BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

Psychological interventions for positive symptoms in : protocol for a network meta-analysis of randomized evidence

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-019280

Article Type: Protocol

Date Submitted by the Author: 22-Aug-2017

Complete List of Authors: Bighelli, Irene; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Salanti, Georgia; University of Bern, Institute of Social and Preventive Medicine (ISPM) Reitmeir, Cornelia; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Wallis, Sofia; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Barbui, Corrado; University of Verona, WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation, Department of Neuroscience, Biomedicine and Movement Sciences, Section of Psychiatry

Furukawa, Toshi; Kyoto University, Department of Health Promotion and http://bmjopen.bmj.com/ Human Behavior, Graduate School of Medicine and School of Public Health Leucht, Stefan; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy

Schizophrenia & psychotic disorders < PSYCHIATRY, Adult psychiatry < Keywords: PSYCHIATRY, Psychological interventions, Network Meta-Analysis, Psychotherapy

on September 24, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Psychological interventions for positive symptoms in 4 5 schizophrenia: protocol for a network meta-analysis of 6 7 randomized evidence 8 9 10 11 12 Irene Bighelli1*, Georgia Salanti2, Cornelia Reitmeir1, Sofia Wallis1, Corrado Barbui3, Toshi A. 13 Furukawa4, Stefan Leucht1 14 15 16 For peer review only 17 18 19 1 Department of Psychiatry and Psychotherapy, Klinikum rechts der Isar, Technical University of 20 21 Munich, Germany 22 23 2 Institute of Social and Preventive Medicine (ISPM), University of Bern, Switzerland 24 25 3 WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation, 26 27 Department of Neuroscience, Biomedicine and Movement Sciences, Section of Psychiatry, University 28 of Verona, Italy 29 30 4 Department of Health Promotion and Human Behavior, Kyoto University Graduate School of 31 Medicine / School of Public Health, Japan 32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 *Corresponding author 40 Department of Psychiatry and Psychotherapy on September 24, 2021 by guest. Protected copyright. 41 Klinikum rechts der Isar, Technische Universität München 42 Ismaningerstr. 22, 81675 München, Germany 43 44 E-mail: [email protected] 45 46 GS: [email protected] 47 CR: [email protected] 48 SW: [email protected] 49 50 CB: [email protected] 51 TAF: [email protected] 52 SL: [email protected] 53 54

55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 ABSTRACT 4 5 Introduction 6 There is rising awareness that we need multi-disciplinary approaches integrating psychological 7 treatments for schizophrenia, but a comprehensive evidence base on their relative efficacy is lacking. 8 We will conduct a network meta-analysis (NMA), integrating direct and indirect comparisons from 9 10 randomised controlled trials (RCTs) to rank psychological treatments for schizophrenia according to 11 their efficacy, acceptability and tolerability. 12 13 Methods and analysis 14 15 We will include all RCTs comparing a psychological treatment aimed at positive symptoms of 16 schizophrenia withFor another psychologicalpeer intervention review or with a noonly treatment condition (waiting list, 17 treatment as usual). We will include studies on adult patients with schizophrenia, excluding specific 18 subpopulations (e.g. prodromal or first episode patients). Primary outcome will be the change in 19 20 positive symptoms on a published rating scale. Secondary outcomes will be acceptability (dropout), 21 change in overall and negative symptoms of schizophrenia, response, relapse, adherence, 22 , quality of life, functioning and adverse events. Published and unpublished studies will be 23 sought through database searches, trial registries and websites. Study selection and data extraction 24 will be conducted by at least two independent reviewers. We will conduct random-effects NMA to 25 26 synthesize all evidence for each outcome and obtain a comprehensive ranking of all treatments. 27 NMA will be conducted in Stata and R within a frequentist framework. The risk of bias in studies will 28 be evaluated using the Cochrane Risk of Bias tool and the credibility of the evidence will be evaluated 29 using an adaptation of the GRADE framework to NMA, recommended by the Cochrane guidance. 30 Subgroup and sensitivity analyses will be conducted to assess the robustness of the findings. 31 32 http://bmjopen.bmj.com/ 33 Ethics and dissemination 34 No ethical issues are foreseen. Results from this study will be published in peer-reviewed journals 35 and presented at relevant conferences. 36 37 38 PROSPERO registration number: CRD42017067795 39 40 on September 24, 2021 by guest. Protected copyright. 41 42 STRENGTHS AND LIMITATIONS OF THIS STUDY 43 • We will conduct a random-effects network meta-analysis to synthesize all available published 44 45 or unpublished randomized controlled trials for each pre-specified outcome, and obtain a 46 comprehensive ranking of all treatments. 47 • This will be the first network meta-analysis on psychological treatments for schizophrenia; 48 the findings from this study have the potential to inform and influence clinical decision- 49 making and guideline development. 50 51 • The risk of heterogeneity and inconsistency is high, given the different psychological 52 interventions that will be included: however, we try to control variability by carefully framing 53 the inclusion criteria about population and interventions, and we will evaluate consistency 54 employing local as well as global methods. 55 56 • The limitations of primary studies will be addressed with the Cochrane risk of bias tool and 57 the quality of evidence for network estimates will be assessed with an appropriate 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 adaptation of the GRADE framework; these approaches are considered the gold standard for 4 critical appraisal of evidence quality. 5 6 7 INTRODUCTION 8 Schizophrenia is a debilitating, and often life-long disorder that ranks among the top 20 causes of 9 disability according to the World Health Report (1). Although pharmacological interventions have 10 been the mainstay of treatment for schizophrenia, antipsychotics have a number of limitations 11 12 (limited response, high incidence of disabling side effects, poor adherence to treatment) (2) and are 13 problematic in many situations (such as medical comorbidities, tolerability problems and pregnancy). 14 Besides, there has been growing recognition of the importance of psychological processes in 15 , both as contributors to onset and persistence, and in terms of the negative psychological 16 For peer review only 17 impact of a diagnosis of schizophrenia on the individual’s well-being, psychosocial functioning and 18 life opportunities. Psychological interventions for psychosis and schizophrenia have been developed 19 to address these aspects, and, in accordance with guidelines from the National Institute for Health 20 and Care Excellence in the United Kingdom (3) and the Schizophrenia Patient Outcomes Research 21 Team in the United States (4), psychological treatments are widely regarded as a necessary 22 23 intervention for schizophrenia. 24 A broad range of interventions that can be defined as “psychological” have been studied in the 25 treatment of schizophrenia. These interventions can be provided at different stages of the illness and 26 address different aspects, like social and cognitive functioning, adherence to medication and 27 symptoms of schizophrenia. Table 1 presents the panorama of existing systematic reviews of 28 29 randomized controlled trials that have been conducted on the topic. These reviews have mainly 30 included studies comparing the intervention under examination with so called no treatment 31 conditions (waiting-list, treatment as usual (TAU)) (5, 6). Other reviews included also active 32 comparisons with other psychological treatments (7–9). An attempt to provide information on active http://bmjopen.bmj.com/ 33 comparisons was made by Turner and colleagues, who performed separate meta-analysis when 34 35 there were at least five eligible randomized controlled trials comparing one intervention to another 36 psychological intervention (10). However, all the available reviews applied pairwise meta-analysis as 37 a method, being able to pool results only when a comparison of two treatments was considered in 38 existing studies. The comparative efficacy and tolerability of the existing interventions has not been 39 checked yet; as a result, it is still currently unclear which are the most efficacious, the most 40 on September 24, 2021 by guest. Protected copyright. 41 acceptable and the best tolerable psychological treatments for schizophrenia. 42 To overcome this gap in the current knowledge, a network meta-analysis would be necessary to 43 consider both direct and indirect comparisons, and produce hierarchies of the effects of the various 44 psychological treatments in the various efficacy and tolerability outcomes. Such hierarchies are 45 46 essential for guidelines, which should ideally be able to indicate which treatment is likely to be the 47 best, the second best and so on for a given outcome. Only the method of network meta-analysis can 48 provide such hierarchies by combining all the randomised evidence. Our aim is to produce such a 49 network meta-analysis of all psychological interventions for schizophrenia in multiple outcomes. We 50 focus here on the interventions primarily aimed at treating positive symptoms in the acute phase of 51 52 the illness. 53 54 55 56

57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Intervention Existing reviews RCT* Comparator Acceptance and Ongoing Cochrane TAU, pharmacological intervention, another 4 - 5 commitment therapy review (11) psychosocial intervention 6 Adherence interventions Gray 2016 (12) 6 TAU, didactic health education 7 Active comparisons Befriending, CBT, Cognitive remediation, 8 (Befriending, CBT, Psychoeducation, Social skills training, Supportive 9 Cognitive remediation, Turner 2014 (10) 48 counseling, Family intervention, , Body 10 Psychoeducation, Social psychotherapy, Occupational therapy, Problem- 11 skills training, Supportive solving therapy 12 counseling) 13 Art therapy Ruddy 2005 (6) 2 Standard care 14 Assertive community TAU, hospital-based rehabilitation, case Marshall 1998 (13) 20** 15 treatment management 16 Befriending - For peer review only 17 Bibliotherapy -

18 Body oriented - 19 psychological therapy 20 Assertive Community Treatment, Assertive 21 Case management Dieterich 2017 (14) 40** Outreach model, Case Management model, 22 standard community care 23 Zimmermann 2005 24 (positive symptoms) 15 Waiting-list, TAU or another therapeutic treatment 25 (9) 26 Active (Psychoeducation, Family Intervention, 27 Supportive Psychotherapy, Supportive Counselling, 28 Jones 2012 (7) 20 Cognitive Remediation) and nonactive control 29 treatments (Recreation and support, social 30 activities, befriending, non-specific counselling) 31 Waiting list, TAU or an intervention designed to 32 control for the non-specific effects of http://bmjopen.bmj.com/ 33 psychotherapy (recreation and support, group 34 Cognitive behavioural support, befriending, supportive Jauhar 2013 (15) 52 35 therapy counselling/therapy, social activity therapy and 36 goal-focused supportive contact) or active 37 treatments (cognitive remediation, 38 psychoeducation) 39 Van der Gaag 2014 40 (individually tailored) 18 Any control condition was accepted on September 24, 2021 by guest. Protected copyright. 41 (16) 42 Hazell 2016 (low 8 TAU, supportive psychotherapy 43 intensity) (17) 44 Kennedy 2016 Non-specialized therapy (focused on supportive 45 (auditory 2 interactions and social integration) 46 hallucinations) (18) 47 TAU, active control (e.g. computer games) another Cognitive remediation Cella 2016 (19) 45 48 active treatment (e.g. CBT) 49 Ren 2013 (20) 1 Standard care plus supportive counselling 50 Pitschel-Walz 2001 TAU, patient intervention, other family 25 51 (21) interventions Family interventions 52 TAU, discussion groups, psychoeducation, Pharoah 2006 (22) 25 53 supportive psychotherapy, psychosocial support 54 TAU, other groups (active discussion group, 55 Group psychotherapeutic , counselling group, occupational Orfanos 2015 (23) 34 56 treatments therapy group or problem-solving discussion 57 group) 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 Integrated psychological TAU, placebo-attention condition, other active 3 Roder 2006 (24) 16 4 Therapy (IPT) treatments Psychological and 5 TAU, active comparator (including psychosocial 6 Lutgens 2017 (8) 95 psychoeducation, supportive therapy, cognitive interventions for negative 7 remediation) 8 symptoms in psychosis 9 TAU, Wait list control, Supportive Therapy, 10 Metacognitive training Eichner 2016 (25) 12 Newspaper discussion group, CogPack (=Cognitive 11 remediation) Active control intervention (e.g. Befriending, 12 Mindfulness Aust 2017 (26) 11** 13 Progressive Muscle Relaxation), TAU Placebo defined as an alternative therapy designed 14 Geretsegger 2017 18 to control for effects of the therapist’s attention; 15 (27) TAU or no treatment 16 For peer reviewReality-adaptive, only supportive psychotherapy, 17 Hospital comparison, Ataraxic drugs, electro 18 Psychodynamic therapy Malmberg 2012 (28) 4 convulsive therapy, Milieu therapy, individual vs. 19 group 20 TAU, supportive psychotherapy, behavioral 21 Psychoeducation Pekkala 2006 (29) 10 intervention, leisure-time group 22 TAU, structured activities, discussion group, 23 Social skills training Almerie 2015 (30) 13 interaction group, no treatment control 24 Standard care, any other treatment (biological, 25 psychological or social) such as medication, 26 Supportive therapy Buckley 2015 (31) 24 problem-solving therapy, psychoeducation, social 27 skills training, CBT, or 28 psychodynamic psychotherapy 29 Systemic therapy Pinquart 2016 (5) 7 No treatment 30

31 Table 1. Existing reviews about psychological treatments for schizophrenia. 32 RCT: randomized controlled trial; TAU: treatment as usual. http://bmjopen.bmj.com/ 33 *number of RCTs on schizophrenic patients; **RCTs about patients with severe mental illness including schizophrenia 34 35 36 Objectives 37 38 To estimate relative treatment effects and obtain a hierarchy for the psychological treatments in 39 patients with schizophrenia, in terms of: 40

1. Efficacy on positive symptoms on September 24, 2021 by guest. Protected copyright. 41 2. Acceptability 42 43 3. Other efficacy measures, such as overall symptoms, negative symptoms, response, relapse, 44 adherence, depression, quality of life, functioning 45 4. Tolerability. 46 47 48 METHODS AND ANALYSIS 49 Criteria for considering studies for this review 50 Methods for this systematic review have been developed according to the Preferred Reporting Items 51 52 for Systematic review and Meta-Analysis Protocols (PRISMA-P) checklist, and the PRISMA extension 53 statement for reporting of systematic reviews incorporating network meta-analyses of healthcare 54 interventions (32, 33). This systematic review and NMA is registered in the PROSPERO database 55 (registration number: CRD42017067795); the record in PROSPERO will be updated with any 56 amendment made to the protocol. 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 4 Types of studies 5 6 We will include all randomised trials (RCTs) in which participants with schizophrenia received a 7 psychological intervention as defined below (see Types of interventions). Studies whose sequence 8 generation was at high risk of bias (e.g. randomization by the date of birth or day of the week) will be 9 excluded. We will accept open and blinded RCTs; this choice is particularly relevant in trials on 10 psychological interventions, in which in best case only the assessor of outcome can be blind, but not 11 12 the therapist. Open RCTs will be excluded in a sensitivity analysis. We will include both trials in which 13 psychological interventions were compared with a control condition and trials in which they were 14 compared with another intervention. There will be no language restriction in order to avoid the 15 problem of ‘language bias’ (34). As an exception, we will exclude studies conducted in mainland 16 China (while studiesFor conducted peer in Taiwan andreview Hong Kong will not only be excluded) to avoid a systematic 17 18 bias, since many of these studies do not use appropriate randomisation procedures and often do not 19 report their methods (35, 36). In the case of cross-over studies we will use only the first cross-over 20 phase in order to avoid the problem of carry-over effects which are very likely in schizophrenia and 21 with psychological treatments. We will exclude cluster randomized trials. 22

23 24 Types of participants 25 Our aim is to collect information on the efficacy of psychological treatments on patients with positive 26 symptoms. In order to select this population, we operationalized the inclusion criteria as follows. We 27 will include adults, however defined by study authors, with a diagnosis of schizophrenia or related 28 29 disorders (such as schizophreniform, or schizoaffective disorders); there is no clear evidence that the 30 latter schizophrenia-like psychoses are caused by fundamentally different disease processes or 31 require different treatment approaches (37). We will include trials irrespective of the diagnostic 32 criteria used. Here we will follow the strategy of the Cochrane Schizophrenia Group (38) to include http://bmjopen.bmj.com/ 33 not only studies that used specific diagnostic criteria such as ICD-10 or DSM-V, because these criteria 34 35 are not meticulously used in clinical routine either. This decision should increase generalizability and 36 representativeness. 37 Studies including participants with other diagnoses part of the psychosis spectrum will be included 38 only if participants with a diagnosis of schizophrenia, schizophreniform or schizoaffective disorders 39 were more than 80% of the participants considered. We will include studies recruiting patients with 40 on September 24, 2021 by guest. Protected copyright. 41 positive symptoms, either , hallucinations or both, or acute phase of illness, however 42 defined by inclusion criteria of the trial. 43 We will exclude studies focused on specific subpopulations of patients, such as (1) studies recruiting 44 patients with predominant negative symptoms, (2) studies on patients with comorbid psychiatric 45 46 disorders including substance abuse, (3) studies recruiting patients with concomitant medical 47 illnesses, (4) trials enrolling stable patients ( studies), (5) studies on first episode 48 patients, (6) trials on patients who show prodromal signs of psychosis (also defined as “at risk for 49 psychosis”). 50

51 52 Types of interventions 53 Any psychological intervention that occurs through interaction between therapist and patient, either 54 face-to-face individually or in group, with the primary aim to reduce positive symptoms. 55 Interventions with an explicit primary aim different from positive symptoms (e.g. functioning, 56 57 cognition, adherence to medication, knowledge of the illness) will be excluded. Psychological 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 treatments will be compared to each other and to any non-pharmacological control condition 4 considered in the included studies. Comparators will include the so called “treatment as usual”, 5 waiting list and inactive treatments. “Non-active” comparators have been associated with a nocebo 6 effect, therefore their effect on the network will be analysed in a sensitivity analysis (39). Patients 7 8 also receiving treatment as usual, including pharmacological interventions will be included. If 9 psychological treatments that we do not include among the interventions (e.g. psychoeducation, 10 supportive therapy) are used as control condition in the studies, they will be included as nodes in 11 order to strengthen the network, but will not be part of our decision set. 12 13 14 Outcome measures 15 Outcomes will be measured at study endpoint, as defined in each study. 16 Primary outcomeFor peer review only 17 18 Change in positive symptoms of schizophrenia, examined accordingly to the respective subscale of 19 20 the Positive and Negative Syndrome Scale (PANSS), the Brief Psychiatric Rating Scale (BPRS) or the 21 Scales for Assessment of Positive Symptoms (SAPS) or any other published scale. 22 As not all studies will have used the same scale, we will extract data according to the following 23 hierarchy: mean change of the PANSS positive symptoms subscale from baseline to endpoint, if not 24 available mean change of the BPRS positive symptoms subscale, or if again not available the mean 25 26 values at endpoint of the PANSS/BPRS positive symptoms subscale. The results of other rating scales 27 will only be used if the instrument has been published in a peer-reviewed journal, because it has 28 been shown that non-validated schizophrenia scales exaggerate differences (40). 29 30 31 Secondary outcomes 32 1. Acceptability, defined as the percentage of patients leaving the study early (‘dropout’) for http://bmjopen.bmj.com/ 33 any reason. All-cause discontinuation due to any reason combines efficacy, tolerability, 34 and other factors and can therefore be considered as a measure of ‘acceptability of 35 treatment’ (38) or of overall “effectiveness”; 36 37 2. Change in overall symptoms, measured by rating scales such as the PANSS or the BPRS, or 38 any other published scale (e.g. the Manchester Scale) for the assessment of overall 39 schizophrenic symptomatology. The results of other rating scales will only be used if the 40

instrument has been published in a peer-reviewed journal; on September 24, 2021 by guest. Protected copyright. 41 3. Change in negative symptoms, measured by the respective subscale of the PANSS, or the 42 43 “Scales for Assessment of Negative Symptoms” (SANS) or any other published scale; 44 4. Response, measured by the percentage of responders defined by reduction on the PANSS, 45 BPRS or CGI scores, accepting the criteria used by study authors; 46 5. Percentage of patients with relapse, by definitions operationalized by rating scales, and, if 47 not available, number of rehospitalisations due to psychopathology. We will not include 48 49 data from studies that used non-operationalized relapse criteria (e.g. “clinical 50 judgement”); 51 6. Adherence, measured by any published rating scale (e.g. “Adherence Therapy Patients 52 Satisfaction Questionnaire”, “Adherence Rating Scale”); 53 7. Depression, measured by the Calgary Depression Scale for Schizophrenia, the Hamilton 54 55 Depression Rating Scale, the Montgomery Asberg Depression Scale or other published 56 symptom scales; 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 8. Quality of life, measured by any published rating scale (e.g. “Heinrichs quality of life 4 scale”, Quality of Life Scale (QOLS); 5 9. Functioning, measured by rating scales such as the Global Assessment of Functioning or 6 the Psychosocial Performance Scale, or any other published rating scale. 7 8 10. Tolerability, measured as the percentage of patients experiencing adverse events. 9 Adverse events associated with psychological treatments are not covered as 10 comprehensively as in trials on pharmacological treatments (41). However, there is a 11 raising awareness of the importance of considering possible harms associated with 12 psychological interventions (42). Therefore we will collect any available information in 13 14 clinical studies about this outcome, using a classification proposed by Linden and 15 colleagues (42): a) emergence of new symptoms; b) deterioration of existing symptoms; c) 16 lack ofFor improvement peer or deterioration review of illness; d) prolongation only of treatment; e) patient’s 17 non-compliance; f) strains in the patient-therapist relationship; g) very good patient- 18 therapist relationship, therapy dependency; h) strains or changes in family relations; i) 19 20 strains or changes in work relations; l) any change in the life circumstances of the 21 patient; m) stigmatization. Suicide attempts and any other possible adverse event related 22 to psychological treatment will also be considered. 23 11. Mortality. Psychosocial treatments may actually reduce or, by contrast, increase overall 24 mortality, in particular connected to suicidality. To test this, we will examine this outcome 25 26 in terms of a) death for any reason, b) death due to natural causes and c) due to suicide. 27 28 Search strategy 29 Electronic searches 30 31 The following sources will be searched without restrictions for language or publication period: 32 EMBASE, MEDLINE, PsycINFO, PUBMED. The search terms that will be used for PubMed are provided http://bmjopen.bmj.com/ 33 as Supplement material. We will also search the following international databases: 34 1. WHO International Clinical Trials Registry Platform (ICTRP) 35 2. BIOSIS 36 37 3. Cochrane Collaboration Controlled Trials Register 38 4. ClinicalTrials.gov. 39 40

