Life Goals Program,50 which includes psychoeducation and individualized problem solving skills with a focus on selfmanagement. The fourth study evaluated customized adherence enhancement (CAE), a customized behavioral multicomponent intervention in which participants were assigned to between one to four modules (psychoeducation, substance use/modified Motivational Enhancement Therapy [MET], provider communication, and/or medication management) based on a baseline assessment.49 Interventions aimed at patients with bipolar disorder found some improvement in medication adherence and were associated with improvements in several clinical outcomes; however, the strength of evidence was insufficient from which to draw conclusions (Table 2). There were no studies examining depot antipsychotics, adherence to treatment for comorbid medical conditions, harms, or costs in patients with bipolar disorder. Psychoeducation (SOE: Insufficient). The 2 small RCTs examining psychoeducation, delivered both individually in concert with psychotherapy51 and within groups,52 resulted in improved adherence and clinical outcomes. Individual psychoeducation was associated with reduced hospital readmissions and recurrence of depression and mania.51 Group psychoeducation was associated with improvement in patient functioning.52 Psychoeducation Plus Problem Solving (SOE: Insufficient). One RCT evaluated a manualized structured group psychotherapy program that included psychoeducation and individualized problem solving.50 The trial reported no significant differences in adherence at 3, 6, and 12 months as compared to treatment as usual, nor were there differences in depression and mania severity or patient functioning.

13 Multicomponent Behavioral Intervention (SOE: Insufficient). One uncontrolled (singlearm) study evaluating CAE in patients with bipolar disorder found improved adherence and several clinical outcomes.49

DISCUSSION The studies examining interventions to improve medication adherence in individuals with psychotic spectrum or bipolar disorders evaluated a wide range of heterogeneous interventions and found mixed results. In individuals with psychotic spectrum disorders, there was limited evidence to support improved psychopharmacological adherence associated with interventions involving family members or technology, and a potential for benefit associated with combining depot antipsychotics with another adherence intervention. However, these findings must be interpreted with caution, given the heterogeneity among interventions and settings, the difficulty in determining the contribution of the depot antipsychotic versus the adherence intervention, methodological limitations, and the lack of consistent replication of any specific intervention. Findings were mixed regarding the effectiveness of multicomponent behavioral interventions, with insufficient evidence to determine the effectiveness of AT or CT. In addition, no clear evidence exists to support conclusions regarding the effect of medication adherence interventions on patient outcomes. Very few studies examined interventions for medication adherence in individuals with bipolar disorder; while in general these interventions appear beneficial to psychopharmacological adherence in this population, interventions were heterogeneous. With the exception of interventions involving technology and system-level interventions, many interventions include behavioral techniques that are flexible and designed to adapt to different settings and patients. While evidence supports the use of these techniques for the

14 treatment of other mental health conditions such as anxiety, depression, and substance abuse, additional research of standardized forms of these interventions for medication adherence is needed to replicate findings across settings and populations in order to better understand their effect on adherence and patient outcomes. Similarly, many of the included studies compare medication adherence interventions to usual care, rather than an active comparator. Given the nature of mental illness, the intervention group may experience greater benefit as a result of more frequent provider interaction alone, making uncontrolled studies difficult to synthesize and interpret. Finally, findings may not extrapolate to all populations with psychotic spectrum or bipolar disorder because individuals recruited for these studies – particularly in studies examining long-acting injectable depot antipsychotics – often have especially poor baseline medication adherence. Our review has several limitations. Many of the studies were small and likely underpowered for several outcomes. Ideally, statistical power can be enhanced in systematic reviews through meta-analysis. However, heterogeneity of study design, measures, interventions, and populations precluded us from performing a formal quantitative combining of studies. In addition, we included studies using both objective and validated subjective measures due to the lack of a gold standard measure of adherence and the range of common outcomes utilized in the field. As a result, many of the outcomes were indirect measures of adherence (e.g., self-report measures, pharmacy claims), and despite restricting our inclusion to only studies using validated subjective measures, the accuracy of these indirect measures is not entirely clear. Finally, although we did conduct a search for grey literature, we were unable to conduct a formal assessment of publication bias.