Reference lists and other sources on September 24, 2021 by guest. Protected copyright. 41 References of all selected studies will be inspected for other published reports and citations of 42 43 unpublished studies. We will also inspect previous reviews conducted on psychological treatments 44 for schizophrenia to check if some studies meet our inclusion criteria as well. In addition, we will 45 contact the first author of each included study published in the last 30 years for missing information 46 about their studies. 47

48 49 Identification and selection of studies 50 Studies identified through electronic and manual searches will be listed with citation, titles and 51 abstracts, in Citavi; duplicates will be excluded. The eligibility for inclusion process will be conducted 52 53 in two separate stages: 54 1. Two authors will independently inspect title and abstracts identified in the literature searches, 55 and exclude those not pertinent. Disagreement will be resolved by discussion, and where doubt 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 still remains, we will acquire the full article for further inspection and the article will proceed to 4 the next stage; 5 2. Once the full articles are obtained, two reviewers will independently assess them for eligibility. 6 Disagreements will be resolved by discussion, and, if needed, a third senior author will be 7 8 involved. When required, further information will be obtained from study authors. 9 10 Data extraction 11 Two authors will independently extract data from all selected trials. When disagreement arises we 12 13 will resolve it by discussion and, if needed, involving a third senior author. Where this is not sufficient 14 we will contact the study authors. 15 The following data will be collected from each included study: 16 • Study citation,For year(s) ofpeer study, registration review number to trialsonly registries, year of publication, 17 18 location, setting, number of centers, sample size, diagnostic criteria, funding/sponsor (industry or 19 academic); 20 • Methodology, including study design (type of RCT), number of arms, risk of bias (see below); 21 • Characteristics of study participants, including gender, age, details on diagnosis, number 22 randomized to each arm, sociodemographic characteristics, whether psychological treatments 23 24 naïve at baseline, or with previous experience with the experimental intervention); 25 • Characteristics of intervention, including number and frequency of sessions, therapy setting, 26 expertise of therapist, researcher allegiance at study arm level; 27 • Outcome measures, including information on whether an Intention to Treat (ITT) approach has 28 29 been used and how it was defined. 30 The two reviewers will independently input data into an Access database especially created for this 31 study. The software will automatically detect any inconsistencies, and they will be resolved by 32 discussion. http://bmjopen.bmj.com/ 33

34 35 Measurement of treatment effect 36 Relative treatment effects 37 • Continuous outcomes: For continuous outcomes we will use the standardized mean difference 38 (SMD), because we expect that the studies use different rating scales of overall 39 40 schizophrenia symptomatology. on September 24, 2021 by guest. Protected copyright. 41 • Dichotomous outcomes: The effect size for dichotomous outcomes will be the risk ratio (RR) and 42 its 95% confidence intervals (CIs). 43 44 45 Relative treatment ranking 46 We will estimate the probability for each intervention to be ranked at each possible place, given the 47 relative effect sizes as estimated in NMA. As described in Salanti et al (43) we will obtain a hierarchy 48 of the competing interventions using the Surface Under the Cumulative Ranking curve (SUCRA) and 49 mean ranks. SUCRA values will be expressed as percentage, showing the relative probability of an 50 51 intervention to be among the best options. 52 53 Dealing with missing outcome data and missing statistics 54 For continuous outcomes we will extract data for all randomized patients if possible, and we will give 55 56 preference to data based on mixed-effect models of repeated measurements of multiple imputations 57 over last-observation-carried-forward data. 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 We will use published standard deviations (SDs), where available. When standard errors instead of 4 SDs are presented, the former will be converted to SDs (44). If both are missing we will estimate SDs 5 from p-values or confidence intervals, as described in Section 7.7.3 of the Cochrane Handbook for 6 Systematic Reviews (45). If none of these options is viable we will contact the original authors. When 7 8 no information can be obtained we will derive SDs from those of the other studies using a validated 9 imputation technique (44). 10 For dichotomous outcomes, everyone allocated to the intervention will be counted whether they 11 completed the follow up or not. If the authors applied such a strategy, we will use their results. If the 12 original authors presented only the results of the per-protocol or completer population, we will 13 14 assume that those participants lost to follow-up would not have changed in a given outcome. In 15 terms of efficacy this means that they would be conservatively considered to have not responded to 16 treatment or control.For In terms peer of tolerability review it would mean only that participants would not have 17 developed a side-effect. 18 19 20 Risk of bias assessment 21 Risk of bias will be assessed for each included study using the Cochrane Collaboration ‘risk of bias’ 22 23 tool (45, 46). The following domains will be considered: 24 1. Sequence generation: was the allocation sequence adequately generated? 25 2. Allocation concealment: was allocation adequately concealed? 26 3. Blinding of participants: was knowledge of the allocated treatment adequately prevented during 27 28 the study? Given the peculiarity of the included studies, in which the therapist cannot be blind, we 29 will consider under this item only if a way was found to keep patients unaware of the treatment they 30 were receiving (even if we expect this will not be likely); 31 4. Blinding of outcome assessors: were outcome evaluated by blind raters? Were adequate measures 32 taken to prevent them to discover treatment allocation during the study? http://bmjopen.bmj.com/ 33 34 5. Incomplete outcome data: were incomplete outcome data adequately addressed? 35 6. Selective reporting: are reports of the study free from suggestion of selective outcome reporting? 36 7. Researcher’s allegiance: do the researchers involved have a vested interest for the psychological 37 treatment under investigation? We will additionally consider this point as possible source of bias, 38 since it has been claimed to be relevant in trials on psychological interventions (47, 48, 49). 39 40 A description of what was reported about the same domains in each study will be provided, and a on September 24, 2021 by guest. Protected copyright. 41 judgement on the risk of bias will be made for each one of them, based on the following three 42 categories: ‘high risk of bias’, ‘low risk of bias’ and ‘unclear risk of bias’ where information are not 43 sufficient to make a judgement. Two independent review authors will assess the risk of bias in 44 selected studies. Any disagreement will be resolved through discussion. Where necessary, the 45 46 authors of the studies will be contacted for further information. Studies will be classified as having 47 low risk of bias if none of the domains above was rated as high risk of bias and three or less were 48 rated as unclear risk; moderate if one was rated as high risk of bias or none was rated as high risk of 49 bias but four or more were rated as unclear risk, and all other cases will be assumed to pertain to 50 high risk of bias (50). We will not include studies in the data analyses whose sequence generation 51 52 was at high risk of bias (e.g. randomization by the date of birth or day of the week). Effects of high 53 risk of bias in the other domains will be analyzed by sensitivity analyses. 54 55 56 Data analysis 57 Characteristics of the included studies 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 We will produce descriptive statistics and study population characteristics across all eligible trials, 4 describing the types of comparisons and other clinical or methodological variables, such as age, 5 duration of illness, co-medication, country, duration of study and number of session. 6 7 8 Two-step procedure 9 In a first step we will perform series of conventional pair-wise meta-analyses by combining studies 10 that compared the same interventions, including the comparison between active treatments and the 11 different control arms. In subgroups with very few RCTs available or if the requirements of network 12 meta-analysis are not met it can be that network meta-analysis will not be appropriate and in this 13 14 case, conventional pairwise meta-analysis will be the most straightforward approach. As 15 heterogeneity is likely, a random effects model will be used. In a second step we will then perform a 16 network meta-analysisFor within peer a frequentist framework.review only 17 18 19 Assessment of heterogeneity 20 The heterogeneity (variability in relative treatment effects within the same treatment comparison) 21 will be measured with the tau-squared (the variance of the random effects distribution). The 22 23 heterogeneity variance will be assumed common across the various treatment comparisons (grouped 24 by comparison type) and the empirical distributions will be used to characterize the amount of 25 heterogeneity as low, moderate or high using the first and third quantiles (51–53). Potential reasons 26 for heterogeneity will be explored by subgroup analysis (see below). 27

28 29 Assessment of the transitivity assumption 30 Joint analysis of treatments can be misleading if the network is substantially intransitive. We assume 31 that patients who fulfill the inclusion criteria outlined in criteria for considering studies for this 32 review section are equally likely to be randomised to any of the interventions that we plan to http://bmjopen.bmj.com/ 33 34 compare. We will need to investigate the distribution of clinical and methodological variables that 35 can act as effect modifiers across treatment comparisons (54). We have maximized the chances of 36 transitivity in our network with regard to clinical variables by limiting our samples to participants 37 with schizophrenia and excluding specific subgroups like first episode patients or patients with 38 prevalent negative symptoms. Other clinical or methodological variables that may influence the 39 40 efficacy of psychological interventions include administration mode and frequency of the treatment on September 24, 2021 by guest. Protected copyright. 41 (like number of sessions and experience of the therapist), baseline severity (see below, “Investigation 42 of heterogeneity and inconsistency”), and blinding, which will also be assessed in sensitivity analyses. 43 We will investigate if these variables are similarly distributed across studies grouped by comparison. 44 The comparability of studies comparing the intervention with treatment as usual or waiting list 45 46 conditions with those that provide head-to-head evidence will be examined carefully. 47 48 Network meta-analysis 49 50 Network meta-analysis combines direct and indirect evidence for all relative treatment effects and 51 can therefore provide estimates with maximum power and increased precision (55). If the collected 52 studies appear to be sufficiently similar with respect to the distribution of effect modifiers (refer 53 Assessment of transitivity assumption section), we will conduct a random-effects NMA to synthesize 54 all evidence for each outcome, and obtain a comprehensive ranking of all treatments. We will 55 56 assume a single heterogeneity parameter for each network. We will present the summary SMDs or 57 RRs for all pairwise comparisons in a league table. We will also estimate the prediction intervals to 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 assess how much the common heterogeneity affects the relative effect with respect to the extra 4 uncertainty anticipated in a future study. To rank the various treatments for each outcome, we will 5 use the surface under the cumulative ranking curve (SUCRA) and the mean ranks. 6 7 8 Assessment of inconsistency 9 The strategical and conceptual evaluation of transitivity will be supplemented with a statistical 10 evaluation of consistency, the agreement between direct and indirect evidence. We will employ local 11 12 as well as global methods to evaluate consistency (56). Local methods detect ‘hot spots’ of 13 inconsistency, evidence loops that are inconsistent or comparisons for which direct and indirect 14 evidence disagree. We will employ a method that separates direct evidence from indirect evidence 15 provided by the entire network and then evaluate the agreement of these two pieces of evidence 16 (57). We will also Forevaluate consistencypeer in thereview entire network by onlycalculating the design-by-treatment 17 18 interaction test and I-squared for network heterogeneity, inconsistency, and for both (58). Tests for 19 inconsistency are known to have low power, and empirical evidence has suggested that 10% of 20 evidence loops published in the medical literature are expected to be inconsistent (60). Therefore, 21 interpretation of the statistical inference about inconsistency will be carried out with caution and 22 possible sources of inconsistency will be explored even in the absence of evidence for inconsistency. 23 24 25 Investigation of heterogeneity and inconsistency 26 We expect small amounts of heterogeneity and inconsistency to be present given the variety of study 27 settings we plan to include. The following potential effect modifiers of the primary outcome will be 28 29 explored by subgroup analyses: 30 a) Number of sessions 31 b) Study duration 32 c) Setting: individual vs group http://bmjopen.bmj.com/ 33 d) Expertise of the therapist 34 35 e) Baseline severity (PANSS or BPRS score at baseline) 36 f) High versus low and middle income countries. 37 38 Sensitivity analyses 39 40 We will explore the following sensitivity analyses by excluding: on September 24, 2021 by guest. Protected copyright. 41 a) studies in which the outcome assessor was not blind (open studies) 42 b) studies that presented only completer analyses 43 c) studies characterized as pertaining to high risk of bias 44 d) studies with high risk of bias in researchers’ allegiance 45 46 e) studies focused on treatment resistant patients (study defined) 47 f) studies with a non-active comparison group. 48 49 Publication bias 50 51 We will first examine funnel plots of pairwise MAs if there are 10 or more studies included. We will 52 also explore the association between study size and effect size with a comparison-adjusted funnel 53 plot that has been adapted to network meta-analysis (59). 54 55 56 Evaluating the quality of the evidence 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 The quality of evidence contributing to each network estimate will be evaluated using an adaptation 4 of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework 5 specifically developed for network meta-analysis (56). We will characterize the credibility of a body 6 of evidence based on the study limitations, imprecision, heterogeneity/inconsistency, indirectness, 7 8 and publication bias. 9 10 Statistical software 11 The analysis and presentation of results will be performed using the Stata packages network and 12 network_graphs, the R package netmeta. 13 14 15 ETHICS AND DISSEMINATION 16 For peer review only 17 This review does not require ethical approval. Findings will be published in peer reviewed scientific 18 journals, granting open access, and the database will be made publicly available (in agreement with 19 European Research Council Guidelines on Implementation of Open Access to Scientific Publications 20 and Research Data (http://ec.europa.eu/research/participants/data/ref/h2020/other/hi/oa- 21 pilot/h2020-hi-erc-oa-guide_en.pdf). 22 23 24 25 REFERENCES 26 1. Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M et al. Years lived with disability (YLDs) for 1160 27 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 28 2010. Lancet 2012; 380(9859):2163–96. 29 30 2. Leucht S, Cipriani A, Spineli L, Mavridis D, Orey D, Richter F et al. Comparative efficacy and tolerability of 15 31 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet 2013; 382(9896):951–62. 32 http://bmjopen.bmj.com/ 33 3. National Collaborating Centre for Mental Health. Core interventions in the treatment and management of 34 schizophrenia in adults in primary and secondary care (Clinical Guideline CG82). London; 2009. 35 4. Buchanan RW, Kreyenbuhl J, Kelly DL, Noel JM, Boggs DL, Fischer BA et al. The 2009 schizophrenia PORT 36 psychopharmacological treatment recommendations and summary statements. Schizophr Bull 2010; 36(1):71– 37 38 93. 39 5. Pinquart M, Oslejsek B, Teubert D. Efficacy of systemic therapy on adults with mental disorders: A meta- 40