15 Future Research Interventions to improve medication adherence in individuals with psychotic spectrum and bipolar disorders warrant further investigation, particularly in the form of well-designed RCTs with active comparators, and with adequate duration and the power to detect differences. Furthermore, replication is needed to draw conclusions about the effectiveness of specific programs. Several small studies suggest the effectiveness of some of the interventions, such as interventions including family members. However, many of the interventions are multicomponent and complex, differ widely in their content and implementation, and research evaluating standardized interventions is needed. Because individuals with serious mental illness such as schizophrenia or bipolar disorder have substantially elevated rates of cardiovascularrelated mortality,53 future interventions should incorporate initiatives to improve adherence for both psychopharmacologic and non-psychopharmacologic medications. Reporting costs or potential harms was also very limited. Future studies should assess potential adverse effects and costs associated with medication adherence interventions. Finally, objective measures should be used to measure medication adherence, and the identification and validation of a gold standard assessment tool for medication adherence is warranted. CONCLUSION Interventions to improve adherence in patients with psychotic spectrum disorders and bipolar disorder are heterogeneous and inconsistent with respect to their impact on medication taking and clinical outcomes. While the strength of evidence was low, interventions involving family members or technology, such as text messages or e-monitoring, showed the most consistent effect on medication adherence in patients with psychotic spectrum disorders.

16 ACKNOWLEDGEMENTS AND DISCLOSURES Funding Support: U.S. Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Quality Enhancement Research Initiative. The U.S. Department of Veterans Affairs had no role in the conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication. Disclaimer: This report is based on research conducted by the Evidence-based Synthesis Program (ESP) Center located at the Portland VA Health Care System, Portland, OR, funded by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Quality Enhancement Research Initiative. The findings and conclusions in this document are those of the authors who are responsible for its contents; the findings and conclusions do not necessarily represent the views of the Department of Veterans Affairs or the United States government. Therefore, no statement in this article should be construed as an official position of the Department of Veterans Affairs. Conflicts of Interest: The authors report no proprietary or commercial interest in any product mentioned or concept discussed in this article. Registration: PROSPERO CRD42015017190 (http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42015017190). The contents of this paper have not been publically published elsewhere. A technical report has been published internally within the VA intranet, to be made publically available upon publication of the journal manuscript.

17 REFERENCES

1. Robinson D, Woerner MG, Alvir JM, et al. Predictors of relapse following response from a first episode of schizophrenia or schizoaffective disorder. Arch Gen Psychiatry. 1999;56(3):241-7. 2. Gonzalez-Pinto A, Mosquera F, Alonso M, et al. Suicidal risk in bipolar I disorder patients and adherence to long-term lithium treatment. Bipolar Disord. 2006;8(5 pt 2):618-24. 3. Gianfrancesco FD, Sajatovic M, Rajagopalan K, Wang R-H. Antipsychotic treatment adherence and associated mental health care use among individuals with bipolar disorder. Clin Ther. 2008;30(7):1358-74. 4. Velligan DI, Weiden PJ, Sajatovic M, et al. The Expert Consensus Guideline Series: Adherence problems in patients with serious and persistent mental illness. J Clin Psychiatry. 2009;70(Suppl 4):1-48. 5. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med. 2005;353(12):1209-23. 6. Moritz S, Peters MJ, Karow A, Deljkovic A, Tonn P, Naber D. Cure or curse? Ambivalent attitudes towards neuroleptic medication in schizophrenia and non-schizophrenia patients. Ment. 2009;1(1):e2. 7. Diaz E, Levine HB, Sullivan MC, et al. Use of the Medication Event Monitoring System to estimate medication compliance in patients with schizophrenia. J Psychiatry Neurosci. 2001;26(4):325-9. 8. Connolly KR, Thase ME. The clinical management of bipolar disorder: a review of evidence-based guidelines. Prim Care Companion CNS Disord. 2011;13(4). 9. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353(5):487-97. 10. Scott J, Pope M. Nonadherence with mood stabilizers: Prevalence and predictors. J Clin Psychiatry. 2002;63(5):384-90. 11. Vieta E, Azorin J-M, Bauer M, et al. Psychiatrists' perceptions of potential reasons for non- and partial adherence to medication: Results of a survey in bipolar disorder from eight European countries. J Affect Disord. 2012;143(1-3):125-30. 12. Newcomer J. Medical risk in patients with bipolar disorder and schizophrenia. J Clin Psychiatry. 2006;67(Suppl 9):25-30; discussion 6-42. 13. Dolder CR, Lacro JP, Jeste DV. Adherence to antipsychotic and nonpsychiatric medications in middle-aged and older patients with psychotic disorders. Psychosom Med. 2003;65(1):156-62. 14. Piette JD, Heisler M, Ganoczy D, McCarthy JF, Valenstein M. Differential medication adherence among patients with schizophrenia and comorbid diabetes and hypertension. Psychiatr Serv. 2007;58(2):207-12. 15. Steiner J, Earnest M. The language of medication-taking. Ann Intern Med. 2000;132(11):926-30. 16. Sendt K-V, Tracy DK, Bhattacharyya S. A systematic review of factors influencing adherence to antipsychotic medication in schizophrenia-spectrum disorders. Psychiatry Res. 2015;225(1-2):14-30.