analysis. Psychother Res 2016; 26(2):241–57. on September 24, 2021 by guest. Protected copyright. 41 42 6. Ruddy R, Milnes D. Art therapy for schizophrenia or schizophrenia-like illnesses. Cochrane Database Syst Rev 43 2005; (4):CD003728. 44 7. Jones C, Hacker D, Cormac I, Meaden A, Irving CB. Cognitive versus other psychosocial 45 46 treatments for schizophrenia. Cochrane Database Syst Rev 2012; (4):CD008712. 47 8. Lutgens D, Gariepy G, Malla A. Psychological and psychosocial interventions for negative symptoms in 48 psychosis: Systematic review and meta-analysis. Br J Psychiatry 2017; 210(5):324–32. 49 50 9. Zimmermann G, Favrod J, Trieu VH, Pomini V. The effect of cognitive behavioral treatment on the positive 51 symptoms of schizophrenia spectrum disorders: A meta-analysis. Schizophr Res 2005; 77(1):1–9. 52 10. Turner DT, van der Gaag M, Karyotaki E, Cuijpers P. Psychological interventions for psychosis: a meta- 53 54 analysis of comparative outcome studies. Am J Psychiatry 2014; 171(5):523–38. 55 11. Naeem F, Asmer MS, Khoury B, Kingdon D, Farooq S. Acceptance and Commitment Therapy for 56 schizophrenia and related disorders (Protocol). Cochrane Database of Systematic Reviews 2015; (6). 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 12. Gray R, Bressington D, Ivanecka A, Hardy S, Jones M, Schulz M et al. Is adherence therapy an effective 4 adjunct treatment for patients with schizophrenia spectrum disorders? A systematic review and meta-analysis. 5 BMC Psychiatry 2016; 16:90. 6 13. Marshall M, Lockwood A. Assertive community treatment for people with severe mental disorders. 7 Cochrane Database Syst Rev 2000; (2):CD001089. 8 9 14. Dieterich M, Irving CB, Bergman H, Khokhar MA, Park B, Marshall M. Intensive case management for severe 10 mental illness. Cochrane Database Syst Rev 2017; 1:CD007906. 11 12 15. Jauhar S, McKenna PJ, Radua J, Fung E, Salvador R, Laws KR. Cognitive-behavioural therapy for the 13 symptoms of schizophrenia: systematic review and meta-analysis with examination of potential bias. Br J 14 Psychiatry 2014; 204(1):20–9. 15 16. van der Gaag M, Valmaggia LR, Smit F. The effects of individually tailored formulation-based cognitive 16 For peer review only 17 behavioural therapy in auditory hallucinations and delusions: A meta-analysis. Schizophr Res 2014; 156(1):30– 18 7. 19 17. Hazell CM, Hayward M, Cavanagh K, Strauss C. A systematic review and meta-analysis of low intensity CBT 20 for psychosis. Clin Psychol Rev 2016; 45:183–92. 21 22 18. Kennedy L, Xyrichis A. Cognitive Behavioral Therapy Compared with Non-specialized Therapy for Alleviating 23 the Effect of Auditory Hallucinations in People with Reoccurring Schizophrenia: A Systematic Review and Meta- 24 analysis. Community Ment Health J 2017; 53(2):127–33. 25 26 19. Cella M, Preti A, Edwards C, Dow T, Wykes T. Cognitive remediation for negative symptoms of 27 schizophrenia: A network meta-analysis. Clin Psychol Rev 2017; 52:43–51. 28 20. Ren J, Xia J. Dance therapy for schizophrenia. Cochrane Database Syst Rev 2013; (10):CD006868. 29 30 21. Pitschel-Walz G, Leucht S, Bauml J, Kissling W, Engel RR. The effect of family interventions on relapse and 31 rehospitalization in schizophrenia--a meta-analysis. Schizophr Bull 2001; 27(1):73–92. 32 http://bmjopen.bmj.com/ 33 22. Pharoah F, Mari J, Rathbone J, Wong W. Family intervention for schizophrenia. Cochrane Database Syst Rev 34 2010; (12):CD000088. 35 23. Orfanos S, Banks C, Priebe S. Are Group Psychotherapeutic Treatments Effective for Patients with 36 Schizophrenia? A Systematic Review and Meta-Analysis. Psychother Psychosom 2015; 84(4):241–9. 37 38 24. Roder V, Mueller DR, Mueser KT, Brenner HD. Integrated psychological therapy (IPT) for schizophrenia: Is it 39 effective? Schizophr Bull 2006; 32 Suppl 1:S81-93. 40 on September 24, 2021 by guest. Protected copyright. 41 25. Eichner C, Berna F. Acceptance and Efficacy of Metacognitive Training (MCT) on Positive Symptoms and 42 Delusions in Patients With Schizophrenia: A Meta-analysis Taking Into Account Important Moderators. 43 Schizophr Bull 2016; 42(4):952–62. 44 26. Aust J, Bradshaw T. Mindfulness interventions for psychosis: A systematic review of the literature. J 45 Psychiatr Ment Health Nurs 2017; 24(1):69–83. 46 47 27. Geretsegger M, Mössler KA, Bieleninik Ł, Chen X-J, Heldal TO, Gold C. Music therapy for people with 48 schizophrenia and schizophrenia-like disorders. Cochrane Database Syst Rev 2017; 5:CD004025. 49 50 28. Malmberg L, Fenton M. Individual psychodynamic psychotherapy and for schizophrenia and 51 severe mental illness. Cochrane Database Syst Rev 2001; (3):CD001360. 52 29. Pekkala E, Merinder L. Psychoeducation for schizophrenia. Cochrane Database Syst Rev 2000; 53 (4):CD002831. 54 55 30. Almerie MQ, Okba Al Marhi M, Jawoosh M, Alsabbagh M, Matar HE, Maayan N et al. Social skills 56 programmes for schizophrenia. Cochrane Database Syst Rev 2015; (6):CD009006. 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 31. Buckley LA, Maayan N, Soares-Weiser K, Adams CE. Supportive therapy for schizophrenia. Cochrane 4 Database Syst Rev 2015; (4):CD004716. 5 32. Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C et al. The PRISMA extension 6 statement for reporting of systematic reviews incorporating network meta-analyses of health care 7 interventions: Checklist and explanations. Ann Intern Med 2015; 162(11):777–84. 8 9 33. Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M et al. Preferred reporting items for 10 systematic review and meta-analysis protocols (PRISMA-P) 2015: Elaboration and explanation. BMJ 2015; 11 349:g7647. 12 13 34. Egger M, Zellweger-Zahner T, Schneider M, Junker C, Lengeler C, Antes G. Language bias in randomised 14 controlled trials published in English and German. Lancet 1997; 350(9074):326–9. 15 35. Wu T, Li Y, Liu G, Bian Z, Li J, Zhang J, Xie L, Ni J, editor. Investigation of authenticity of 'claimed' randomized 16 For peer review only 17 controlled trials (RCTs) and quality assessment of RCT reports published in China; 2006. 18 36. Woodhead M. 80% of China's clinical trial data are fraudulent, investigation finds. BMJ 2016; 355:i5396. 19 20 37. Carpenter WT, JR, Buchanan RW. Schizophrenia. N Engl J Med 1994; 330(10):681–90. 21 38. Adams,C.E., Coutinho,E., Davis,J.M., Duggan,L., Essali,A., Fenton,M., Li,C., Jayaram,M., Leucht,S., Tharyan,P., 22 Välimäki,M. Cochrane Schizophrenia Group. The Cochrane Library. Chichester, UK: John Wiley & Sons Ltd; 23 2011. 24 25 39. Furukawa TA, Noma H, Caldwell DM, Honyashiki M, Shinohara K, Imai H et al. Waiting list may be a nocebo 26 condition in psychotherapy trials: a contribution from network meta-analysis. Acta Psychiatr Scand 2014; 27 130(3):181–92. 28 29 40. Marshall M, Lockwood A, Bradley C, Adams C, Joy C, Fenton M. Unpublished rating scales: a major source of 30 bias in randomised controlled trials of treatments for schizophrenia. Br J Psychiatry 2000; 176:249–52. 31 41. Vaughan B, Goldstein MH, Alikakos M, Cohen LJ, Serby MJ. Frequency of reporting of adverse events in 32 http://bmjopen.bmj.com/ 33 randomized controlled trials of psychotherapy vs. psychopharmacotherapy. Compr Psychiatry 2014; 55(4):849– 34 55. 35 42. Linden M, Schermuly-Haupt M-L. Definition, assessment and rate of psychotherapy side effects. World 36 Psychiatry 2014; 13(3):306–9. 37 38 43. Salanti G, Ades AE, Ioannidis JPA. Graphical methods and numerical summaries for presenting results from 39 multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol 2011; 64(2):163–71. 40 on September 24, 2021 by guest. Protected copyright. 41 44. Furukawa TA, Barbui C, Cipriani A, Brambilla P, Watanabe N. Imputing missing standard deviations in meta- 42 analyses can provide accurate results. J Clin Epidemiol 2006; 59(1):7–10. 43 45. Higgins JPT, editor. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 [updated 44 March 2011]; 2011. 45 46 46. Higgins JPT, Altman DG, Sterne JAC (editors), editor. Assessing risk of bias in included studies. In: Higgins 47 JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic Reviews of 48 Interventions version 5.2.0 (updated June 2017); Cochrane, 2017. Available from: URL: 49 http://training.cochrane.org/handbook. 50 51 47. Dragioti E, Dimoliatis I, Evangelou E. Disclosure of researcher allegiance in meta-analyses and randomised 52 controlled trials of psychotherapy: a systematic appraisal. BMJ Open 2015; 5(6):e007206. 53 48. Lieb K, Osten-Sacken J von der, Stoffers-Winterling J, Reiss N, Barth J. Conflicts of interest and spin in 54 55 reviews of psychological therapies: a systematic review. BMJ Open 2016; 6(4):e010606. 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 49. Munder T, Brutsch O, Leonhart R, Gerger H, Barth J. Researcher allegiance in psychotherapy outcome 4 research: an overview of reviews. Clin Psychol Rev 2013; 33(4):501–11. 5 50. Furukawa TA, Salanti G, Atkinson LZ, Leucht S, Ruhe HG, Turner EH et al. Comparative efficacy and 6 acceptability of first-generation and second-generation antidepressants in the acute treatment of major 7 depression: Protocol for a network meta-analysis. BMJ Open 2016; 6(7):e010919. 8 9 51. Rhodes KM, Turner RM, White IR, Jackson D, Spiegelhalter DJ, Higgins JPT. Implementing informative priors 10 for heterogeneity in meta-analysis using meta-regression and pseudo data. Stat Med 2016; 35(29):5495–511. 11 12 52. Rhodes KM, Turner RM, Higgins JPT. Empirical evidence about inconsistency among studies in a pair-wise 13 meta-analysis. Res Synth Methods 2016; 7(4):346–70. 14 53. Turner RM, Davey J, Clarke MJ, Thompson SG, Higgins JP. Predicting the extent of heterogeneity in meta- 15 analysis, using empirical data from the Cochrane Database of Systematic Reviews. Int J Epidemiol 2012; 16 For peer review only 17 41(3):818–27. 18 54. Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: Many 19 names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods 20 2012; 3(2):80–97. 21 22 55. Salanti G, Higgins JPT, Ades AE, Ioannidis JPA. Evaluation of networks of randomized trials. Stat Methods 23 Med Res 2008; 17(3):279–301. 24 25 56. Salanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JPT. Evaluating the quality of evidence from a 26 network meta-analysis. PLoS One 2014; 9(7):e99682. 27 57. Dias S, Welton NJ, Caldwell DM, Ades AE. Checking consistency in mixed treatment comparison meta- 28 analysis. Stat Med 2010; 29(7-8):932–44. 29 30 58. Higgins JPT, Jackson D, Barrett JK, Lu G, Ades AE, White IR. Consistency and inconsistency in network meta- 31 analysis: Concepts and models for multi-arm studies. Res Synth Methods 2012; 3(2):98–110. 32 http://bmjopen.bmj.com/ 33 59. Chaimani A, Salanti G. Using network meta-analysis to evaluate the existence of small-study effects in a 34 network of interventions. Res Synth Methods 2012; 3(2):161–76. 35 60. Veroniki AA, Vasiliadis HS, Higgins JPT, Salanti G. Evaluation of inconsistency in networks of interventions. 36 Int J Epidemiol 2013; 42(1):332–45. 37 38 39 Authors’ contributions: IB and SL designed this study, drafted and critically revised the 40 on September 24, 2021 by guest. Protected copyright. 41 protocol. IB will screen search results for inclusion, conduct data extraction and data analysis, and 42 draft the final manuscript. SL will assist with data extraction and analysis, and revise the final 43 manuscript. CR and SW will screen search results for inclusion and conduct data extraction. GS 44 provided substantial methodological advice in planning the study, and will assist with data analysis. 45 46 CB and TF contributed with clinical and methodological input in planning the study. All authors 47 contributed to and have approved the final manuscript. 48 49 Collaborators: Samantha Roberts helped us to conduct the literature searches. Maximilian Huhn, 50 51 Johannes Schneider-Thoma, Marc Krause and Costanza Carmi provided help and suggestions. 52 53 Funding: This project has received funding from the European Union’s Horizon 2020 Research and 54 55 Innovation Programme under the Marie Skłodowska-Curie grant agreement N° 701717. This work 56 was also supported by the German Research Foundation (DFG) and the Technical University of 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Munich within the Open Access Publishing Funding Programme. The funder had no role in developing 4 the protocol. 5 6 7 Competing interests: SL in the past 3 years has received honoraria for consulting from Roche, 8 TEVA, Otsuka, Lundbeck, and LB Pharma; for lectures from Otsuka, Lundbeck, Janssen, ICON, Lilly, 9 Sanofi Aventis, AOP Orphan, Roche, and Servier; and for a publication from Roche. TAF has received 10 lecture fees from Eli Lilly, Janssen, Meiji, Mitsubishi-Tanabe, MSD and Pfizer and consultancy fees 11 from Takeda Science Foundation. He has received royalties from Igaku-Shoin and Nihon Bunka 12 13 Kagaku-sha publishers. He has received research support from Mochida and Mitsubishi-Tanabe. 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 40 on September 24, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Search terms for Pubmed 4 5 ("Schizophrenia"[Mesh] OR "Paranoid Disorders"[Mesh] OR schizo*[Title/Abstract] OR 6 7 psychotic*[Title/Abstract] OR psychosis[Title/Abstract] OR psychoses[Title/Abstract]) AND 8 ("Psychotherapy"[Mesh] or "Behavior Therapy"[Mesh] or ""[Mesh] or 9 "Complementary Therapies"[Mesh] or "Psychoanalysis"[Mesh] or "Counseling"[Mesh] or 10 "Hypnosis"[Mesh] or "Association"[Mesh] or "Association Learning"[Mesh] OR 11 abreaction[Title/Abstract] OR "acceptance[Title/Abstract] AND commitment therapy"[Title/Abstract] 12 13 OR "acting out"[Title/Abstract] OR adlerian[Title/Abstract] OR "analytical 14 psychotherapy"[Title/Abstract] OR "analytical "[Title/Abstract] OR "anger 15 control"[Title/Abstract] OR "anger management"[Title/Abstract] OR "animal therapy"[Title/Abstract] 16 OR "animal For therapies"[Title/Abstract] peer review OR "art therapy"[Title/Abstract] only OR "art 17 18 therapies"[Title/Abstract] OR "assertive training"[Title/Abstract] OR "assertiveness 19 training"[Title/Abstract] OR "attention training technique"[Title/Abstract] OR "autogenic 20 training"[Title/Abstract] OR autosuggestion[Title/Abstract] OR ""[Title/Abstract] OR 21 "aversion therapies"[Title/Abstract] OR "balint group"[Title/Abstract] OR befriending[Title/Abstract] 22 OR "behavior contracting"[Title/Abstract] OR "behavior modification"[Title/Abstract] OR "behavior 23 24 regulation"[Title/Abstract] OR "behavior therapy"[Title/Abstract] OR "behavior 25 therapies"[Title/Abstract] OR "behaviour contracting"[Title/Abstract] OR "behaviour 26 modification"[Title/Abstract] OR "behaviour regulation"[Title/Abstract] OR "behaviour 27 therapy"[Title/Abstract] OR "behaviour therapies"[Title/Abstract] OR bibliotherapy[Title/Abstract] 28 OR bibliotherapies[Title/Abstract] OR [Title/Abstract] OR "body 29 30 psychotherapy"[Title/Abstract] OR "body psychotherapies"[Title/Abstract] OR "brief 31 psychotherapy"[Title/Abstract] OR "brief psychotherapies"[Title/Abstract] OR "caregiver 32 support"[Title/Abstract] OR cbt[Title/Abstract] OR "client centre"[Title/Abstract] OR "client http://bmjopen.bmj.com/ 33 center"[Title/Abstract] OR "cognitive behavior"[Title/Abstract] OR "cognitive 34 behaviorial"[Title/Abstract] OR "cognitive intervention"[Title/Abstract] OR "cognitive 35 36 interventions"[Title/Abstract] OR "cognitive rehabilitation"[Title/Abstract] OR "cognitive 37 remediation"[Title/Abstract] OR "cognitive technique"[Title/Abstract] OR "cognitive 38 techniques"[Title/Abstract] OR "cognitive therapy"[Title/Abstract] OR "cognitive 39 therapies"[Title/Abstract] OR "cognitive treatment"[Title/Abstract] OR "cognitive 40 treatments"[Title/Abstract] OR "color therapy"[Title/Abstract] OR "color therapies"[Title/Abstract] on September 24, 2021 by guest. Protected copyright. 41 42 OR "colour therapy"[Title/Abstract] OR "colour therapies"[Title/Abstract] OR "compassionate mind 43 training"[Title/Abstract] OR "conjoint therapy"[Title/Abstract] OR "conjoint therapies"[Title/Abstract] 44 OR ""[Title/Abstract] OR "conversational therapy"[Title/Abstract] OR 45 "conversational therapies"[Title/Abstract] OR "conversion therapy"[Title/Abstract] OR "conversion 46 47 therapies"[Title/Abstract] OR "coping skills"[Title/Abstract] OR counseling[Title/Abstract] OR 48 counselling[Title/Abstract] OR countertransference[Title/Abstract] OR "couples 49 therapy"[Title/Abstract] OR "couples therapies"[Title/Abstract] OR "covert 50 sensitization"[Title/Abstract] OR "covert sensitisation"[Title/Abstract] OR "crisis 51 intervention"[Title/Abstract] OR "dance therapy"[Title/Abstract] OR "dance 52 53 therapies"[Title/Abstract] OR dialectic[Title/Abstract] OR dialectical[Title/Abstract] OR "dream 54 analysis"[Title/Abstract] OR eclectic[Title/Abstract] OR "emotion focused"[Title/Abstract] OR 55 "emotionally focused"[Title/Abstract] OR "emotional freedom technique"[Title/Abstract] OR 56 "encounter group therapy"[Title/Abstract] OR "encounter group therapies"[Title/Abstract] OR 57 ""[Title/Abstract] OR "existential therapies"[Title/Abstract] OR "experiential 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 psychotherapy"[Title/Abstract] OR "experiential psychotherapies"[Title/Abstract] OR "exposure 4 therapy"[Title/Abstract] OR "exposure therapies"[Title/Abstract] OR "expressive 5 psychotherapy"[Title/Abstract] OR "expressive psychotherapies"[Title/Abstract] OR "eye movement 6 desensitization"[Title/Abstract] OR "eye movement desensitisation"[Title/Abstract] OR "family 7 8 intervention"[Title/Abstract] OR "family interventions"[Title/Abstract] OR "family 9 therapy"[Title/Abstract] OR "family therapies"[Title/Abstract] OR ""[Title/Abstract] 10 OR "feminist therapies"[Title/Abstract] OR "free association"[Title/Abstract] OR 11 freudian[Title/Abstract] OR "geriatric psychotherapy"[Title/Abstract] OR "geriatric 12 psychotherapies"[Title/Abstract] OR ""[Title/Abstract] OR "gestalt 13 14 therapies"[Title/Abstract] OR griefwork[Title/Abstract] OR "group intervention"[Title/Abstract] OR 15 "group interventions"[Title/Abstract] OR ""[Title/Abstract] OR "group 16 psychotherapies"[Title/Abstract]For peer OR review "group therapy"[Title/Abstract] only OR "group 17 therapies"[Title/Abstract] OR "guided imagery"[Title/Abstract] OR "holistic 18 psychotherapy"[Title/Abstract] OR "holistic psychotherapies"[Title/Abstract] OR "humanistic 19 20 psychotherapy"[Title/Abstract] OR "humanistic psychotherapies"[Title/Abstract] OR 21 hypnosis[Title/Abstract] OR [Title/Abstract] OR hypnotherapies[Title/Abstract] OR 22 hypnotizability[Title/Abstract] OR hypnotisability[Title/Abstract] OR imagery[Title/Abstract] OR 23 "implosive therapy"[Title/Abstract] OR "implosive therapies"[Title/Abstract] OR "individual 24 psychotherapy"[Title/Abstract] OR "individual psychotherapies"[Title/Abstract] OR "insight 25 26 therapy"[Title/Abstract] OR "insight therapies"[Title/Abstract] OR "integrated psychological 27 therapy"[Title/Abstract] OR "integrative psychotherapy"[Title/Abstract] OR "integrative 28 psychotherapies"[Title/Abstract] OR "integrative therapy"[Title/Abstract] OR "integrative 29 therapies"[Title/Abstract] OR interpersonal[Title/Abstract] OR jungian[Title/Abstract] OR 30 kleinian[Title/Abstract] OR [Title/Abstract] OR logotherapies[Title/Abstract] OR 31 32 "marathon group therapy"[Title/Abstract] OR "marathon group therapies"[Title/Abstract] OR http://bmjopen.bmj.com/ 33 "marital therapy"[Title/Abstract] OR "marital therapies"[Title/Abstract] OR meditation[Title/Abstract] 34 OR "mental healing"[Title/Abstract] OR ""[Title/Abstract] OR "metacognitive 35 therapies"[Title/Abstract] OR "metacognitive training"[Title/Abstract] OR "milieu 36 37 therapy"[Title/Abstract] OR "milieu therapies"[Title/Abstract] OR mindfulness[Title/Abstract] OR 38 "morita therapy"[Title/Abstract] OR "morita therapies"[Title/Abstract] OR multimodal[Title/Abstract] 39 OR "music therapy"[Title/Abstract] OR "music therapies"[Title/Abstract] OR "narrative 40

therapy"[Title/Abstract] OR "narrative therapies"[Title/Abstract] OR "nondirective on September 24, 2021 by guest. Protected copyright. 41 therapy"[Title/Abstract] OR "nondirective therapies"[Title/Abstract] OR "object 42 43 relations"[Title/Abstract] OR "person centred therapy"[Title/Abstract] OR "person centred 44 therapies"[Title/Abstract] OR "person centered therapy"[Title/Abstract] OR "person centered 45 therapies"[Title/Abstract] OR "personal construct therapy"[Title/Abstract] OR "personal construct 46 therapies"[Title/Abstract] OR "persuasion therapy"[Title/Abstract] OR "persuasion 47 therapies"[Title/Abstract] OR "pet therapy"[Title/Abstract] OR "pet therapies"[Title/Abstract] OR 48 49 ""[Title/Abstract] OR "play therapies"[Title/Abstract] OR "primal therapy"[Title/Abstract] 50 OR "primal therapies"[Title/Abstract] OR "problem solving"[Title/Abstract] OR 51 psychoanalyse[Title/Abstract] OR psychoanalysed[Title/Abstract] OR psychoanalysis[Title/Abstract] 52 OR psychoanalytic[Title/Abstract] OR [Title/Abstract] OR psychodynamic[Title/Abstract] 53 OR psychoeducate[Title/Abstract] OR psychoeducation[Title/Abstract] OR 54 55 psychoeducating[Title/Abstract] OR psychologic[Title/Abstract] OR psychological[Title/Abstract] OR 56 psychologically[Title/Abstract] OR "psychological therapy"[Title/Abstract] OR "psychological 57 therapies"[Title/Abstract] OR "psychosocial treatment"[Title/Abstract] OR "psychosocial 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 treatments"[Title/Abstract] OR psychotherapy[Title/Abstract] OR psychotherapies[Title/Abstract] OR 4 "psychotherapeutic counsel"[Title/Abstract] OR "psychotherapeutic counseling"[Title/Abstract] OR 5 "psychotherapeutic counselling"[Title/Abstract] OR "psychotherapeutic processes"[Title/Abstract] OR 6 "psychotherapeutic training"[Title/Abstract] OR "psychotherapeutic treatment"[Title/Abstract] OR 7 8 "psychotherapeutic treatments"[Title/Abstract] OR "rational emotive"[Title/Abstract] OR "reality 9 therapy"[Title/Abstract] OR "reality therapies"[Title/Abstract] OR "reciprocal 10 inhibition"[Title/Abstract] OR rehabilitation[Title/Abstract] OR rehabilitating[Title/Abstract] OR 11 "relationship therapy"[Title/Abstract] OR "relationship therapies"[Title/Abstract] OR 12 relaxation[Title/Abstract] OR "reminiscence therapy"[Title/Abstract] OR "reminiscence 13 14 therapies"[Title/Abstract] OR rogerian[Title/Abstract] OR "role play"[Title/Abstract] OR "role 15 plays"[Title/Abstract] OR "role playing"[Title/Abstract] OR "self analysis"[Title/Abstract] OR "self 16 analysing"[Title/Abstract]For OR peer "self esteem"[Title/Abstract] review OR "sensitivity only training"[Title/Abstract] OR 17 "sex therapy"[Title/Abstract] OR "sex therapies"[Title/Abstract] OR "sleep phase 18 chronotherapy"[Title/Abstract] OR "sleep phase chronotherapies"[Title/Abstract] OR "social skills 19 20 education"[Title/Abstract] OR "social skills training"[Title/Abstract] OR "socioenvironmental 21 therapy"[Title/Abstract] OR "socioenvironmental therapies"[Title/Abstract] OR 22 sociotherapy[Title/Abstract] OR sociotherapies[Title/Abstract] OR "solution focused"[Title/Abstract] 23 OR "stress management"[Title/Abstract] OR "support group"[Title/Abstract] OR "support 24 groups"[Title/Abstract] OR "supportive therapy"[Title/Abstract] OR "supportive 25 26 therapies"[Title/Abstract] OR "systematic desensitization"[Title/Abstract] OR "systematic 27 desensitisation"[Title/Abstract] OR "systemic therapy"[Title/Abstract] OR "systemic 28 therapies"[Title/Abstract] OR "therapeutic community"[Title/Abstract] OR "therapeutic 29 communities"[Title/Abstract] OR ""[Title/Abstract] OR 30 transference[Title/Abstract] OR transtheoretical[Title/Abstract] OR "validation 31 32 therapy"[Title/Abstract] OR "validation therapies"[Title/Abstract]) AND (((randomized controlled http://bmjopen.bmj.com/ 33 trial[pt]) OR (controlled clinical trial[pt]) OR (randomized[tiab]) OR (placebo[tiab]) OR (drug 34 therapy[sh]) OR (randomly[tiab]) OR (trial[tiab]) OR (groups[tiab])) NOT (animals[mh] NOT 35 humans[mh])) 36 37 38 39 40 on September 24, 2021 by guest. Protected copyright. 41 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: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

2 8 5 3 1 5 16 16 58 58 1617 1617 1617 1617 1 (title) 1 Reported on Reported pageon Supplementary Supplementary file BMJ Open Checklist item http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected copyright.