18 17. Sajatovic M, Valenstein M, Blow FC, Ganoczy D, Ignacio RV. Treatment adherence with antipsychotic medications in bipolar disorder. Bipolar Disord. 2006;8(3):232-41. 18. Doshi J, Zhu J, Lee B, Kimmel S, Volpp K. Impact of a prescription copayment increase on lipid-lowering medication adherence in veterans. Circulation. 2009 119(3):390-7. 19. Viswanathan M, Golin CE, Jones CD, et al. Closing the quality gap: revisiting the state of the science (vol. 4: medication adherence interventions: comparative effectiveness). Evid rep/technol assess. 2012(208.4):1-685. 20. Kondo K, Low A, Jindai K, et al. Interventions to improve pharmacological adherence among adults with psychotic spectrum disorders, bipolar disorder, and posttraumatic stress disorder. VA ESP Project #05-225; 2015 [May 19, 2016]; Available from: (Internal VA access only) http://www.hsrd.research.va.gov/publications/esp/reports.cfm. 21. Nguyen TM, La Caze A, Cottrell N. What are validated self-report adherence scales really measuring?: a systematic review. Br J Clin Pharmacol. 2014;77(3):427-45. 22. Viswanathan M, Berkman N, Dryden D, Hartling L. Assessing risk of bias and confounding in observational studies of interventions or exposures: Further development of the RTI Item Bank. Rockville, MD: Agency for Healthcare Research and Quality; Methods Research Report (AHRQ Publication No. 13-EHC106-EF), 2012. 23. Viswanathan M, Ansari M, Berkman N, et al. Assessing the risk of bias of individual studies in systematic reviews of health care interventions. Rockville, MD: Agency for Healthcare Research and Quality; Methods Guide for Comparative Effectiveness Reviews (AHRQ Publication No. 12-EHC047-EF), 2012. 24. Berkman N, Lohr K, Ansari M, et al. Grading the strength of a body of evidence when assessing health care interventions for the effective health care program of the Agency for Healthcare Research and Quality: An update. Rockville, MD: Agency for Healthcare Research and Quality; Methods guide for comparative effectiveness reviews (AHRQ Publication No. 13(14)-EHC130-EF), 2013. 25. Kemp R, Kirov G, Everitt B, Hayward P, David A. Randomised controlled trial of compliance therapy: 18-month follow-up. Br J Psychiatry. 1998;172:413-9. 26. Healey A, Knapp M, Astin J, et al. Cost-effectiveness evaluation of compliance therapy for people with psychosis. Br J Psychiatry. 1998;172:420-4. 27. Hayward P, Chan N, Kemp R, Youle S. Medication self-management: A preliminary report on an intervention to improve medication compliance. J Ment Health. 1995;4(5):511-7. 28. Kemp R, Hayward P, David A. Compliance therapy manual. London: King's College School of Medicine and Dentistry and Institute of Psychiatry; 1997. 29. Gray R. Adherence therapy: Working together to improve health. A treatment manual for healthcare workers: Available at: http://www.academia.edu/2436503/Adherence_therapy_manual. 30. Beebe L, Smith KD, Phillips C. A comparison of telephone and texting interventions for persons with schizophrenia spectrum disorders. Issues Ment Health Nurs. 2014;35(5):323-9. 31. Kopelowicz A, Zarate R, Wallace CJ, Liberman RP, Lopez SR, Mintz J. The ability of multifamily groups to improve treatment adherence in Mexican Americans with schizophrenia. Arch Gen Psychiatry. 2012;69(3):265-73. 32. Pitschel-Walz G, Bauml J, Bender W, Engel RR, Wagner M, Kissling W. Psychoeducation and compliance in the treatment of schizophrenia: Results of the Munich Psychosis Information Project Study. J Clin Psychiatry. 2006;67(3):443-52.