For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml registers registers or grey literatureother sources) with planned coveragedates of Provide anexplicit Provide statement the review question(s)of address will the with referenceto participants, Provide name, Provide institutional affiliation, email address authors; of all protocol provide physical address mailing Describe all intended Describe information (suchsources as electronic contact databases, with study trial authors, Specify study characteristicsSpecify the as (such PICO, design,study setting, time frame) report and characteristics (such as years considered, as (such language, years publication to status) be as used criteria for eligibilityreview for the that it could it repeatedthat be interventions, interventions, comparators, and (PICO) outcomes of corresponding authorof If the the protocolIf anrepresents amendment previously a completed of or published protocol, and as identify such changes; list otherwise, state fordocumentingplan important amendments protocol

6 6 the rationale Describe for theincontext the review of what is already known 2 2 registered, If provide namethe the registry (such as PROSPERO)of and registration number 7 7 9 9 5c rolesfunder(s), Describe of and/orsponsor(s), institution(s), if in any, developing the protocol

Role of Role of sponsor

Sponsor 5b name Provide review for the funder and/or sponsor Sources funderor 5a Indicate financialsources of support reviewor forthe other Contact 3a Update 1b the foranprotocol If update is a systematicprevious of identifyreview, as such Identification 1a report the as Identify protocol systematicaa of review Contributions 3b contributions Describe authorsof protocol and identify the the guarantor reviewof Study records:Study Search Search strategy 10 search draft Present forto usedatbe strategy leastof electronic one database, including planned limits, such Information Information sources Eligibility criteriaEligibility 8 METHODS Objectives Rationale Rationale Support: Authors: Title: ADMINISTRATIVE INFORMATION PRISMA-P (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols) 2015 checklist: recommended items to to items recommended checklist: 2015 Protocols) Meta-Analysis review for and Systematic Items (PreferredReporting PRISMA-P reviewprotocol* a systematic in address Section topic and No Item Registration INTRODUCTION Amendments 4 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 Page 21 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from Page 22 of 9 9 9 13 13 10 10 12 12 89 89 89 89 78 78 11, 12 12, 12, 13 9,

BMJ Open http://bmjopen.bmj.com/ bias(es) (suchbias(es) as publication bias across studies, selective reporting ti ti A,Petticrew M, P, Shekelle Stewart L, PRISMA-P Preferred Group. reportingitems for systematic andreview d ind conjunctionwith Explanationthe PRISMA-P and Elaboration (citewhen available) for important on September 24, 2021 by guest. Protected copyright.

For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml , τ)Kendall’s 2 Describe Describe planned extractingmethod of from (suchdata reports as piloting forms, independently,in done State the the State that process be (suchwill for used selecting studies independenttwo reviewers)as eachthrough Describe anticipatedDescribe methods assessing for risk of bias of studies, individual including whether this be will the at done level,or study outcome both; state this or informationhow will in used databe synthesis Specify any planned Specify assessment of meta within studies)within List and all define List forvariables which data (such be soughtwill asPICO items, funding sources), any pre List and all define List for outcomes whichdata sought, be including will prioritization of mainadditional and rationaleoutcomes, with duplicate), any duplicate), processes for and obtaining confirming data from investigators data and and combiningdata methods of data from studies, including planned explorationany of consistency(such as I phase of the phase review of is,(that screening, andeligibility inclusionin metaanalysis) planned data assumptionsandplanned simplifications 12 12 14 14 17 17 strength Describe how the the body evidenceof of willassessed be as(such GRADE) 13 13 11c 11c 11a 11a the mechanism(s) Describe that be will to used recordsmanage and thedata throughout review 15c 15c any Describe proposed additional analyses as (such or subgroupsensitivity analyses, metaregression) 11b 15b data are for If appropriate synthesis,quantitative describe planned summary methods handling measures, of 15d quantitative If synthesis is not appropriate, describe typethe of summary planned

collection Data Selection Data management process process It is It stronglyrecommendedchecklist this that be rea

clarification theclarification items. on to Amendments a review protocol should be tracked copyrightand dated. The for (includingPRISMA-P checklist) byheld the is Group PRISMA-P and is distributedunder a Creative Commons Attribution Licence 4.0.

* Confidence inConfidence evidencecumulative Data Data synthesis 15a criteria under Describe data which study be synthesisedwill quantitatively Risk Risk in of bias studiesindividual Metabias(es) 16 Outcomes and Outcomes meta-analysis protocols(PRISMA-P) elaboration2015: explanation. andBMJ. Jan 2015 2;349(jan02 1):g7647. Data itemsData From: Shamseer From: L, Moher D, Clarke M, Ghersi Libera D, prioritization 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 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

Psychological interventions for positive symptoms in schizophrenia: protocol for a network meta-analysis of randomized controlled trials

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-019280.R1

Article Type: Protocol

Date Submitted by the Author: 09-Nov-2017

Complete List of Authors: Bighelli, Irene; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Salanti, Georgia; University of Bern, Institute of Social and Preventive Medicine (ISPM) Reitmeir, Cornelia; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Wallis, Sofia; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Barbui, Corrado; University of Verona, WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation, Department of Neuroscience, Biomedicine and Movement Sciences, Section of Psychiatry

Furukawa, Toshi; Kyoto University, Department of Health Promotion and http://bmjopen.bmj.com/ Human Behavior, Graduate School of Medicine and School of Public Health Leucht, Stefan; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy

Primary Subject Mental health Heading:

Secondary Subject Heading: Mental health

Schizophrenia & psychotic disorders < PSYCHIATRY, Adult psychiatry < on September 24, 2021 by guest. Protected copyright. Keywords: PSYCHIATRY, Psychological interventions, Network Meta-Analysis, Psychotherapy

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Psychological interventions for positive symptoms in 4 5 schizophrenia: protocol for a network meta-analysis of 6 7 randomized controlled trials 8 9 10 11 12 Irene Bighelli1*, Georgia Salanti2, Cornelia Reitmeir1, Sofia Wallis1, Corrado Barbui3, Toshi A. 13 Furukawa4, Stefan Leucht1 14 15 16 For peer review only 17 18 19 1 Department of Psychiatry and Psychotherapy, Klinikum rechts der Isar, Technical University of 20 21 Munich, Germany 22 23 2 Institute of Social and Preventive Medicine (ISPM), University of Bern, Switzerland 24 25 3 WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation, 26 27 Department of Neuroscience, Biomedicine and Movement Sciences, Section of Psychiatry, University 28 of Verona, Italy 29 30 4 Department of Health Promotion and Human Behavior, Kyoto University Graduate School of 31 Medicine / School of Public Health, Japan 32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 *Corresponding author 40 Department of Psychiatry and Psychotherapy on September 24, 2021 by guest. Protected copyright. 41 Klinikum rechts der Isar, Technische Universität München 42 Ismaningerstr. 22, 81675 München, Germany 43 44 E-mail: [email protected] 45 46 GS: [email protected] 47 CR: [email protected] 48 SW: [email protected] 49 50 CB: [email protected] 51 TAF: [email protected] 52 SL: [email protected] 53 54

55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 ABSTRACT 4 5 Introduction 6 There is rising awareness that we need multi-disciplinary approaches integrating psychological 7 treatments for schizophrenia, but a comprehensive evidence base on their relative efficacy is lacking. 8 We will conduct a network meta-analysis (NMA), integrating direct and indirect comparisons from 9 10 randomised controlled trials (RCTs) to rank psychological treatments for schizophrenia according to 11 their efficacy, acceptability and tolerability. 12 13 Methods and analysis 14 15 We will include all RCTs comparing a psychological treatment aimed at positive symptoms of 16 schizophrenia withFor another psychologicalpeer intervention review or with a noonly treatment condition (waiting list, 17 treatment as usual). We will include studies on adult patients with schizophrenia, excluding specific 18 subpopulations (e.g. first episode patients, or patients with psychiatric comorbidities). Primary 19 20 outcome will be the change in positive symptoms on a published rating scale. Secondary outcomes 21 will be acceptability (dropout), change in overall and negative symptoms of schizophrenia, response, 22 relapse, adherence, depression, quality of life, functioning and adverse events. Published and 23 unpublished studies will be sought through database searches, trial registries and websites. Study 24 selection and data extraction will be conducted by at least two independent reviewers. We will 25 26 conduct random-effects NMA to synthesize all evidence for each outcome and obtain a 27 comprehensive ranking of all treatments. NMA will be conducted in Stata and R within a frequentist 28 framework. The risk of bias in studies will be evaluated using the Cochrane Risk of Bias tool and the 29 credibility of the evidence will be evaluated using an adaptation of the GRADE framework to NMA, 30 recommended by the Cochrane guidance. Subgroup and sensitivity analyses will be conducted to 31 32 assess the robustness of the findings. http://bmjopen.bmj.com/ 33 34 Ethics and dissemination 35 No ethical issues are foreseen. Results from this study will be published in peer-reviewed journals 36 37 and presented at relevant conferences. 38 39 PROSPERO registration number: CRD42017067795 40 on September 24, 2021 by guest. Protected copyright. 41 42 43 STRENGTHS AND LIMITATIONS OF THIS STUDY 44 45 • We will conduct a random-effects network meta-analysis to synthesize all available published 46 or unpublished randomized controlled trials for each pre-specified outcome, and obtain a 47 comprehensive ranking of all treatments. 48 • This will be the first network meta-analysis on psychological treatments for schizophrenia; 49 the findings from this study have the potential to inform and influence clinical decision- 50 51 making and guideline development. 52 • The risk of heterogeneity and inconsistency is high, given the different psychological 53 interventions that will be included: however, we try to control variability by carefully framing 54 the inclusion criteria about population and interventions, and we will evaluate consistency 55 56 employing local as well as global methods. 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 • The limitations of primary studies will be addressed with the Cochrane risk of bias tool and 4 the quality of evidence for network estimates will be assessed with an appropriate 5 adaptation of the GRADE framework; these approaches are considered the gold standard for 6 critical appraisal of evidence quality. 7 8 9 INTRODUCTION 10 11 Schizophrenia is a debilitating, and often life-long disorder that ranks among the top 20 causes of 12 disability according to the World Health Report (1). Although pharmacological interventions have 13 been the mainstay of treatment for schizophrenia, antipsychotics have a number of limitations 14 (limited response, high incidence of disabling side effects, poor adherence to treatment) (2) and are 15 problematic in many situations (such as medical comorbidities, tolerability problems and pregnancy). 16 For peer review only 17 Besides, there has been growing recognition of the importance of psychological processes in 18 psychosis, both as contributors to onset and persistence, and in terms of the negative psychological 19 impact of a diagnosis of schizophrenia on the individual’s well-being, psychosocial functioning and 20 life opportunities. Psychological interventions for psychosis and schizophrenia have been developed 21 to address these aspects, and, in accordance with guidelines from the National Institute for Health 22 23 and Care Excellence in the United Kingdom (3) and the Schizophrenia Patient Outcomes Research 24 Team in the United States (4), psychological treatments are widely regarded as a necessary 25 intervention for schizophrenia. 26 A broad range of interventions that can be defined as “psychological” have been studied in the 27 treatment of schizophrenia. These interventions can be provided at different stages of the illness and 28 29 address different aspects, like social and cognitive functioning, adherence to medication and 30 symptoms of schizophrenia. Table 1 presents the panorama of existing systematic reviews of 31 randomized controlled trials that have been conducted on the topic. These reviews have mainly 32 included studies comparing the intervention under examination with so called no treatment http://bmjopen.bmj.com/ 33 34 conditions (waiting-list, treatment as usual (TAU)) (5, 6). Other reviews included also active 35 comparisons with other psychological treatments (7–9). An attempt to provide information on active 36 comparisons was made by Turner and colleagues, who performed separate meta-analysis when 37 there were at least five eligible randomized controlled trials comparing one intervention to another 38 psychological intervention (10). However, all the available reviews applied pairwise meta-analysis as 39 40 a method, being able to pool results only when a comparison of two treatments was considered in on September 24, 2021 by guest. Protected copyright. 41 existing studies. The comparative efficacy and tolerability of the existing interventions has not been 42 checked yet; as a result, it is still currently unclear which are the most efficacious, the most 43 acceptable and the best tolerable psychological treatments for schizophrenia. 44 To overcome this gap in the current knowledge, a network meta-analysis would be necessary to 45 46 consider both direct and indirect comparisons, and produce hierarchies of the effects of the various 47 psychological treatments in the various efficacy and tolerability outcomes. Such hierarchies are 48 essential for guidelines, which should ideally be able to indicate which treatment is likely to be the 49 best, the second best and so on for a given outcome. Only the method of network meta-analysis can 50 provide such hierarchies by combining all the randomised evidence. Our aim is to produce such a 51 52 network meta-analysis of all psychological interventions for schizophrenia in multiple outcomes. We 53 focus here on the interventions primarily aimed at treating positive symptoms in the acute phase of 54 the illness. 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3

4 5 Intervention Existing reviews RCT* Comparator Acceptance and Ongoing Cochrane TAU, pharmacological intervention, another 6 - 7 commitment therapy review (11) psychosocial intervention 8 Adherence interventions Gray 2016 (12) 6 TAU, didactic health education Active comparisons 9 Befriending, CBT, Cognitive remediation, (Befriending, CBT, 10 Psychoeducation, Social skills training, Supportive Cognitive remediation, 11 Turner 2014 (10) 48 counseling, Family intervention, Art therapy, Body Psychoeducation, Social 12 psychotherapy, Occupational therapy, Problem- skills training, Supportive 13 solving therapy 14 counseling) 15 Art therapy Ruddy 2005 (6) 2 Standard care 16 Assertive community TAU, hospital-based rehabilitation, case ForMarshall peer 1998 (13) review20** only 17 treatment management 18 Befriending -

19 Bibliotherapy -

20 Body oriented - 21 psychological therapy 22 Assertive Community Treatment, Assertive 23 Case management Dieterich 2017 (14) 40** Outreach model, Case Management model, 24 standard community care 25 Zimmermann 2005 26 (positive symptoms) 15 Waiting-list, TAU or another therapeutic treatment 27 (9) 28 Active (Psychoeducation, Family Intervention, 29 Supportive Psychotherapy, Supportive Counselling, 30 Jones 2012 (7) 20 Cognitive Remediation) and nonactive control 31 treatments (Recreation and support, social 32 activities, befriending, non-specific counselling) http://bmjopen.bmj.com/ 33 Waiting list, TAU or an intervention designed to 34 control for the non-specific effects of 35 psychotherapy (recreation and support, group 36 Cognitive behavioural support, befriending, supportive Jauhar 2013 (15) 52 37 therapy counselling/therapy, social activity therapy and 38 goal-focused supportive contact) or active 39 treatments (cognitive remediation, 40 psychoeducation) on September 24, 2021 by guest. Protected copyright. 41 Van der Gaag 2014 42 (individually tailored) 18 Any control condition was accepted 43 (16) Hazell 2016 (low 44 8 TAU, supportive psychotherapy 45 intensity) (17) Kennedy 2016 46 Non-specialized therapy (focused on supportive (auditory 2 47 interactions and social integration) 48 hallucinations) (18) TAU, active control (e.g. computer games) another 49 Cognitive remediation Cella 2016 (19) 45 50 active treatment (e.g. CBT) 51 Dance therapy Ren 2013 (20) 1 Standard care plus supportive counselling Pitschel-Walz 2001 TAU, patient intervention, other family 52 25 (21) interventions 53 Family interventions TAU, discussion groups, psychoeducation, 54 Pharoah 2006 (22) 25 55 supportive psychotherapy, psychosocial support Group psychotherapeutic TAU, other groups (active discussion group, 56 Orfanos 2015 (23) 34 57 treatments support group, counselling group, occupational 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 therapy group or problem-solving discussion 4 group) Integrated psychological TAU, placebo-attention condition, other active 5 Roder 2006 (24) 16 6 Therapy (IPT) treatments Psychological and 7 TAU, active comparator (including psychosocial 8 Lutgens 2017 (8) 95 psychoeducation, supportive therapy, cognitive interventions for negative 9 remediation) 10 symptoms in psychosis 11 TAU, Wait list control, Supportive Therapy, 12 Metacognitive training Eichner 2016 (25) 12 Newspaper discussion group, CogPack (=Cognitive 13 remediation) Active control intervention (e.g. Befriending, 14 Mindfulness Aust 2017 (26) 11** 15 Progressive Muscle Relaxation), TAU Placebo defined as an alternative therapy designed 16 Geretsegger 2017 Music therapy For peer review18 to control for effectsonly of the therapist’s attention; 17 (27) TAU or no treatment 18 Reality-adaptive, supportive psychotherapy, 19 Hospital comparison, Ataraxic drugs, electro 20 Psychodynamic therapy Malmberg 2012 (28) 4 convulsive therapy, Milieu therapy, individual vs. 21 group 22 TAU, supportive psychotherapy, behavioral 23 Psychoeducation Pekkala 2006 (29) 10 intervention, leisure-time group 24 TAU, structured activities, discussion group, 25 Social skills training Almerie 2015 (30) 13 interaction group, no treatment control 26 Standard care, any other treatment (biological, 27 psychological or social) such as medication, 28 Supportive therapy Buckley 2015 (31) 24 problem-solving therapy, psychoeducation, social 29 skills training, CBT, family therapy or 30 psychodynamic psychotherapy 31 Systemic therapy Pinquart 2016 (5) 7 No treatment 32

http://bmjopen.bmj.com/ 33 Table 1. Existing reviews about psychological treatments for schizophrenia. 34 RCT: randomized controlled trial; TAU: treatment as usual. 35 *number of RCTs on schizophrenic patients; **RCTs about patients with severe mental illness including schizophrenia 36 37 38 Objectives 39 40 To estimate relative treatment effects and obtain a hierarchy for the psychological treatments in on September 24, 2021 by guest. Protected copyright. 41 patients with schizophrenia, in terms of: 42 1. Efficacy on positive symptoms 43 2. Acceptability 44 45 3. Other efficacy measures, such as overall symptoms, negative symptoms, response, relapse, 46 adherence, depression, quality of life, functioning 47 4. Tolerability. 48 49 50 METHODS AND ANALYSIS 51 Criteria for considering studies for this review 52 Methods for this systematic review have been developed according to the Preferred Reporting Items 53 54 for Systematic review and Meta-Analysis Protocols (PRISMA-P) checklist, and the PRISMA extension 55 statement for reporting of systematic reviews incorporating network meta-analyses of healthcare 56 interventions (32, 33). This systematic review and NMA is registered in the PROSPERO database 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 (registration number: CRD42017067795); the record in PROSPERO will be updated with any 4 amendment made to the protocol. 5 6 7 Types of studies 8 We will include all randomised trials (RCTs) in which participants with schizophrenia received a 9 psychological intervention as defined below (see Types of interventions). Studies whose sequence 10 generation was at high risk of bias (e.g. randomization by the date of birth or day of the week) will be 11 12 excluded. We will accept open and blinded RCTs; this choice is particularly relevant in trials on 13 psychological interventions, in which in best case only the assessor of outcome can be blind, but not 14 the therapist. Open RCTs will be excluded in a sensitivity analysis. We will include both trials in which 15 psychological interventions were compared with a control condition and trials in which they were 16 compared with anotherFor intervention. peer There reviewwill be no language restrictiononly in order to avoid the 17 18 problem of ‘language bias’ (34). In case we retrieve references in languages in which we are not 19 fluent, study authors will be contacted to check inclusion criteria and eventually ask for study data. 20 As an exception, we will not search Chinese databases, since serious concerns have been raised on 21 the trustworthiness of Chinese trials found in these databases (35, 36). Chinese studies found in 22 Western databases will be considered for inclusion. In the case of cross-over studies we will use only 23 24 the first cross-over phase in order to avoid the problem of carry-over effects which are very likely in 25 schizophrenia and with psychological treatments. We will exclude cluster randomized trials. 26 27 28 Types of participants 29 Our aim is to collect information on the efficacy of psychological treatments on patients with positive 30 31 symptoms. In order to select this population, we operationalized the inclusion criteria as follows. We 32 will include adults, however defined by study authors, with a diagnosis of schizophrenia or related http://bmjopen.bmj.com/ 33 disorders (such as schizophreniform, or schizoaffective disorders); there is no clear evidence that the 34 latter schizophrenia-like psychoses are caused by fundamentally different disease processes or 35 require different treatment approaches (37). We will include trials irrespective of the diagnostic 36 37 criteria used. Here we will follow the strategy of the Cochrane Schizophrenia Group (38) to include 38 not only studies that used specific diagnostic criteria such as ICD-10 or DSM-V, because these criteria 39 are not meticulously used in clinical routine either. This decision should increase generalizability and 40 representativeness. on September 24, 2021 by guest. Protected copyright. 41 Studies including participants with other diagnoses part of the psychosis spectrum will be included 42 43 only if participants with a diagnosis of schizophrenia, schizophreniform or schizoaffective disorders 44 were more than 80% of the participants considered. We will include studies recruiting patients with 45 positive symptoms, either delusions, hallucinations or both, or in the phase of acute exacerbation of 46 positive symptoms, however defined by inclusion criteria of the trial. 47 48 We will exclude studies focused on specific subpopulations of patients, such as (1) studies recruiting 49 patients in which negative symptoms are predominant, according to authors’ definition, (2) studies 50 on patients with comorbid psychiatric disorders including substance abuse, (3) studies recruiting 51 patients with concomitant medical illnesses, (4) trials enrolling stable patients (relapse prevention 52 studies), (5) studies on first episode patients, (6) trials on patients who show prodromal signs of 53 54 psychosis (also defined as “at risk for psychosis”). 55 56 Types of interventions 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Any psychological intervention that occurs through interaction between therapist and patient, either 4 face-to-face individually or in group, with the primary aim to reduce positive symptoms. 5 Interventions with an explicit primary aim different from positive symptoms (e.g. functioning, 6 cognition, adherence to medication, knowledge of the illness) will be excluded. The identified 7 8 treatments will be classified after identification of eligible studies. Psychological treatments will be 9 compared to each other and to any non-pharmacological control condition considered in the 10 included studies. Comparators will include the so called “treatment as usual”, waiting list and inactive 11 treatments. The effect of “non-active” comparators will be analysed in a sensitivity analysis (39). 12 Patients also receiving treatment as usual, including pharmacological interventions, will be included. 13 14 If psychological treatments that we do not include among the interventions (e.g. psychoeducation, 15 supportive therapy) are used as control condition in the studies, they will be included as nodes in 16 order to strengthenFor the network, peer but will not review be part of our decision only set. 17 18 19 Outcome measures 20 Outcomes will be measured at study endpoint, as defined in each study. 21 22 23 Primary outcome 24 25 Change in positive symptoms of schizophrenia, examined accordingly to the respective subscale of 26 the Positive and Negative Syndrome Scale (PANSS), the Brief Psychiatric Rating Scale (BPRS) or the 27 28 Scales for Assessment of Positive Symptoms (SAPS) or any other published scale. 29 As not all studies will have used the same scale, we will extract data according to the following 30 hierarchy: mean change of the PANSS positive symptoms subscale from baseline to endpoint, if not 31 available mean change of the BPRS positive symptoms subscale, or if again not available the mean 32 values at endpoint of the PANSS/BPRS positive symptoms subscale. The results of other rating scales http://bmjopen.bmj.com/ 33 34 will only be used if the instrument has been published in a peer-reviewed journal, because it has 35 been shown that non-validated schizophrenia scales exaggerate differences (40). 36 37 Secondary outcomes 38 39 Given the focus on treatments for positive symptoms, the results of this review will be informative 40 for the treatment of positive symptoms. They will also describe how these interventions can have an on September 24, 2021 by guest. Protected copyright. 41 effect on a number of other outcomes. With this aim, the following secondary outcomes will be 42 assessed: 43

44 45 1. Acceptability, defined as the percentage of patients leaving the study early (‘dropout’) for 46 any reason. All-cause discontinuation due to any reason combines efficacy, tolerability, 47 and other factors and can therefore be considered as a measure of ‘acceptability of 48 treatment’ (38) or of overall “effectiveness”; 49 2. Change in overall symptoms, measured by rating scales such as the PANSS or the BPRS, or 50 51 any other published scale (e.g. the Manchester Scale) for the assessment of overall 52 schizophrenic symptomatology. The results of other rating scales will only be used if the 53 instrument has been published in a peer-reviewed journal; 54 3. Change in negative symptoms, measured by the respective subscale of the PANSS, or the 55 “Scales for Assessment of Negative Symptoms” (SANS) or any other published scale; 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 4. Response, measured by the percentage of responders defined by reduction on the PANSS, 4 BPRS or CGI scores, accepting the criteria used by study authors; 5 5. Percentage of patients with relapse, by definitions operationalized by rating scales, and, if 6 not available, number of rehospitalisations due to psychopathology. We will not include 7 8 data from studies that used non-operationalized relapse criteria (e.g. “clinical 9 judgement”); 10 6. Adherence, measured by any published rating scale (e.g. “Adherence Therapy Patients 11 Satisfaction Questionnaire”, “Adherence Rating Scale”); 12 7. Depression, measured by the Calgary Depression Scale for Schizophrenia, the Hamilton 13 14 Depression Rating Scale, the Montgomery Asberg Depression Scale or other published 15 symptom scales; 16 8. QualityFor of life, measuredpeer by any review published rating scaleonly (e.g. “Heinrichs quality of life 17 scale”, Quality of Life Scale (QOLS); 18 9. Functioning, measured by rating scales such as the Global Assessment of Functioning or 19 20 the Psychosocial Performance Scale, or any other published rating scale. 21 10. Tolerability, measured as the percentage of patients experiencing adverse events. 22 Adverse events associated with psychological treatments are not covered as 23 comprehensively as in trials on pharmacological treatments (41). However, there is a 24 raising awareness of the importance of considering possible harms associated with 25 26 psychological interventions (42). Therefore we will collect any available information in 27 clinical studies about this outcome, using a classification proposed by Linden and 28 colleagues (42): a) emergence of new symptoms; b) deterioration of existing symptoms; c) 29 lack of improvement or deterioration of illness; d) prolongation of treatment; e) patient’s 30 non-compliance; f) strains in the patient-therapist relationship; g) very good patient- 31 32 therapist relationship, therapy dependency; h) strains or changes in family relations; i) http://bmjopen.bmj.com/ 33 strains or changes in work relations; l) any change in the life circumstances of the 34 patient; m) stigmatization. Suicide attempts and any other possible adverse event related 35 to psychological treatment will also be considered. 36 37 11. Mortality. Psychosocial treatments may actually reduce or, by contrast, increase overall 38 mortality, in particular connected to suicidality. To test this, we will examine this outcome 39 in terms of a) death for any reason, b) death due to natural causes and c) due to suicide. 40 on September 24, 2021 by guest. Protected copyright. 41 42 43 Search strategy 44 45 Electronic searches 46 The following sources will be searched without restrictions for language or publication period: 47 EMBASE, MEDLINE, PsycINFO, PUBMED. The search terms that will be used for PubMed are provided 48 as Supplement material. We will also search the following international databases: 49 1. WHO International Clinical Trials Registry Platform (ICTRP) 50 51 2. BIOSIS 52 3. Cochrane Collaboration Controlled Trials Register 53 4. ClinicalTrials.gov. 54 55 Reference lists and other sources 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 References of all selected studies will be inspected for other published reports and citations of 4 unpublished studies. We will also inspect previous reviews conducted on psychological treatments 5 for schizophrenia to check if some studies meet our inclusion criteria as well. In addition, we will 6 contact the first author of each included study published in the last 30 years for missing information 7 8 about their studies. 9 10 Identification and selection of studies 11 Studies identified through electronic and manual searches will be listed with citation, titles and 12 13 abstracts, in Citavi; duplicates will be excluded. The eligibility for inclusion process will be conducted 14 in two separate stages: 15 1. Two authors will independently inspect title and abstracts identified in the literature searches, 16 and exclude thoseFor not pertinent. peer Disagreement review will be resolved only by discussion, and where doubt 17 18 still remains, we will acquire the full article for further inspection and the article will proceed to 19 the next stage; 20 2. Once the full articles are obtained, two reviewers will independently assess them for eligibility. 21 Disagreements will be resolved by discussion, and, if needed, a third senior author will be 22 involved. When required, further information will be obtained from study authors. 23 24 25 Data extraction 26 Two authors will independently extract data from all selected trials. When disagreement arises we 27 28 will resolve it by discussion and, if needed, involving a third senior author. Where this is not sufficient 29 we will contact the study authors. 30 The following data will be collected from each included study: 31 • Study citation, year(s) of study, registration number to trials registries, year of publication, 32 location, setting, number of centers, sample size, diagnostic criteria, funding/sponsor (industry or http://bmjopen.bmj.com/ 33 34 academic); 35 • Methodology, including study design (type of RCT), number of arms, risk of bias (see below); 36 • Characteristics of study participants, including gender, age, details on diagnosis, number 37 randomized to each arm, sociodemographic characteristics, whether psychological treatments 38 39 naïve at baseline, or with previous experience with the experimental intervention); 40 • Characteristics of intervention, including number and frequency of sessions, therapy setting, on September 24, 2021 by guest. Protected copyright. 41 expertise of therapist, researcher allegiance at study arm level; 42 • Outcome measures, including information on whether an Intention to Treat (ITT) approach has 43 been used and how it was defined. 44 45 The two reviewers will independently input data into an Access database especially created for this 46 study. The software will automatically detect any inconsistencies, and they will be resolved by 47 discussion. 48 49 50 Measurement of treatment effect 51 Relative treatment effects 52 • Continuous outcomes: For continuous outcomes we will use the standardized mean difference 53 (SMD), because we expect that the studies use different rating scales of overall 54 55 schizophrenia symptomatology. 56 • Dichotomous outcomes: The effect size for dichotomous outcomes will be the risk ratio (RR) and 57 its 95% confidence intervals (CIs). 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 4 Relative treatment ranking 5 We will estimate the probability for each intervention to be ranked at each possible place, given the 6 relative effect sizes as estimated in NMA. As described in Salanti et al (43) we will obtain a hierarchy 7 8 of the competing interventions using the Surface Under the Cumulative Ranking curve (SUCRA) and 9 mean ranks. SUCRA values will be expressed as percentage, showing the relative probability of an 10 intervention to be among the best options. 11 12 Dealing with missing outcome data and missing statistics 13 14 For continuous outcomes we will extract data for all randomized patients if possible, and we will give 15 preference to data based on mixed-effect models of repeated measurements of multiple imputations 16 over last-observation-carried-forwardFor peer data. review only 17 We will use published standard deviations (SDs), where available. When standard errors instead of 18 SDs are presented, the former will be converted to SDs (44). If both are missing we will estimate SDs 19 20 from p-values or confidence intervals, as described in Section 7.7.3 of the Cochrane Handbook for 21 Systematic Reviews (45). If none of these options is viable we will contact the original authors. When 22 no information can be obtained we will derive SDs from those of the other studies using a validated 23 imputation technique (44). 24 25 For dichotomous outcomes, everyone allocated to the intervention will be counted whether they 26 completed the follow up or not. If the authors applied such a strategy, we will use their results. If the 27 original authors presented only the results of the per-protocol or completer population, we will 28 assume that those participants lost to follow-up would not have changed in a given outcome. In 29 terms of efficacy this means that they would be conservatively considered to have not responded to 30 31 treatment or control. In terms of tolerability it would mean that participants would not have 32 developed a side-effect. http://bmjopen.bmj.com/ 33 34 35 Risk of bias assessment 36 Risk of bias will be assessed for each included study using the Cochrane Collaboration ‘risk of bias’ 37 tool (45, 46). The following domains will be considered: 38 39 1. Sequence generation: was the allocation sequence adequately generated? 40 2. Allocation concealment: was allocation adequately concealed? on September 24, 2021 by guest. Protected copyright. 41 3. Blinding of participants: was knowledge of the allocated treatment adequately prevented during 42 the study? Given the peculiarity of the included studies, in which the therapist cannot be blind, we 43 will consider under this item only if a way was found to keep patients unaware of the treatment they 44 45 were receiving (even if we expect this will not be likely); 46 4. Blinding of outcome assessors: were outcome evaluated by blind raters? Were adequate measures 47 taken to prevent them to discover treatment allocation during the study? 48 5. Incomplete outcome data: were incomplete outcome data adequately addressed? 49 6. Selective reporting: are reports of the study free from suggestion of selective outcome reporting? 50 51 7. Researcher’s allegiance: do the researchers involved have a vested interest for the psychological 52 treatment under investigation? We will additionally consider this point as possible source of bias, 53 since it has been claimed to be relevant in trials on psychological interventions (47, 48, 49). An 54 evaluation of high risk of bias will be given, for example, when the authors are founders of the 55 therapy or have written a manual for that therapy. 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 A description of what was reported about the same domains in each study will be provided, and a 4 judgement on the risk of bias will be made for each one of them, based on the following three 5 categories: ‘high risk of bias’, ‘low risk of bias’ and ‘unclear risk of bias’ where information are not 6 sufficient to make a judgement. Two independent review authors will assess the risk of bias in 7 8 selected studies. Any disagreement will be resolved through discussion. Where necessary, the 9 authors of the studies will be contacted for further information. Studies will be classified as having 10 low risk of bias if none of the domains above was rated as high risk of bias and three or less were 11 rated as unclear risk; moderate if one was rated as high risk of bias or none was rated as high risk of 12 bias but four or more were rated as unclear risk, and all other cases will be assumed to pertain to 13 14 high risk of bias (50). We will not include studies in the data analyses whose sequence generation 15 was at high risk of bias (e.g. randomization by the date of birth or day of the week). Effects of high 16 risk of bias in the otherFor domains peer will be analyzed review by sensitivity analyses. only 17 18 19 Data analysis 20 Characteristics of the included studies 21 We will produce descriptive statistics and study population characteristics across all eligible trials, 22 describing the types of comparisons and other clinical or methodological variables, such as age, 23 24 duration of illness, co-medication, country, duration of study and number of session. 25 26 Two-step procedure 27 In a first step we will perform series of conventional pair-wise meta-analyses by combining studies 28 29 that compared the same interventions, including the comparison between active treatments and the 30 different control arms. In subgroups with very few RCTs available or if the requirements of network 31 meta-analysis are not met it can be that network meta-analysis will not be appropriate and in this 32 case, conventional pairwise meta-analysis will be the most straightforward approach. As http://bmjopen.bmj.com/ 33 heterogeneity is likely, a random effects model will be used. In a second step we will then perform a 34 35 network meta-analysis within a frequentist framework. 36 37 38 Assessment of heterogeneity 39 The heterogeneity (variability in relative treatment effects within the same treatment comparison) 40 will be measured with the tau-squared (the variance of the random effects distribution). The on September 24, 2021 by guest. Protected copyright. 41 heterogeneity variance will be assumed common across the various treatment comparisons (grouped 42 by comparison type) and the empirical distributions will be used to characterize the amount of 43 44 heterogeneity as low, moderate or high using the first and third quantiles (51–53). Potential reasons 45 for heterogeneity will be explored by subgroup analysis (see below). 46 47 Assessment of the transitivity assumption 48 49 Joint analysis of treatments can be misleading if the network is substantially intransitive. We assume 50 that patients who fulfill the inclusion criteria outlined in criteria for considering studies for this 51 review section are equally likely to be randomised to any of the interventions that we plan to 52 compare. We will need to investigate the distribution of clinical and methodological variables that 53 can act as effect modifiers across treatment comparisons (54). We have maximized the chances of 54 55 transitivity in our network with regard to clinical variables by limiting our samples to participants 56 with schizophrenia and excluding specific subgroups like first episode patients or patients with 57 prevalent negative symptoms. Other clinical or methodological variables that may influence the 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 efficacy of psychological interventions include administration mode and frequency of the treatment 4 (like number of sessions and experience of the therapist), baseline severity (see below, “Investigation 5 of heterogeneity and inconsistency”), and blinding, which will also be assessed in sensitivity analyses. 6 We will investigate if these variables are similarly distributed across studies grouped by comparison. 7 8 The comparability of studies comparing the intervention with treatment as usual or waiting list 9 conditions with those that provide head-to-head evidence will be examined carefully. 10 11 Network meta-analysis 12 13 Network meta-analysis combines direct and indirect evidence for all relative treatment effects and 14 can therefore provide estimates with maximum power and increased precision (55). If the collected 15 studies appear to be sufficiently similar with respect to the distribution of effect modifiers (refer 16 Assessment of transitivityFor assumption peer section), review we will conduct aonly random-effects NMA to synthesize 17 all evidence for each outcome, and obtain a comprehensive ranking of all treatments. We will 18 19 assume a single heterogeneity parameter for each network. We will present the summary SMDs or 20 RRs for all pairwise comparisons in a league table. We will also estimate the prediction intervals to 21 assess how much the common heterogeneity affects the relative effect with respect to the extra 22 uncertainty anticipated in a future study. To rank the various treatments for each outcome, we will 23 use the surface under the cumulative ranking curve (SUCRA) and the mean ranks. 24 25 26 Assessment of inconsistency 27 28 The strategical and conceptual evaluation of transitivity will be supplemented with a statistical 29 evaluation of consistency, the agreement between direct and indirect evidence. We will employ local 30 as well as global methods to evaluate consistency (56). Local methods detect ‘hot spots’ of 31 inconsistency, evidence loops that are inconsistent or comparisons for which direct and indirect 32 evidence disagree. We will employ a method that separates direct evidence from indirect evidence http://bmjopen.bmj.com/ 33 34 provided by the entire network and then evaluate the agreement of these two pieces of evidence 35 (57). We will also evaluate consistency in the entire network by calculating the design-by-treatment 36 interaction test and I-squared for network heterogeneity, inconsistency, and for both (58). Tests for 37 inconsistency are known to have low power, and empirical evidence has suggested that 10% of 38 evidence loops published in the medical literature are expected to be inconsistent (60). Therefore, 39 40 interpretation of the statistical inference about inconsistency will be carried out with caution and on September 24, 2021 by guest. Protected copyright. 41 possible sources of inconsistency will be explored even in the absence of evidence for inconsistency. 42 43 Investigation of heterogeneity and inconsistency 44 45 We expect small amounts of heterogeneity and inconsistency to be present given the variety of study 46 settings we plan to include. The following potential effect modifiers of the primary outcome will be 47 explored by subgroup analyses: 48 a) Number of sessions 49 b) Study duration 50 51 c) Setting: individual vs group 52 d) Expertise of the therapist 53 e) Baseline severity (PANSS or BPRS score at baseline) 54 f) Different types of patients, with a different clinical outline concerning symptoms (if identified). 55