19 33. Valencia M, Rascon ML, Juarez F, Escamilla R, Saracco R, Liberman RP. Application in Mexico of psychosocial rehabilitation with schizophrenia patients. Psychiatry. 2010;73(3):24863. 34. Velligan D, Mintz J, Maples N, et al. A randomized trial comparing in person and electronic interventions for improving adherence to oral medications in schizophrenia. Schizophr Bull. 2013;39(5):999-1007. 35. Montes JM, Medina E, Gomez-Beneyto M, Maurino J. A short message service (SMS)based strategy for enhancing adherence to antipsychotic medication in schizophrenia. Psychiatry Res. 2012;200(2-3):89-95. 36. Frangou S, Sachpazidis I, Stassinakis A, Sakas G. Telemonitoring of medication adherence in patients with schizophrenia. Telemedicine and e-Health. 2005;11(6):675-83. 37. Sajatovic M, Levin J, Ramirez LF, et al. Prospective trial of customized adherence enhancement plus long-acting injectable antipsychotic medication in homeless or recently homeless individuals with schizophrenia or schizoaffective disorder. J Clin Psychiatry. 2013;74(12):1249-55. 38. Lee S-H, Choi TK, Suh S, et al. Effectiveness of a psychosocial intervention for relapse prevention in patients with schizophrenia receiving risperidone via long-acting injection. Psychiatry Res. 2010;175(3):195-9. 39. Kavanagh K, Duncan-McConnell D, Greenwood K, Trivedi P, Wykes T. Educating acute inpatients about their medication: is it worth it? An exploratory study of group education for patients on a psychiatric intensive care unit. J Ment Health. 2003;12(1):71-80. 40. Skarsholm H, Stoevring H, Nielsen B. Effect of a system-oriented intervention on compliance problems in schizophrenia: a pragmatic controlled trial. Schizophr Res Treatment. 2014;2014:789403. 41. O'Donnell C, Donohoe G, Sharkey L, et al. Compliance therapy: a randomised controlled trial in schizophrenia. British Medical Journal. 2003;327(7419):834. 42. Byerly MJ, Fisher R, Carmody T, Rush A. A trial of Compliance Therapy in outpatients with schizophrenia or schizoaffective disorder. J Clin Psychiatry. 2005;66(8):997-1001. 43. Gray R, Leese M, Bindman J, et al. Adherence therapy for people with schizophrenia. European multicentre randomised controlled trial. Br J Psychiatry. 2006;189:508-14. 44. Schulz M, Gray R, Spiekermann A, Abderhalden C, Behrens J, Driessen M. Adherence therapy following an acute episode of schizophrenia: A multi-centre randomised controlled trial. Schizophr Res. 2013;146(1-3):59-63. 45. Bechdolf A, Kohn D, Knost B, Pukrop R, Klosterkotter J. A randomized comparison of group cognitive-behavioural therapy and group psychoeducation in acute patients with schizophrenia: Outcome at 24 months. Acta Psychiatr Scand. 2005;112(3):173-9. 46. Velligan DI, Diamond PM, Mintz J, et al. The use of individually tailored environmental supports to improve medication adherence and outcomes in schizophrenia. Schizophr Bull. 2008;34(3):483-93. 47. Barkhof E, Meijer CJ, de Sonneville LM, Linszen DH, de Haan L. The effect of motivational interviewing on medication adherence and hospitalization rates in nonadherent patients with multi-episode schizophrenia. Schizophr Bull. 2013;39(6):1242-51. 48. Hamann J, Cohen R, Leucht S, Busch R, Kissling W. Shared decision making and longterm outcome in schizophrenia treatment. J Clin Psychiatry. 2007;68(7):992-7. 49. Sajatovic M, Levin J, Tatsuoka C, et al. Six-month outcomes of customized adherence enhancement (CAE) therapy in bipolar disorder. Bipolar Disord. 2012;14(3):291-300.