56 57 Sensitivity analyses 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 We will explore the following sensitivity analyses by excluding: 4 a) studies in which the outcome assessor was not blind (open studies) 5 b) studies that presented only completer analyses 6 c) studies characterized as pertaining to high risk of bias 7 8 d) studies with high risk of bias in researchers’ allegiance 9 e) studies focused on treatment resistant patients (study defined) 10 f) studies with a non-active comparison group. 11 12 Publication bias 13 14 We will first examine funnel plots of pairwise MAs if there are 10 or more studies included. We will 15 also explore the association between study size and effect size with a comparison-adjusted funnel 16 plot that has beenFor adapted topeer network meta-analysis review (59). only 17 18 19 Evaluating the quality of the evidence 20 The quality of evidence contributing to each network estimate will be evaluated using an adaptation 21 of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework 22 specifically developed for network meta-analysis (56). We will characterize the credibility of a body 23 of evidence based on the study limitations, imprecision, heterogeneity/inconsistency, indirectness, 24 25 and publication bias. 26 27 Statistical software 28 The analysis and presentation of results will be performed using the Stata packages network and 29 30 network_graphs, the R package netmeta. 31 32 ETHICS AND DISSEMINATION http://bmjopen.bmj.com/ 33 34 This review does not require ethical approval. Findings will be published in peer reviewed scientific 35 journals, granting open access, and the database will be made publicly available (in agreement with 36 European Research Council Guidelines on Implementation of Open Access to Scientific Publications 37 and Research Data (http://ec.europa.eu/research/participants/data/ref/h2020/other/hi/oa- 38 39 pilot/h2020-hi-erc-oa-guide_en.pdf). 40 on September 24, 2021 by guest. Protected copyright. 41 42 REFERENCES 43 1. Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M et al. Years lived with disability (YLDs) for 1160 44 45 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global Burden of Disease Study 46 2010. Lancet 2012; 380(9859):2163–96. 47 2. Leucht S, Cipriani A, Spineli L, Mavridis D, Orey D, Richter F et al. Comparative efficacy and tolerability of 15 48 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet 2013; 382(9896):951–62. 49 50 3. National Collaborating Centre for Mental Health. Core interventions in the treatment and management of 51 schizophrenia in adults in primary and secondary care (Clinical Guideline CG82). London; 2009. 52 53 4. Buchanan RW, Kreyenbuhl J, Kelly DL, Noel JM, Boggs DL, Fischer BA et al. The 2009 schizophrenia PORT 54 psychopharmacological treatment recommendations and summary statements. Schizophr Bull 2010; 36(1):71– 55 93. 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 5. Pinquart M, Oslejsek B, Teubert D. Efficacy of systemic therapy on adults with mental disorders: A meta- 4 analysis. Psychother Res 2016; 26(2):241–57. 5 6. Ruddy R, Milnes D. Art therapy for schizophrenia or schizophrenia-like illnesses. Cochrane Database Syst Rev 6 2005; (4):CD003728. 7 8 7. Jones C, Hacker D, Cormac I, Meaden A, Irving CB. Cognitive behaviour therapy versus other psychosocial 9 treatments for schizophrenia. Cochrane Database Syst Rev 2012; (4):CD008712. 10 11 8. Lutgens D, Gariepy G, Malla A. Psychological and psychosocial interventions for negative symptoms in 12 psychosis: Systematic review and meta-analysis. Br J Psychiatry 2017; 210(5):324–32. 13 9. Zimmermann G, Favrod J, Trieu VH, Pomini V. The effect of cognitive behavioral treatment on the positive 14 symptoms of schizophrenia spectrum disorders: A meta-analysis. Schizophr Res 2005; 77(1):1–9. 15 16 10. Turner DT, vanFor der Gaag M,peer Karyotaki E, Cuijpersreview P. Psychological only interventions for psychosis: a meta- 17 analysis of comparative outcome studies. Am J Psychiatry 2014; 171(5):523–38. 18 19 11. Naeem F, Asmer MS, Khoury B, Kingdon D, Farooq S. Acceptance and Commitment Therapy for 20 schizophrenia and related disorders (Protocol). Cochrane Database of Systematic Reviews 2015; (6). 21 12. Gray R, Bressington D, Ivanecka A, Hardy S, Jones M, Schulz M et al. Is adherence therapy an effective 22 adjunct treatment for patients with schizophrenia spectrum disorders? A systematic review and meta-analysis. 23 BMC Psychiatry 2016; 16:90. 24 25 13. Marshall M, Lockwood A. Assertive community treatment for people with severe mental disorders. 26 Cochrane Database Syst Rev 2000; (2):CD001089. 27 28 14. Dieterich M, Irving CB, Bergman H, Khokhar MA, Park B, Marshall M. Intensive case management for severe 29 mental illness. Cochrane Database Syst Rev 2017; 1:CD007906. 30 15. Jauhar S, McKenna PJ, Radua J, Fung E, Salvador R, Laws KR. Cognitive-behavioural therapy for the 31 symptoms of schizophrenia: systematic review and meta-analysis with examination of potential bias. Br J 32 http://bmjopen.bmj.com/ 33 Psychiatry 2014; 204(1):20–9. 34 16. van der Gaag M, Valmaggia LR, Smit F. The effects of individually tailored formulation-based cognitive 35 behavioural therapy in auditory hallucinations and delusions: A meta-analysis. Schizophr Res 2014; 156(1):30– 36 7. 37 38 17. Hazell CM, Hayward M, Cavanagh K, Strauss C. A systematic review and meta-analysis of low intensity CBT 39 for psychosis. Clin Psychol Rev 2016; 45:183–92. 40 on September 24, 2021 by guest. Protected copyright. 41 18. Kennedy L, Xyrichis A. Cognitive Behavioral Therapy Compared with Non-specialized Therapy for Alleviating 42 the Effect of Auditory Hallucinations in People with Reoccurring Schizophrenia: A Systematic Review and Meta- 43 analysis. Community Ment Health J 2017; 53(2):127–33. 44 19. Cella M, Preti A, Edwards C, Dow T, Wykes T. Cognitive remediation for negative symptoms of 45 schizophrenia: A network meta-analysis. Clin Psychol Rev 2017; 52:43–51. 46 47 20. Ren J, Xia J. Dance therapy for schizophrenia. Cochrane Database Syst Rev 2013; (10):CD006868. 48 49 21. Pitschel-Walz G, Leucht S, Bauml J, Kissling W, Engel RR. The effect of family interventions on relapse and 50 rehospitalization in schizophrenia--a meta-analysis. Schizophr Bull 2001; 27(1):73–92. 51 22. Pharoah F, Mari J, Rathbone J, Wong W. Family intervention for schizophrenia. Cochrane Database Syst Rev 52 2010; (12):CD000088. 53 54 23. Orfanos S, Banks C, Priebe S. Are Group Psychotherapeutic Treatments Effective for Patients with 55 Schizophrenia? A Systematic Review and Meta-Analysis. Psychother Psychosom 2015; 84(4):241–9. 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 24. Roder V, Mueller DR, Mueser KT, Brenner HD. Integrated psychological therapy (IPT) for schizophrenia: Is it 4 effective? Schizophr Bull 2006; 32 Suppl 1:S81-93. 5 25. Eichner C, Berna F. Acceptance and Efficacy of Metacognitive Training (MCT) on Positive Symptoms and 6 Delusions in Patients With Schizophrenia: A Meta-analysis Taking Into Account Important Moderators. 7 Schizophr Bull 2016; 42(4):952–62. 8 9 26. Aust J, Bradshaw T. Mindfulness interventions for psychosis: A systematic review of the literature. J 10 Psychiatr Ment Health Nurs 2017; 24(1):69–83. 11 12 27. Geretsegger M, Mössler KA, Bieleninik Ł, Chen X-J, Heldal TO, Gold C. Music therapy for people with 13 schizophrenia and schizophrenia-like disorders. Cochrane Database Syst Rev 2017; 5:CD004025. 14 28. Malmberg L, Fenton M. Individual psychodynamic psychotherapy and psychoanalysis for schizophrenia and 15 severe mental illness. Cochrane Database Syst Rev 2001; (3):CD001360. 16 For peer review only 17 29. Pekkala E, Merinder L. Psychoeducation for schizophrenia. Cochrane Database Syst Rev 2000; 18 (4):CD002831. 19 20 30. Almerie MQ, Okba Al Marhi M, Jawoosh M, Alsabbagh M, Matar HE, Maayan N et al. Social skills 21 programmes for schizophrenia. Cochrane Database Syst Rev 2015; (6):CD009006. 22 31. Buckley LA, Maayan N, Soares-Weiser K, Adams CE. Supportive therapy for schizophrenia. Cochrane 23 Database Syst Rev 2015; (4):CD004716. 24 25 32. Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C et al. The PRISMA extension 26 statement for reporting of systematic reviews incorporating network meta-analyses of health care 27 interventions: Checklist and explanations. Ann Intern Med 2015; 162(11):777–84. 28 29 33. Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M et al. Preferred reporting items for 30 systematic review and meta-analysis protocols (PRISMA-P) 2015: Elaboration and explanation. BMJ 2015; 31 349:g7647. 32 http://bmjopen.bmj.com/ 33 34. Egger M, Zellweger-Zahner T, Schneider M, Junker C, Lengeler C, Antes G. Language bias in randomised 34 controlled trials published in English and German. Lancet 1997; 350(9074):326–9. 35 35. Wu T, Li Y, Liu G, Bian Z, Li J, Zhang J, Xie L, Ni J, editor. Investigation of authenticity of 'claimed' randomized 36 controlled trials (RCTs) and quality assessment of RCT reports published in China; 2006. 37 38 36. Woodhead M. 80% of China's clinical trial data are fraudulent, investigation finds. BMJ 2016; 355:i5396. 39 40 37. Carpenter WT, JR, Buchanan RW. Schizophrenia. N Engl J Med 1994; 330(10):681–90. on September 24, 2021 by guest. Protected copyright. 41 38. Adams,C.E., Coutinho,E., Davis,J.M., Duggan,L., Essali,A., Fenton,M., Li,C., Jayaram,M., Leucht,S., Tharyan,P., 42 Välimäki,M. Cochrane Schizophrenia Group. The Cochrane Library. Chichester, UK: John Wiley & Sons Ltd; 43 2011. 44 45 39. Furukawa TA, Noma H, Caldwell DM, Honyashiki M, Shinohara K, Imai H et al. Waiting list may be a nocebo 46 condition in psychotherapy trials: a contribution from network meta-analysis. Acta Psychiatr Scand 2014; 47 130(3):181–92. 48 49 40. Marshall M, Lockwood A, Bradley C, Adams C, Joy C, Fenton M. Unpublished rating scales: a major source of 50 bias in randomised controlled trials of treatments for schizophrenia. Br J Psychiatry 2000; 176:249–52. 51 41. Vaughan B, Goldstein MH, Alikakos M, Cohen LJ, Serby MJ. Frequency of reporting of adverse events in 52 randomized controlled trials of psychotherapy vs. psychopharmacotherapy. Compr Psychiatry 2014; 55(4):849– 53 55. 54 55 42. Linden M, Schermuly-Haupt M-L. Definition, assessment and rate of psychotherapy side effects. World 56 Psychiatry 2014; 13(3):306–9. 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 43. Salanti G, Ades AE, Ioannidis JPA. Graphical methods and numerical summaries for presenting results from 4 multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol 2011; 64(2):163–71. 5 44. Furukawa TA, Barbui C, Cipriani A, Brambilla P, Watanabe N. Imputing missing standard deviations in meta- 6 analyses can provide accurate results. J Clin Epidemiol 2006; 59(1):7–10. 7 8 45. Higgins JPT, editor. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 [updated 9 March 2011]; 2011. 10 11 46. Higgins JPT, Altman DG, Sterne JAC (editors), editor. Assessing risk of bias in included studies. In: Higgins 12 JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic Reviews of 13 Interventions version 5.2.0 (updated June 2017); Cochrane, 2017. Available from: URL: 14 http://training.cochrane.org/handbook. 15 47. Dragioti E, Dimoliatis I, Evangelou E. Disclosure of researcher allegiance in meta-analyses and randomised 16 For peer review only 17 controlled trials of psychotherapy: a systematic appraisal. BMJ Open 2015; 5(6):e007206. 18 48. Lieb K, Osten-Sacken J von der, Stoffers-Winterling J, Reiss N, Barth J. Conflicts of interest and spin in 19 reviews of psychological therapies: a systematic review. BMJ Open 2016; 6(4):e010606. 20 21 49. Munder T, Brutsch O, Leonhart R, Gerger H, Barth J. Researcher allegiance in psychotherapy outcome 22 research: an overview of reviews. Clin Psychol Rev 2013; 33(4):501–11. 23 50. Furukawa TA, Salanti G, Atkinson LZ, Leucht S, Ruhe HG, Turner EH et al. Comparative efficacy and 24 25 acceptability of first-generation and second-generation antidepressants in the acute treatment of major 26 depression: Protocol for a network meta-analysis. BMJ Open 2016; 6(7):e010919. 27 51. Rhodes KM, Turner RM, White IR, Jackson D, Spiegelhalter DJ, Higgins JPT. Implementing informative priors 28 for heterogeneity in meta-analysis using meta-regression and pseudo data. Stat Med 2016; 35(29):5495–511. 29 30 52. Rhodes KM, Turner RM, Higgins JPT. Empirical evidence about inconsistency among studies in a pair-wise 31 meta-analysis. Res Synth Methods 2016; 7(4):346–70. 32 http://bmjopen.bmj.com/ 33 53. Turner RM, Davey J, Clarke MJ, Thompson SG, Higgins JP. Predicting the extent of heterogeneity in meta- 34 analysis, using empirical data from the Cochrane Database of Systematic Reviews. Int J Epidemiol 2012; 35 41(3):818–27. 36 54. Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: Many 37 names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods 38 39 2012; 3(2):80–97. 40

55. Salanti G, Higgins JPT, Ades AE, Ioannidis JPA. Evaluation of networks of randomized trials. Stat Methods on September 24, 2021 by guest. Protected copyright. 41 Med Res 2008; 17(3):279–301. 42 43 56. Salanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JPT. Evaluating the quality of evidence from a 44 network meta-analysis. PLoS One 2014; 9(7):e99682. 45 57. Dias S, Welton NJ, Caldwell DM, Ades AE. Checking consistency in mixed treatment comparison meta- 46 47 analysis. Stat Med 2010; 29(7-8):932–44. 48 58. Higgins JPT, Jackson D, Barrett JK, Lu G, Ades AE, White IR. Consistency and inconsistency in network meta- 49 analysis: Concepts and models for multi-arm studies. Res Synth Methods 2012; 3(2):98–110. 50 51 59. Chaimani A, Salanti G. Using network meta-analysis to evaluate the existence of small-study effects in a 52 network of interventions. Res Synth Methods 2012; 3(2):161–76. 53 60. Veroniki AA, Vasiliadis HS, Higgins JPT, Salanti G. Evaluation of inconsistency in networks of interventions. 54 55 Int J Epidemiol 2013; 42(1):332–45. 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Authors’ contributions: IB and SL designed this study, drafted and critically revised the 4 protocol. IB will screen search results for inclusion, conduct data extraction and data analysis, and 5 draft the final manuscript. SL will assist with data extraction and analysis, and revise the final 6 7 manuscript. CR and SW will screen search results for inclusion and conduct data extraction. GS 8 provided substantial methodological advice in planning the study, and will assist with data analysis. 9 CB and TF contributed with clinical and methodological input in planning the study. All authors 10 contributed to and have approved the final manuscript. 11

12 13 Collaborators: Samantha Roberts helped us to conduct the literature searches. Maximilian Huhn, 14 Johannes Schneider-Thoma, Marc Krause and Costanza Carmi provided help and suggestions. 15

16 For peer review only 17 Funding: This project has received funding from the European Union’s Horizon 2020 Research and 18 Innovation Programme under the Marie Skłodowska-Curie grant agreement N° 701717. This work 19 was also supported by the German Research Foundation (DFG) and the Technical University of 20 21 Munich within the Open Access Publishing Funding Programme. The funder had no role in developing 22 the protocol. 23 24 Competing interests: SL in the past 3 years has received honoraria for consulting from Roche, 25 26 TEVA, Otsuka, Lundbeck, and LB Pharma; for lectures from Otsuka, Lundbeck, Janssen, ICON, Lilly, 27 Sanofi Aventis, AOP Orphan, Roche, and Servier; and for a publication from Roche. TAF has received 28 lecture fees from Eli Lilly, Janssen, Meiji, Mitsubishi-Tanabe, MSD and Pfizer and consultancy fees 29 from Takeda Science Foundation. He has received royalties from Igaku-Shoin and Nihon Bunka 30 31 Kagaku-sha publishers. He has received research support from Mochida and Mitsubishi-Tanabe. 32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 40 on September 24, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 22

1 2 3

Search terms for Pubmed BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from 4 5 6 ("Schizophrenia"[Mesh] OR "Paranoid Disorders"[Mesh] OR schizo*[Title/Abstract] OR 7 psychotic*[Title/Abstract] OR psychosis[Title/Abstract] OR psychoses[Title/Abstract]) AND 8 9 ("Psychotherapy"[Mesh] or "Behavior Therapy"[Mesh] or "Cognitive Therapy"[Mesh] or 10 "Complementary Therapies"[Mesh] or "Psychoanalysis"[Mesh] or "Counseling"[Mesh] or 11 "Hypnosis"[Mesh] or "Association"[Mesh] or "Association Learning"[Mesh] OR 12 abreaction[Title/Abstract] OR "acceptance[Title/Abstract] AND commitment therapy"[Title/Abstract] 13 14 OR "acting out"[Title/Abstract] OR adlerian[Title/Abstract] OR "analytical 15 psychotherapy"[Title/Abstract] OR "analytical psychotherapies"[Title/Abstract] OR "anger 16 control"[Title/Abstract] OR "anger management"[Title/Abstract] OR "animal therapy"[Title/Abstract] 17 18 OR "animal therapies"[Title/Abstract]For peer review OR "art therapy"[Title/Abstract]only OR "art 19 therapies"[Title/Abstract] OR "assertive training"[Title/Abstract] OR "assertiveness 20 training"[Title/Abstract] OR "attention training technique"[Title/Abstract] OR "autogenic 21 22 training"[Title/Abstract] OR autosuggestion[Title/Abstract] OR "aversion therapy"[Title/Abstract] OR 23 "aversion therapies"[Title/Abstract] OR "balint group"[Title/Abstract] OR befriending[Title/Abstract] 24 OR "behavior contracting"[Title/Abstract] OR "behavior modification"[Title/Abstract] OR "behavior 25 regulation"[Title/Abstract] OR "behavior therapy"[Title/Abstract] OR "behavior 26 27 therapies"[Title/Abstract] OR "behaviour contracting"[Title/Abstract] OR "behaviour 28 modification"[Title/Abstract] OR "behaviour regulation"[Title/Abstract] OR "behaviour 29 therapy"[Title/Abstract] OR "behaviour therapies"[Title/Abstract] OR bibliotherapy[Title/Abstract] 30 31 OR bibliotherapies[Title/Abstract] OR biofeedback[Title/Abstract] OR "body 32 psychotherapy"[Title/Abstract] OR "body psychotherapies"[Title/Abstract] OR "brief 33 psychotherapy"[Title/Abstract] OR "brief psychotherapies"[Title/Abstract] OR "caregiver 34 support"[Title/Abstract] OR cbt[Title/Abstract] OR "client centre"[Title/Abstract] OR "client 35 36 center"[Title/Abstract] OR "cognitive behavior"[Title/Abstract] OR "cognitive 37 behaviorial"[Title/Abstract] OR "cognitive intervention"[Title/Abstract] OR "cognitive http://bmjopen.bmj.com/ 38 interventions"[Title/Abstract] OR "cognitive rehabilitation"[Title/Abstract] OR "cognitive 39 40 remediation"[Title/Abstract] OR "cognitive technique"[Title/Abstract] OR "cognitive 41 techniques"[Title/Abstract] OR "cognitive therapy"[Title/Abstract] OR "cognitive 42 therapies"[Title/Abstract] OR "cognitive treatment"[Title/Abstract] OR "cognitive 43 treatments"[Title/Abstract] OR "color therapy"[Title/Abstract] OR "color therapies"[Title/Abstract] 44 45 OR "colour therapy"[Title/Abstract] OR "colour therapies"[Title/Abstract] OR "compassionate mind on September 24, 2021 by guest. Protected copyright. 46 training"[Title/Abstract] OR "conjoint therapy"[Title/Abstract] OR "conjoint therapies"[Title/Abstract] 47 OR "contingency management"[Title/Abstract] OR "conversational therapy"[Title/Abstract] OR 48 49 "conversational therapies"[Title/Abstract] OR "conversion therapy"[Title/Abstract] OR "conversion 50 therapies"[Title/Abstract] OR "coping skills"[Title/Abstract] OR counseling[Title/Abstract] OR 51 counselling[Title/Abstract] OR countertransference[Title/Abstract] OR "couples 52 therapy"[Title/Abstract] OR "couples therapies"[Title/Abstract] OR "covert 53 54 sensitization"[Title/Abstract] OR "covert sensitisation"[Title/Abstract] OR "crisis 55 intervention"[Title/Abstract] OR "dance therapy"[Title/Abstract] OR "dance 56 therapies"[Title/Abstract] OR dialectic[Title/Abstract] OR dialectical[Title/Abstract] OR "dream 57 58 analysis"[Title/Abstract] OR eclectic[Title/Abstract] OR "emotion focused"[Title/Abstract] OR 59 "emotionally focused"[Title/Abstract] OR "emotional freedom technique"[Title/Abstract] OR 60 "encounter group therapy"[Title/Abstract] OR "encounter group therapies"[Title/Abstract] OR "existential therapy"[Title/Abstract] OR "existential therapies"[Title/Abstract] OR "experiential

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 22 BMJ Open

1 2 3 psychotherapy"[Title/Abstract] OR "experiential psychotherapies"[Title/Abstract] OR "exposure 4 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from therapy"[Title/Abstract] OR "exposure therapies"[Title/Abstract] OR "expressive 5 6 psychotherapy"[Title/Abstract] OR "expressive psychotherapies"[Title/Abstract] OR "eye movement 7 desensitization"[Title/Abstract] OR "eye movement desensitisation"[Title/Abstract] OR "family 8 intervention"[Title/Abstract] OR "family interventions"[Title/Abstract] OR "family 9 10 therapy"[Title/Abstract] OR "family therapies"[Title/Abstract] OR "feminist therapy"[Title/Abstract] 11 OR "feminist therapies"[Title/Abstract] OR "free association"[Title/Abstract] OR 12 freudian[Title/Abstract] OR "geriatric psychotherapy"[Title/Abstract] OR "geriatric 13 psychotherapies"[Title/Abstract] OR "gestalt therapy"[Title/Abstract] OR "gestalt 14 15 therapies"[Title/Abstract] OR griefwork[Title/Abstract] OR "group intervention"[Title/Abstract] OR 16 "group interventions"[Title/Abstract] OR "group psychotherapy"[Title/Abstract] OR "group 17 psychotherapies"[Title/Abstract] OR "group therapy"[Title/Abstract] OR "group 18 For peer review only 19 therapies"[Title/Abstract] OR "guided imagery"[Title/Abstract] OR "holistic 20 psychotherapy"[Title/Abstract] OR "holistic psychotherapies"[Title/Abstract] OR "humanistic 21 psychotherapy"[Title/Abstract] OR "humanistic psychotherapies"[Title/Abstract] OR 22 hypnosis[Title/Abstract] OR hypnotherapy[Title/Abstract] OR hypnotherapies[Title/Abstract] OR 23 24 hypnotizability[Title/Abstract] OR hypnotisability[Title/Abstract] OR imagery[Title/Abstract] OR 25 "implosive therapy"[Title/Abstract] OR "implosive therapies"[Title/Abstract] OR "individual 26 psychotherapy"[Title/Abstract] OR "individual psychotherapies"[Title/Abstract] OR "insight 27 28 therapy"[Title/Abstract] OR "insight therapies"[Title/Abstract] OR "integrated psychological 29 therapy"[Title/Abstract] OR "integrative psychotherapy"[Title/Abstract] OR "integrative 30 psychotherapies"[Title/Abstract] OR "integrative therapy"[Title/Abstract] OR "integrative 31 therapies"[Title/Abstract] OR interpersonal[Title/Abstract] OR jungian[Title/Abstract] OR 32 33 kleinian[Title/Abstract] OR logotherapy[Title/Abstract] OR logotherapies[Title/Abstract] OR 34 "marathon group therapy"[Title/Abstract] OR "marathon group therapies"[Title/Abstract] OR 35 "marital therapy"[Title/Abstract] OR "marital therapies"[Title/Abstract] OR meditation[Title/Abstract] 36 37 OR "mental healing"[Title/Abstract] OR "metacognitive therapy"[Title/Abstract] OR "metacognitive http://bmjopen.bmj.com/ 38 therapies"[Title/Abstract] OR "metacognitive training"[Title/Abstract] OR "milieu 39 therapy"[Title/Abstract] OR "milieu therapies"[Title/Abstract] OR mindfulness[Title/Abstract] OR 40 "morita therapy"[Title/Abstract] OR "morita therapies"[Title/Abstract] OR multimodal[Title/Abstract] 41 42 OR "music therapy"[Title/Abstract] OR "music therapies"[Title/Abstract] OR "narrative 43 therapy"[Title/Abstract] OR "narrative therapies"[Title/Abstract] OR "nondirective 44

therapy"[Title/Abstract] OR "nondirective therapies"[Title/Abstract] OR "object on September 24, 2021 by guest. Protected copyright. 45 46 relations"[Title/Abstract] OR "person centred therapy"[Title/Abstract] OR "person centred 47 therapies"[Title/Abstract] OR "person centered therapy"[Title/Abstract] OR "person centered 48 therapies"[Title/Abstract] OR "personal construct therapy"[Title/Abstract] OR "personal construct 49 therapies"[Title/Abstract] OR "persuasion therapy"[Title/Abstract] OR "persuasion 50 51 therapies"[Title/Abstract] OR "pet therapy"[Title/Abstract] OR "pet therapies"[Title/Abstract] OR 52 "play therapy"[Title/Abstract] OR "play therapies"[Title/Abstract] OR "primal therapy"[Title/Abstract] 53 OR "primal therapies"[Title/Abstract] OR "problem solving"[Title/Abstract] OR 54 55 psychoanalyse[Title/Abstract] OR psychoanalysed[Title/Abstract] OR psychoanalysis[Title/Abstract] 56 OR psychoanalytic[Title/Abstract] OR psychodrama[Title/Abstract] OR psychodynamic[Title/Abstract] 57 OR psychoeducate[Title/Abstract] OR psychoeducation[Title/Abstract] OR 58 psychoeducating[Title/Abstract] OR psychologic[Title/Abstract] OR psychological[Title/Abstract] OR 59 60 psychologically[Title/Abstract] OR "psychological therapy"[Title/Abstract] OR "psychological therapies"[Title/Abstract] OR "psychosocial treatment"[Title/Abstract] OR "psychosocial

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 22

1 2 3 treatments"[Title/Abstract] OR psychotherapy[Title/Abstract] OR psychotherapies[Title/Abstract] OR 4 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from "psychotherapeutic counsel"[Title/Abstract] OR "psychotherapeutic counseling"[Title/Abstract] OR 5 6 "psychotherapeutic counselling"[Title/Abstract] OR "psychotherapeutic processes"[Title/Abstract] OR 7 "psychotherapeutic training"[Title/Abstract] OR "psychotherapeutic treatment"[Title/Abstract] OR 8 "psychotherapeutic treatments"[Title/Abstract] OR "rational emotive"[Title/Abstract] OR "reality 9 10 therapy"[Title/Abstract] OR "reality therapies"[Title/Abstract] OR "reciprocal 11 inhibition"[Title/Abstract] OR rehabilitation[Title/Abstract] OR rehabilitating[Title/Abstract] OR 12 "relationship therapy"[Title/Abstract] OR "relationship therapies"[Title/Abstract] OR 13 relaxation[Title/Abstract] OR "reminiscence therapy"[Title/Abstract] OR "reminiscence 14 15 therapies"[Title/Abstract] OR rogerian[Title/Abstract] OR "role play"[Title/Abstract] OR "role 16 plays"[Title/Abstract] OR "role playing"[Title/Abstract] OR "self analysis"[Title/Abstract] OR "self 17 analysing"[Title/Abstract] OR "self esteem"[Title/Abstract] OR ""[Title/Abstract] OR 18 For peer review only 19 "sex therapy"[Title/Abstract] OR "sex therapies"[Title/Abstract] OR "sleep phase 20 chronotherapy"[Title/Abstract] OR "sleep phase chronotherapies"[Title/Abstract] OR "social skills 21 education"[Title/Abstract] OR "social skills training"[Title/Abstract] OR "socioenvironmental 22 therapy"[Title/Abstract] OR "socioenvironmental therapies"[Title/Abstract] OR 23 24 sociotherapy[Title/Abstract] OR sociotherapies[Title/Abstract] OR "solution focused"[Title/Abstract] 25 OR "stress management"[Title/Abstract] OR "support group"[Title/Abstract] OR "support 26 groups"[Title/Abstract] OR "supportive therapy"[Title/Abstract] OR "supportive 27 28 therapies"[Title/Abstract] OR "systematic desensitization"[Title/Abstract] OR "systematic 29 desensitisation"[Title/Abstract] OR "systemic therapy"[Title/Abstract] OR "systemic 30 therapies"[Title/Abstract] OR "therapeutic community"[Title/Abstract] OR "therapeutic 31 communities"[Title/Abstract] OR "transactional analysis"[Title/Abstract] OR 32 33 transference[Title/Abstract] OR transtheoretical[Title/Abstract] OR "validation 34 therapy"[Title/Abstract] OR "validation therapies"[Title/Abstract]) AND (((randomized controlled 35 trial[pt]) OR (controlled clinical trial[pt]) OR (randomized[tiab]) OR (placebo[tiab]) OR (drug 36 37 therapy[sh]) OR (randomly[tiab]) OR (trial[tiab]) OR (groups[tiab])) NOT (animals[mh] NOT http://bmjopen.bmj.com/ 38 humans[mh])) 39 40 41 42 43 44 45 on September 24, 2021 by guest. Protected copyright. 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: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

2 8 5 3 1 5 16 16 58 58 1617 1617 1617 1617 1 (title) 1 Reported on Reported pageon Supplementary Supplementary file BMJ Open Checklist item http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected copyright.

For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml registers registers or grey literatureother sources) with planned coveragedates of Provide anexplicit Provide statement the review question(s)of address will the with referenceto participants, Provide name, Provide institutional affiliation, email address authors; of all protocol provide physical address mailing Describe all intended Describe information (suchsources as electronic contact databases, with study trial authors, Specify study characteristicsSpecify the as (such PICO, design,study setting, time frame) report and characteristics (such as years considered, as (such language, years publication to status) be as used criteria for eligibilityreview for the that it could it repeatedthat be interventions, interventions, comparators, and (PICO) outcomes of corresponding authorof If the the protocolIf anrepresents amendment previously a completed of or published protocol, and as identify such changes; list otherwise, state fordocumentingplan important amendments protocol

6 6 the rationale Describe for theincontext the review of what is already known 2 2 registered, If provide namethe the registry (such as PROSPERO)of and registration number 7 7 9 9 5c rolesfunder(s), Describe of and/orsponsor(s), institution(s), if in any, developing the protocol

Role of Role of sponsor

Sponsor 5b name Provide review for the funder and/or sponsor Sources funderor 5a Indicate financialsources of support reviewor forthe other Contact 3a Update 1b the foranprotocol If update is a systematicprevious of identifyreview, as such Identification 1a report the as Identify protocol systematicaa of review Contributions 3b contributions Describe authorsof protocol and identify the the guarantor reviewof Study records:Study Search Search strategy 10 search draft Present forto usedatbe strategy leastof electronic one database, including planned limits, such Information Information sources Eligibility criteriaEligibility 8 METHODS Objectives Rationale Rationale Support: Authors: Title: ADMINISTRATIVE INFORMATION PRISMA-P (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols) 2015 checklist: recommended items to to items recommended checklist: 2015 Protocols) Meta-Analysis review for and Systematic Items (PreferredReporting PRISMA-P reviewprotocol* a systematic in address Section topic and No Item Registration INTRODUCTION Amendments 4 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 Page 21 of 22 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from Page 22 of 9 9 9 13 13 10 10 12 12 89 89 89 89 78 78 11, 12 12, 12, 13 9,

BMJ Open http://bmjopen.bmj.com/ bias(es) (suchbias(es) as publication bias across studies, selective reporting ti ti A,Petticrew M, P, Shekelle Stewart L, PRISMA-P Preferred Group. reportingitems for systematic andreview d ind conjunctionwith Explanationthe PRISMA-P and Elaboration (citewhen available) for important on September 24, 2021 by guest. Protected copyright.

For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml , τ)Kendall’s 2 Describe Describe planned extractingmethod of from (suchdata reports as piloting forms, independently,in done State the the State that process be (suchwill for used selecting studies independenttwo reviewers)as eachthrough Describe anticipatedDescribe methods assessing for risk of bias of studies, individual including whether this be will the at done level,or study outcome both; state this or informationhow will in used databe synthesis Specify any planned Specify assessment of meta within studies)within List and all define List forvariables which data (such be soughtwill asPICO items, funding sources), any pre List and all define List for outcomes whichdata sought, be including will prioritization of mainadditional and rationaleoutcomes, with duplicate), any duplicate), processes for and obtaining confirming data from investigators data and and combiningdata methods of data from studies, including planned explorationany of consistency(such as I phase of the phase review of is,(that screening, andeligibility inclusionin metaanalysis) planned data assumptionsandplanned simplifications 12 12 14 14 17 17 strength Describe how the the body evidenceof of willassessed be as(such GRADE) 13 13 11c 11c 11a 11a the mechanism(s) Describe that be will to used recordsmanage and thedata throughout review 15c 15c any Describe proposed additional analyses as (such or subgroupsensitivity analyses, metaregression) 11b 15b data are for If appropriate synthesis,quantitative describe planned summary methods handling measures, of 15d quantitative If synthesis is not appropriate, describe typethe of summary planned

collection Data Selection Data management process process It is It stronglyrecommendedchecklist this that be rea

clarification theclarification items. on to Amendments a review protocol should be tracked copyrightand dated. The for (includingPRISMA-P checklist) byheld the is Group PRISMA-P and is distributedunder a Creative Commons Attribution Licence 4.0.

* Confidence inConfidence evidencecumulative Data Data synthesis 15a criteria under Describe data which study be synthesisedwill quantitatively Risk Risk in of bias studiesindividual Metabias(es) 16 Outcomes and Outcomes meta-analysis protocols(PRISMA-P) elaboration2015: explanation. andBMJ. Jan 2015 2;349(jan02 1):g7647. Data itemsData From: Shamseer From: L, Moher D, Clarke M, Ghersi Libera D, prioritization 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 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

Psychological interventions for positive symptoms in schizophrenia: protocol for a network meta-analysis of randomized controlled trials

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2017-019280.R2

Article Type: Protocol

Date Submitted by the Author: 28-Dec-2017

Complete List of Authors: Bighelli, Irene; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Salanti, Georgia; University of Bern, Institute of Social and Preventive Medicine (ISPM) Reitmeir, Cornelia; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Wallis, Sofia; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy Barbui, Corrado; University of Verona, WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation, Department of Neuroscience, Biomedicine and Movement Sciences, Section of Psychiatry

Furukawa, Toshi; Kyoto University, Department of Health Promotion and http://bmjopen.bmj.com/ Human Behavior, Graduate School of Medicine and School of Public Health Leucht, Stefan; Klinikum rechts der Isar der Technischen Universitat Munchen, Department of Psychiatry and Psychotherapy

Primary Subject Mental health Heading:

Secondary Subject Heading: Mental health

Schizophrenia & psychotic disorders < PSYCHIATRY, Adult psychiatry < on September 24, 2021 by guest. Protected copyright. Keywords: PSYCHIATRY, Psychological interventions, Network Meta-Analysis, Psychotherapy

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Psychological interventions for positive symptoms in 4 5 schizophrenia: protocol for a network meta-analysis of 6 7 randomized controlled trials 8 9 10 11 12 Irene Bighelli1*, Georgia Salanti2, Cornelia Reitmeir1, Sofia Wallis1, Corrado Barbui3, Toshi A. 13 Furukawa4, Stefan Leucht1 14 15 16 For peer review only 17 18 19 1 Department of Psychiatry and Psychotherapy, Klinikum rechts der Isar, Technical University of 20 21 Munich, Germany 22 23 2 Institute of Social and Preventive Medicine (ISPM), University of Bern, Switzerland 24 25 3 WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation, 26 27 Department of Neuroscience, Biomedicine and Movement Sciences, Section of Psychiatry, University 28 of Verona, Italy 29 30 4 Department of Health Promotion and Human Behavior, Kyoto University Graduate School of 31 Medicine / School of Public Health, Japan 32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 *Corresponding author 40 Department of Psychiatry and Psychotherapy on September 24, 2021 by guest. Protected copyright. 41 Klinikum rechts der Isar, Technische Universität München 42 Ismaningerstr. 22, 81675 München, Germany 43 44 E-mail: [email protected] 45 46 GS: [email protected] 47 CR: [email protected] 48 SW: [email protected] 49 50 CB: [email protected] 51 TAF: [email protected] 52 SL: [email protected] 53 54

55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 ABSTRACT 4 5 Introduction 6 There is rising awareness that we need multi-disciplinary approaches integrating psychological 7 treatments for schizophrenia, but a comprehensive evidence base on their relative efficacy is lacking. 8 We will conduct a network meta-analysis (NMA), integrating direct and indirect comparisons from 9 10 randomised controlled trials (RCTs) to rank psychological treatments for schizophrenia according to 11 their efficacy, acceptability and tolerability. 12 13 Methods and analysis 14 15 We will include all RCTs comparing a psychological treatment aimed at positive symptoms of 16 schizophrenia withFor another psychologicalpeer intervention review or with a noonly treatment condition (waiting list, 17 treatment as usual). We will include studies on adult patients with schizophrenia, excluding specific 18 subpopulations (e.g. first episode patients, or patients with psychiatric comorbidities). Primary 19 20 outcome will be the change in positive symptoms on a published rating scale. Secondary outcomes 21 will be acceptability (dropout), change in overall and negative symptoms of schizophrenia, response, 22 relapse, adherence, depression, quality of life, functioning and adverse events. Published and 23 unpublished studies will be sought through database searches, trial registries and websites. Study 24 selection and data extraction will be conducted by at least two independent reviewers. We will 25 26 conduct random-effects NMA to synthesize all evidence for each outcome and obtain a 27 comprehensive ranking of all treatments. NMA will be conducted in Stata and R within a frequentist 28 framework. The risk of bias in studies will be evaluated using the Cochrane Risk of Bias tool and the 29 credibility of the evidence will be evaluated using an adaptation of the GRADE framework to NMA, 30 recommended by the Cochrane guidance. Subgroup and sensitivity analyses will be conducted to 31 32 assess the robustness of the findings. http://bmjopen.bmj.com/ 33 34 Ethics and dissemination 35 No ethical issues are foreseen. Results from this study will be published in peer-reviewed journals 36 37 and presented at relevant conferences. 38 39 PROSPERO registration number: CRD42017067795 40 on September 24, 2021 by guest. Protected copyright. 41 42 43 STRENGTHS AND LIMITATIONS OF THIS STUDY 44 45 • We will conduct a random-effects network meta-analysis to synthesize all available published 46 or unpublished randomized controlled trials for each pre-specified outcome, and obtain a 47 comprehensive ranking of all treatments. 48 • This will be the first network meta-analysis on psychological treatments for schizophrenia; 49 the findings from this study have the potential to inform and influence clinical decision- 50 51 making and guideline development. 52 • The risk of heterogeneity and inconsistency is high, given the different psychological 53 interventions that will be included: however, we try to control variability by carefully framing 54 the inclusion criteria about population and interventions, and we will evaluate consistency 55 56 employing local as well as global methods. 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 • The limitations of primary studies will be addressed with the Cochrane risk of bias tool and 4 the quality of evidence for network estimates will be assessed with an appropriate 5 adaptation of the GRADE framework; these approaches are considered the gold standard for 6 critical appraisal of evidence quality. 7 8 9 INTRODUCTION 10 11 Schizophrenia is a debilitating, and often life-long disorder that ranks among the top 20 causes of 12 disability according to the World Health Report (1). Although pharmacological interventions have 13 been the mainstay of treatment for schizophrenia, antipsychotics have a number of limitations 14 (limited response, high incidence of disabling side effects, poor adherence to treatment) (2) and are 15 problematic in many situations (such as medical comorbidities, tolerability problems and pregnancy). 16 For peer review only 17 Besides, there has been growing recognition of the importance of psychological processes in 18 psychosis, both as contributors to onset and persistence, and in terms of the negative psychological 19 impact of a diagnosis of schizophrenia on the individual’s well-being, psychosocial functioning and 20 life opportunities. Psychological interventions for psychosis and schizophrenia have been developed 21 to address these aspects, and, in accordance with guidelines from the National Institute for Health 22 23 and Care Excellence in the United Kingdom (3) and the Schizophrenia Patient Outcomes Research 24 Team in the United States (4), psychological treatments are widely regarded as a necessary 25 intervention for schizophrenia. 26 A broad range of interventions that can be defined as “psychological” have been studied in the 27 treatment of schizophrenia. These interventions can be provided at different stages of the illness and 28 29 address different aspects, like social and cognitive functioning, adherence to medication and 30 symptoms of schizophrenia. Table 1 presents the panorama of existing systematic reviews of 31 randomized controlled trials that have been conducted on the topic. These reviews have mainly 32 included studies comparing the intervention under examination with so called no treatment http://bmjopen.bmj.com/ 33 34 conditions (waiting-list, treatment as usual (TAU)) (5, 6). Other reviews included also active 35 comparisons with other psychological treatments (7–9). An attempt to provide information on active 36 comparisons was made by Turner and colleagues, who performed separate meta-analysis when 37 there were at least five eligible randomized controlled trials comparing one intervention to another 38 psychological intervention (10). However, all the available reviews applied pairwise meta-analysis as 39 40 a method, being able to pool results only when a comparison of two treatments was considered in on September 24, 2021 by guest. Protected copyright. 41 existing studies. The comparative efficacy and tolerability of the existing interventions has not been 42 checked yet; as a result, it is still currently unclear which are the most efficacious, the most 43 acceptable and the best tolerable psychological treatments for schizophrenia. 44 To overcome this gap in the current knowledge, a network meta-analysis would be necessary to 45 46 consider both direct and indirect comparisons, and produce hierarchies of the effects of the various 47 psychological treatments in the various efficacy and tolerability outcomes. Such hierarchies are 48 essential for guidelines, which should ideally be able to indicate which treatment is likely to be the 49 best, the second best and so on for a given outcome. Only the method of network meta-analysis can 50 provide such hierarchies by combining all the randomised evidence. Our aim is to produce such a 51 52 network meta-analysis of all psychological interventions for schizophrenia in multiple outcomes. We 53 focus here on the interventions primarily aimed at treating positive symptoms in the acute phase of 54 the illness. 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3

4 5 Intervention Existing reviews RCT* Comparator Acceptance and Ongoing Cochrane TAU, pharmacological intervention, another 6 - 7 commitment therapy review (11) psychosocial intervention 8 Adherence interventions Gray 2016 (12) 6 TAU, didactic health education Active comparisons 9 Befriending, CBT, Cognitive remediation, (Befriending, CBT, 10 Psychoeducation, Social skills training, Supportive Cognitive remediation, 11 Turner 2014 (10) 48 counseling, Family intervention, Art therapy, Body Psychoeducation, Social 12 psychotherapy, Occupational therapy, Problem- skills training, Supportive 13 solving therapy 14 counseling) 15 Art therapy Ruddy 2005 (6) 2 Standard care 16 Assertive community TAU, hospital-based rehabilitation, case ForMarshall peer 1998 (13) review20** only 17 treatment management 18 Befriending -