20 50. Sajatovic M, Davies MA, Ganocy SJ, et al. A comparison of the life goals program and treatment as usual for individuals with bipolar disorder. Psychiatr Serv. 2009;60(9):1182-9. 51. Javadpour A, Hedayati A, Dehbozorgi G-R, Azizi A. The impact of a simple individual psycho-education program on quality of life, rate of relapse and medication adherence in bipolar disorder patients. Asian J Psychiatr. 2013;6(3):208-13. 52. Bahredar MJ, Asgharnejad Farid AA, Ghanizadeh A, Birashk B. The efficacy of psychoeducational group program on medication adherence and global functioning of patients with bipolar disorder type I. Int J Community Based Nurs Midwifery. 2014;2(1):12-9. 53. Newcomer J, Hennekens C. Severe mental illness and risk of cardiovascular disease. Jama. 2007;298(15):1794-6.

21 Figure 1. Literature Flow Diagram

Search results: a 8,470 references

Excluded = 8,318 references

Pulled for full-text review: 152 references

Excluded = 127 references  Ineligible publication type or study design: 78 b  Ineligible outcome : 35  Ineligible population or did not stratify by condition: 7  Ineligible intervention: 7

Included studies: 24 (from 25 publications)

Psychotic spectrum disorders: 20 studies (from 21 publications) a

Bipolar disorder: 4 studies

7,895 were identified through database searches, 518 were identified from grey literature searches of registry websites, and an additional 57 were identified from the bibliographies of relevant systematic reviews and primary studies. b Includes studies that did not report both a clinical outcome and an eligible medication adherence outcome.

22 Table 1. Psychotic Spectrum Disorders: Summary of Medication Adherence Outcomes Type of Intervention Family Interventions

Study Design (Combined N)a Findings 3 RCTs Better adherence with family interventions as (N=497)31-33 measured by clinician rating/blood plasma and pharmacy records/family-report as compared to usual care in 2 studies (moderate ROB). No difference when controlling for time in a third study examining a culturally modified family intervention as compared to the standard family intervention and monthly sessions (moderate ROB).

Technology Interventions

4 RCTs (N=534)30, 34-36

Adherence Intervention Plus Depot Antipsychotic

1 Trial (randomization unclear) (N=57)38; 1 Prospective Cohort (N=30)37

Pharmacist-led Intervention

1 Prospective Cohort w/post hoc comparison (N=30)39

System-level Intervention

1 NRCT (N=70)40

Strength of Evidence Comments Low Heterogeneity among interventions.

Mixed findings on e-monitoring/MEMS: better Low adherence in one study as compared to pill counts and self-reported adherence (high ROB), conflicting results in one study as compared to a pharmacy-based intervention and usual care (low ROB). Telephone plus SMS resulted in nonsignificant adherence improvement versus telephone or SMS alone (moderate ROB); SMS alone resulted in significantly better adherence than usual care (moderate ROB). Findings indicated improved adherence related Insufficient to the use of depot antipsychotic injections plus a behavioral multicomponent intervention (compared to usual care or no comparator) as measured by injection visits up to one year, and TRQ, Morisky scale, DAI, and AMQ up to 25 weeks (moderate ROB). No significant difference over time or between Insufficient groups (high ROB).

Mixed findings and heterogeneous interventions.

Heterogeneity among interventions; ROB due to study design.

Evidence from only one study; potential ROB due to study design flaws. Nonsignificant trend towards better adherence Insufficient Evidence from for the system-level intervention, compared with only one study. Compliance Therapy (moderate ROB).

Multicomponent Behavioral Interventions Behavioral 2 RCTs Mixed findings: one study (low ROB) reported Insufficient Evidence from Multicomponent (N=570)43, 44 better adherence compared to usual care on the only 2 studies, - Adherence MAQ and SAI-C at 12 months, and the other with mixed Therapy (low ROB) reporting no difference from usual findings. care on the CDR, DAI-30, and MARS at 12 weeks post-discharge. Behavioral 2 RCTs Mixed findings: better MARS scores with Insufficient Inconsistent Multicomponent (N=130)25, 41; Compliance Therapy at one month but not 6 findings among - Compliance 1 NRCT months in one study (high ROB); better DAI and 4 studies. ROB Therapy (N=70)40; adherence scores as compared with routine due to study 1 Prospective management plus supportive counseling through design. Cohort (N=30)42 18-month follow-up in one study (high ROB); no benefit to Compliance Therapy up to 6

23

Type of Intervention

Study Design (Combined N)a

Other 1 RCT (N=88)45 Behavioral Multicomponent

Findings months in 2 studies (compared to nonspecific counseling and Compliance Therapy; moderate ROB). No difference between CBT plus MI and group psychoeducation plus MI groups (moderate ROB).

Other Interventions Motivational 1 RCT Interviewing (N=114)47 (MI)

One study found no benefit of MI over usual care as measured by the MAQ or DAI (low ROB).

Cognitive Adaptation Training (CAT)

2 RCTs (N=247)34, 46

Shared Decision 1 RCT Making (N=107)48

Strength of Evidence

Comments

Insufficient Evidence from only one study.

Insufficient Evidence from only one study.

One study found that both CAT and Pharm-CAT Insufficient Evidence from resulted in better adherence than usual care, with 2 studies that no difference between the two (moderate ROB). used different The second study comparing Pharm-CAT to ecomparators. monitoring reported mixed results (low ROB). One study found no benefit to a shared decision Insufficient Evidence from making over usual care as measured by the only one single MARS and plasma levels (high ROB). study.

Abbreviations: AMQ, Attitude towards Medication Questionnaire; CAT, Cognitive Adaptation Training; CBT, cognitive behavioral therapy; CDR, concentration to dose ratio; DAI, Drug Attitude Inventory; e-Monitoring, electronic monitoring; MAQ, Medication Adherence Questionnaire; MARS, Medication Adherence Rating Scale; MEMS, Medication Event Monitoring System; MI, Motivational Interviewing; N, number randomized; NRCT, non-randomized controlled trial; RCT, randomized controlled trial; ROB, risk of bias; SAI-C, Schedule for the Assessment of Insight - C; SMS, short message service; TRQ, Tablet Routine Questionnaire. a

Studies comparing interventions may be accounted for more than once.

Table 2. Bipolar Disorder: Summary of Medication Adherence Outcomes Type of Intervention Psychoeducation (individual/group)

Psychoeducation Plus Problem Solving

Study Design (Combined N) 1 RCT (N=108)51; 1 NRCT (N=45)52

1 RCT (N=164)50

Strength of Findings Evidence Comments Both individual and group Insufficient Evidence from only 2 studies, psychoeducation resulted in external validity concerns due better medication adherence to setting. than pharmacotherapy alone or pharmacotherapy with standard psychotherapy (moderate ROB). There was no improvement in Insufficient Evidence from only one medication adherence associated study. Only 75% of the with the intervention as intervention group and 81% compared to usual care of the control group (moderate ROB). participated in baseline plus one other assessment, and only 49% of the intervention group participated in most or all of the group sessions, with 37% never participating.

24 Customized Behavioral Multicomponent

1 Prospective Customized adherence Insufficient Evidence from only one Cohort enhancement (CAE) a was study; ROB due to study (N=43)49 associated with better adherence design. and attitudes towards medication at 3 and 6 months (moderate ROB).

Abbreviations: CAE, customized adherence enhancement; N, number randomized; NRCT, non-randomized controlled trial; RCT, randomized controlled trial; ROB, risk of bias. a

CAE is a manualized individual behavioral intervention consisting of four modules (psychoeducation, substance use/modified Motivational Enhancement Therapy (MET), provider communication, medication management). Individuals are assigned to one or more of these modules based on an assessment at baseline.