19 Bibliotherapy -

20 Body oriented - 21 psychological therapy 22 Assertive Community Treatment, Assertive 23 Case management Dieterich 2017 (14) 40** Outreach model, Case Management model, 24 standard community care 25 Zimmermann 2005 26 (positive symptoms) 15 Waiting-list, TAU or another therapeutic treatment 27 (9) 28 Active (Psychoeducation, Family Intervention, 29 Supportive Psychotherapy, Supportive Counselling, 30 Jones 2012 (7) 20 Cognitive Remediation) and nonactive control 31 treatments (Recreation and support, social 32 activities, befriending, non-specific counselling) http://bmjopen.bmj.com/ 33 Waiting list, TAU or an intervention designed to 34 control for the non-specific effects of 35 psychotherapy (recreation and support, group 36 Cognitive behavioural support, befriending, supportive Jauhar 2013 (15) 52 37 therapy counselling/therapy, social activity therapy and 38 goal-focused supportive contact) or active 39 treatments (cognitive remediation, 40 psychoeducation) on September 24, 2021 by guest. Protected copyright. 41 Van der Gaag 2014 42 (individually tailored) 18 Any control condition was accepted 43 (16) Hazell 2016 (low 44 8 TAU, supportive psychotherapy 45 intensity) (17) Kennedy 2016 46 Non-specialized therapy (focused on supportive (auditory 2 47 interactions and social integration) 48 hallucinations) (18) TAU, active control (e.g. computer games) another 49 Cognitive remediation Cella 2016 (19) 45 50 active treatment (e.g. CBT) 51 Dance therapy Ren 2013 (20) 1 Standard care plus supportive counselling Pitschel-Walz 2001 TAU, patient intervention, other family 52 25 (21) interventions 53 Family interventions TAU, discussion groups, psychoeducation, 54 Pharoah 2006 (22) 25 55 supportive psychotherapy, psychosocial support Group psychotherapeutic TAU, other groups (active discussion group, 56 Orfanos 2015 (23) 34 57 treatments support group, counselling group, occupational 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 therapy group or problem-solving discussion 4 group) Integrated psychological TAU, placebo-attention condition, other active 5 Roder 2006 (24) 16 6 Therapy (IPT) treatments Psychological and 7 TAU, active comparator (including psychosocial 8 Lutgens 2017 (8) 95 psychoeducation, supportive therapy, cognitive interventions for negative 9 remediation) 10 symptoms in psychosis 11 TAU, Wait list control, Supportive Therapy, 12 Metacognitive training Eichner 2016 (25) 12 Newspaper discussion group, CogPack (=Cognitive 13 remediation) Active control intervention (e.g. Befriending, 14 Mindfulness Aust 2017 (26) 11** 15 Progressive Muscle Relaxation), TAU Placebo defined as an alternative therapy designed 16 Geretsegger 2017 Music therapy For peer review18 to control for effectsonly of the therapist’s attention; 17 (27) TAU or no treatment 18 Reality-adaptive, supportive psychotherapy, 19 Hospital comparison, Ataraxic drugs, electro 20 Psychodynamic therapy Malmberg 2012 (28) 4 convulsive therapy, Milieu therapy, individual vs. 21 group 22 TAU, supportive psychotherapy, behavioral 23 Psychoeducation Pekkala 2006 (29) 10 intervention, leisure-time group 24 TAU, structured activities, discussion group, 25 Social skills training Almerie 2015 (30) 13 interaction group, no treatment control 26 Standard care, any other treatment (biological, 27 psychological or social) such as medication, 28 Supportive therapy Buckley 2015 (31) 24 problem-solving therapy, psychoeducation, social 29 skills training, CBT, family therapy or 30 psychodynamic psychotherapy 31 Systemic therapy Pinquart 2016 (5) 7 No treatment 32

http://bmjopen.bmj.com/ 33 Table 1. Existing reviews about psychological treatments for schizophrenia. 34 RCT: randomized controlled trial; TAU: treatment as usual. 35 *number of RCTs on schizophrenic patients; **RCTs about patients with severe mental illness including schizophrenia 36 37 38 Objectives 39 40 To estimate relative treatment effects and obtain a hierarchy for the psychological treatments in on September 24, 2021 by guest. Protected copyright. 41 patients with schizophrenia, in terms of: 42 1. Efficacy on positive symptoms 43 2. Acceptability 44 45 3. Other efficacy measures, such as overall symptoms, negative symptoms, response, relapse, 46 adherence, depression, quality of life, functioning 47 4. Tolerability. 48 49 50 METHODS AND ANALYSIS 51 Criteria for considering studies for this review 52 Methods for this systematic review have been developed according to the Preferred Reporting Items 53 54 for Systematic review and Meta-Analysis Protocols (PRISMA-P) checklist, and the PRISMA extension 55 statement for reporting of systematic reviews incorporating network meta-analyses of healthcare 56 interventions (32, 33). This systematic review and NMA is registered in the PROSPERO database 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 (registration number: CRD42017067795); the record in PROSPERO will be updated with any 4 amendment made to the protocol. 5 6 7 Types of studies 8 We will include all randomised trials (RCTs) in which participants with schizophrenia received a 9 psychological intervention as defined below (see Types of interventions). Studies whose sequence 10 generation was at high risk of bias (e.g. randomization by the date of birth or day of the week) will be 11 12 excluded. We will accept open and blinded RCTs; this choice is particularly relevant in trials on 13 psychological interventions, in which in best case only the assessor of outcome can be blind, but not 14 the therapist. Open RCTs will be excluded in a sensitivity analysis. We will include both trials in which 15 psychological interventions were compared with a control condition and trials in which they were 16 compared with anotherFor intervention. peer There reviewwill be no language restrictiononly in order to avoid the 17 18 problem of ‘language bias’ (34). In case we retrieve references in languages in which we are not 19 fluent, study authors will be contacted to check inclusion criteria and eventually ask for study data. 20 As an exception, we will not search Chinese databases, since serious concerns have been raised on 21 the trustworthiness of Chinese trials found in these databases (35, 36). Chinese studies found in 22 Western databases will be considered for inclusion. In the case of cross-over studies we will use only 23 24 the first cross-over phase in order to avoid the problem of carry-over effects which are very likely in 25 schizophrenia and with psychological treatments. We will exclude cluster randomized trials. 26 27 28 Types of participants 29 Our aim is to collect information on the efficacy of psychological treatments on patients with positive 30 31 symptoms. In order to select this population, we operationalized the inclusion criteria as follows. We 32 will include adults, however defined by study authors, with a diagnosis of schizophrenia or related http://bmjopen.bmj.com/ 33 disorders (such as schizophreniform, or schizoaffective disorders); there is no clear evidence that the 34 latter schizophrenia-like psychoses are caused by fundamentally different disease processes or 35 require different treatment approaches (37). We will include trials irrespective of the diagnostic 36 37 criteria used. Here we will follow the strategy of the Cochrane Schizophrenia Group (38) to include 38 not only studies that used specific diagnostic criteria such as ICD-10 or DSM-V, because these criteria 39 are not meticulously used in clinical routine either. This decision should increase generalizability and 40 representativeness. on September 24, 2021 by guest. Protected copyright. 41 Studies including participants with other diagnoses part of the psychosis spectrum will be included 42 43 only if participants with a diagnosis of schizophrenia, schizophreniform or schizoaffective disorders 44 were more than 80% of the participants considered. We will include studies recruiting patients with 45 positive symptoms, either delusions, hallucinations or both, or in the phase of acute exacerbation of 46 positive symptoms, however defined by inclusion criteria of the trial. 47 48 We will exclude studies focused on specific subpopulations of patients, such as (1) studies recruiting 49 patients in which negative symptoms are predominant, according to authors’ definition, (2) studies 50 on patients with comorbid psychiatric disorders including substance abuse, (3) studies recruiting 51 patients with concomitant medical illnesses, (4) trials enrolling stable patients (relapse prevention 52 studies), (5) studies on first episode patients, (6) trials on patients who show prodromal signs of 53 54 psychosis (also defined as “at risk for psychosis”). Among other reasons, we exclude first episode 55 patients because they were found to have significantly higher response rates to treatments 56 compared to chronic patients (39, 40). 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 4 Types of interventions 5 Any psychological intervention that occurs through interaction between therapist and patient, either 6 face-to-face individually or in group, with the primary aim to reduce positive symptoms. 7 8 Interventions with an explicit primary aim different from positive symptoms (e.g. functioning, 9 cognition, adherence to medication, knowledge of the illness) will be excluded. The identified 10 treatments will be classified after identification of eligible studies. Psychological treatments will be 11 compared to each other and to any non-pharmacological control condition considered in the 12 included studies. Comparators will include the so called “treatment as usual”, waiting list and inactive 13 14 treatments. The effect of “non-active” comparators will be analysed in a sensitivity analysis (41). 15 Patients also receiving treatment as usual, including pharmacological interventions, will be included. 16 If psychological treatmentsFor thatpeer we do not review include among the interventionsonly (e.g. psychoeducation, 17 supportive therapy) are used as control condition in the studies, they will be included as nodes in 18 order to strengthen the network, but will not be part of our decision set. 19 20 21 Outcome measures 22 23 Outcomes will be measured at study endpoint, as defined in each study. 24 25 26 Primary outcome 27 28 Change in positive symptoms of schizophrenia, examined accordingly to the respective subscale of 29 the Positive and Negative Syndrome Scale (PANSS), the Brief Psychiatric Rating Scale (BPRS) or the 30 Scales for Assessment of Positive Symptoms (SAPS) or any other published scale. 31 As not all studies will have used the same scale, we will extract data according to the following 32 http://bmjopen.bmj.com/ 33 hierarchy: mean change of the PANSS positive symptoms subscale from baseline to endpoint, if not 34 available mean change of the BPRS positive symptoms subscale, or if again not available the mean 35 values at endpoint of the PANSS/BPRS positive symptoms subscale. The results of other rating scales 36 will only be used if the instrument has been published in a peer-reviewed journal, because it has 37 been shown that non-validated schizophrenia scales exaggerate differences (42). 38 39 40 Secondary outcomes on September 24, 2021 by guest. Protected copyright. 41 Given the focus on treatments for positive symptoms, the results of this review will be informative 42 for the treatment of positive symptoms. They will also describe how these interventions can have an 43 effect on a number of other outcomes. With this aim, the following secondary outcomes will be 44 45 assessed: 46 47 1. Acceptability, defined as the percentage of patients leaving the study early (‘dropout’) for 48 any reason. All-cause discontinuation due to any reason combines efficacy, tolerability, 49 50 and other factors and can therefore be considered as a measure of ‘acceptability of 51 treatment’ (38) or of overall “effectiveness”; 52 2. Change in overall symptoms, measured by rating scales such as the PANSS or the BPRS, or 53 any other published scale (e.g. the Manchester Scale) for the assessment of overall 54 schizophrenic symptomatology. The results of other rating scales will only be used if the 55 56 instrument has been published in a peer-reviewed journal; 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 3. Change in negative symptoms, measured by the respective subscale of the PANSS, or the 4 “Scales for Assessment of Negative Symptoms” (SANS) or any other published scale; 5 4. Response, measured by the percentage of responders defined by reduction on the PANSS, 6 BPRS or CGI scores, accepting the criteria used by study authors; 7 8 5. Percentage of patients with relapse, by definitions operationalized by rating scales, and, if 9 not available, number of rehospitalisations due to psychopathology. We will not include 10 data from studies that used non-operationalized relapse criteria (e.g. “clinical 11 judgement”); 12 6. Adherence, measured by any published rating scale (e.g. “Adherence Therapy Patients 13 14 Satisfaction Questionnaire”, “Adherence Rating Scale”); 15 7. Depression, measured by the Calgary Depression Scale for Schizophrenia, the Hamilton 16 DepressionFor Rating peer Scale, the Montgomery review Asberg Depressiononly Scale or other published 17 symptom scales; 18 8. Quality of life, measured by any published rating scale (e.g. “Heinrichs quality of life 19 20 scale”, Quality of Life Scale (QOLS); 21 9. Functioning, measured by rating scales such as the Global Assessment of Functioning or 22 the Psychosocial Performance Scale, or any other published rating scale. 23 10. Tolerability, measured as the percentage of patients experiencing adverse events. 24 Adverse events associated with psychological treatments are not covered as 25 26 comprehensively as in trials on pharmacological treatments (43). However, there is a 27 raising awareness of the importance of considering possible harms associated with 28 psychological interventions (44). Therefore we will collect any available information in 29 clinical studies about this outcome, using a classification proposed by Linden and 30 colleagues (44): a) emergence of new symptoms; b) deterioration of existing symptoms; c) 31 32 lack of improvement or deterioration of illness; d) prolongation of treatment; e) patient’s http://bmjopen.bmj.com/ 33 non-compliance; f) strains in the patient-therapist relationship; g) very good patient- 34 therapist relationship, therapy dependency; h) strains or changes in family relations; i) 35 strains or changes in work relations; l) any change in the life circumstances of the 36 37 patient; m) stigmatization. Suicide attempts and any other possible adverse event related 38 to psychological treatment will also be considered. 39 11. Mortality. Psychosocial treatments may actually reduce or, by contrast, increase overall 40

mortality, in particular connected to suicidality. To test this, we will examine this outcome on September 24, 2021 by guest. Protected copyright. 41 in terms of a) death for any reason, b) death due to natural causes and c) due to suicide. 42

43

44

45 46 Search strategy 47 Electronic searches 48 The following sources will be searched without restrictions for language or publication period: 49 EMBASE, MEDLINE, PsycINFO, PUBMED. The search terms that will be used for PubMed are provided 50 51 as Supplement material. We will also search the following international databases: 52 1. WHO International Clinical Trials Registry Platform (ICTRP) 53 2. BIOSIS 54 3. Cochrane Collaboration Controlled Trials Register 55 4. ClinicalTrials.gov. 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Reference lists and other sources 4 References of all selected studies will be inspected for other published reports and citations of 5 unpublished studies. We will also inspect previous reviews conducted on psychological treatments 6 for schizophrenia to check if some studies meet our inclusion criteria as well. In addition, we will 7 8 contact the first author of each included study published in the last 30 years for missing information 9 about their studies. 10 11 12 Identification and selection of studies 13 Studies identified through electronic and manual searches will be listed with citation, titles and 14 abstracts, in Citavi; duplicates will be excluded. The eligibility for inclusion process will be conducted 15 in two separate stages: 16 1. Two authors willFor independently peer inspect review title and abstracts identifiedonly in the literature searches, 17 18 and exclude those not pertinent. Disagreement will be resolved by discussion, and where doubt 19 still remains, we will acquire the full article for further inspection and the article will proceed to 20 the next stage; 21 2. Once the full articles are obtained, two reviewers will independently assess them for eligibility. 22 Disagreements will be resolved by discussion, and, if needed, a third senior author will be 23 24 involved. When required, further information will be obtained from study authors. 25 26 Data extraction 27 28 Two authors will independently extract data from all selected trials. When disagreement arises we 29 will resolve it by discussion and, if needed, involving a third senior author. Where this is not sufficient 30 we will contact the study authors. 31 The following data will be collected from each included study: 32 http://bmjopen.bmj.com/ 33 • Study citation, year(s) of study, registration number to trials registries, year of publication, 34 location, setting, number of centers, sample size, diagnostic criteria, funding/sponsor (industry or 35 academic); 36 • Methodology, including study design (type of RCT), number of arms, risk of bias (see below); 37 • Characteristics of study participants, including gender, age, details on diagnosis, number 38 39 randomized to each arm, sociodemographic characteristics, whether psychological treatments 40 naïve at baseline, or with previous experience with the experimental intervention); on September 24, 2021 by guest. Protected copyright. 41 • Characteristics of intervention, including number and frequency of sessions, therapy setting, 42 expertise of therapist, researcher allegiance at study arm level; 43 44 • Outcome measures, including information on whether an Intention to Treat (ITT) approach has 45 been used and how it was defined. 46 The two reviewers will independently input data into an Access database especially created for this 47 study. The software will automatically detect any inconsistencies, and they will be resolved by 48 discussion. 49 50 51 Measurement of treatment effect 52 Relative treatment effects 53 • Continuous outcomes: For continuous outcomes we will use the standardized mean difference 54 55 (SMD), because we expect that the studies use different rating scales of overall 56 schizophrenia symptomatology. 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 • Dichotomous outcomes: The effect size for dichotomous outcomes will be the risk ratio (RR) and 4 its 95% confidence intervals (CIs). 5 6 Relative treatment ranking 7 8 We will estimate the probability for each intervention to be ranked at each possible place, given the 9 relative effect sizes as estimated in NMA. As described in Salanti et al (45) we will obtain a hierarchy 10 of the competing interventions using the Surface Under the Cumulative Ranking curve (SUCRA) and 11 mean ranks. SUCRA values will be expressed as percentage, showing the relative probability of an 12 intervention to be among the best options. 13 14 15 Dealing with missing outcome data and missing statistics 16 For continuous outcomesFor we peer will extract data review for all randomized onlypatients if possible, and we will give 17 preference to data based on mixed-effect models of repeated measurements of multiple imputations 18 19 over last-observation-carried-forward data. 20 We will use published standard deviations (SDs), where available. When standard errors instead of 21 SDs are presented, the former will be converted to SDs (46). If both are missing we will estimate SDs 22 from p-values or confidence intervals, as described in Section 7.7.3 of the Cochrane Handbook for 23 Systematic Reviews (47). If none of these options is viable we will contact the original authors. When 24 25 no information can be obtained we will derive SDs from those of the other studies using a validated 26 imputation technique (46). 27 For dichotomous outcomes, everyone allocated to the intervention will be counted whether they 28 completed the follow up or not. If the authors applied such a strategy, we will use their results. If the 29 original authors presented only the results of the per-protocol or completer population, we will 30 31 assume that those participants lost to follow-up would not have changed in a given outcome. In 32 terms of efficacy this means that they would be conservatively considered to have not responded to http://bmjopen.bmj.com/ 33 treatment or control. In terms of tolerability it would mean that participants would not have 34 developed a side-effect. 35 36 37 Risk of bias assessment 38 39 Risk of bias will be assessed for each included study using the Cochrane Collaboration ‘risk of bias’ 40 tool (47, 48). The following domains will be considered: on September 24, 2021 by guest. Protected copyright. 41 1. Sequence generation: was the allocation sequence adequately generated? 42 2. Allocation concealment: was allocation adequately concealed? 43 3. Blinding of participants: was knowledge of the allocated treatment adequately prevented during 44 45 the study? Given the peculiarity of the included studies, in which the therapist cannot be blind, we 46 will consider under this item only if a way was found to keep patients unaware of the treatment they 47 were receiving (even if we expect this will not be likely); 48 4. Blinding of outcome assessors: were outcome evaluated by blind raters? Were adequate measures 49 taken to prevent them to discover treatment allocation during the study? 50 51 5. Incomplete outcome data: were incomplete outcome data adequately addressed? 52 6. Selective reporting: are reports of the study free from suggestion of selective outcome reporting? 53 7. Researcher’s allegiance: do the researchers involved have a vested interest for the psychological 54 treatment under investigation? We will additionally consider this point as possible source of bias, 55 since it has been claimed to be relevant in trials on psychological interventions (49)(50, 51). An 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 evaluation of high risk of bias will be given, for example, when the authors are founders of the 4 therapy or have written a manual for that therapy. 5 6 A description of what was reported about the same domains in each study will be provided, and a 7 8 judgement on the risk of bias will be made for each one of them, based on the following three 9 categories: ‘high risk of bias’, ‘low risk of bias’ and ‘unclear risk of bias’ where information are not 10 sufficient to make a judgement. Two independent review authors will assess the risk of bias in 11 selected studies. Any disagreement will be resolved through discussion. Where necessary, the 12 authors of the studies will be contacted for further information. Studies will be classified as having 13 14 low risk of bias if none of the domains above was rated as high risk of bias and three or less were 15 rated as unclear risk; moderate if one was rated as high risk of bias or none was rated as high risk of 16 bias but four or moreFor were peerrated as unclear review risk, and all other casesonly will be assumed to pertain to 17 high risk of bias (52). We will not include studies in the data analyses whose sequence generation 18 was at high risk of bias (e.g. randomization by the date of birth or day of the week). Effects of high 19 20 risk of bias in the other domains will be analyzed by sensitivity analyses. 21 22 Data analysis 23 24 Characteristics of the included studies 25 We will produce descriptive statistics and study population characteristics across all eligible trials, 26 describing the types of comparisons and other clinical or methodological variables, such as age, 27 duration of illness, co-medication, country, duration of study and number of session. 28 29 30 Two-step procedure 31 In a first step we will perform series of conventional pair-wise meta-analyses by combining studies 32 that compared the same interventions, including the comparison between active treatments and the http://bmjopen.bmj.com/ 33 different control arms. In subgroups with very few RCTs available or if the requirements of network 34 35 meta-analysis are not met it can be that network meta-analysis will not be appropriate and in this 36 case, conventional pairwise meta-analysis will be the most straightforward approach. As 37 heterogeneity is likely, a random effects model will be used. In a second step we will then perform a 38 network meta-analysis within a frequentist framework. 39 40 on September 24, 2021 by guest. Protected copyright. 41 Assessment of heterogeneity 42 The heterogeneity (variability in relative treatment effects within the same treatment comparison) 43 44 will be measured with the tau-squared (the variance of the random effects distribution). The 45 heterogeneity variance will be assumed common across the various treatment comparisons (grouped 46 by comparison type) and the empirical distributions will be used to characterize the amount of 47 heterogeneity as low, moderate or high using the first and third quantiles (53–55). Potential reasons 48 for heterogeneity will be explored by subgroup analysis (see below). 49 50 51 Assessment of the transitivity assumption 52 Joint analysis of treatments can be misleading if the network is substantially intransitive. We assume 53 that patients who fulfill the inclusion criteria outlined in criteria for considering studies for this 54 55 review section are equally likely to be randomised to any of the interventions that we plan to 56 compare. We will need to investigate the distribution of clinical and methodological variables that 57 can act as effect modifiers across treatment comparisons (56). We have maximized the chances of 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 transitivity in our network with regard to clinical variables by limiting our samples to participants 4 with schizophrenia and excluding specific subgroups like first episode patients or patients with 5 prevalent negative symptoms. Other clinical or methodological variables that may influence the 6 efficacy of psychological interventions include administration mode and frequency of the treatment 7 8 (like number of sessions and experience of the therapist), baseline severity (see below, “Investigation 9 of heterogeneity and inconsistency”), and blinding, which will also be assessed in sensitivity analyses. 10 We will investigate if these variables are similarly distributed across studies grouped by comparison. 11 The comparability of studies comparing the intervention with treatment as usual or waiting list 12 conditions with those that provide head-to-head evidence will be examined carefully. 13 14 15 Network meta-analysis 16 Network meta-analysisFor combines peer direct and review indirect evidence foronly all relative treatment effects and 17 can therefore provide estimates with maximum power and increased precision (57). If the collected 18 19 studies appear to be sufficiently similar with respect to the distribution of effect modifiers (refer 20 Assessment of transitivity assumption section), we will conduct a random-effects NMA to synthesize 21 all evidence for each outcome, and obtain a comprehensive ranking of all treatments. We will 22 assume a single heterogeneity parameter for each network. We will present the summary SMDs or 23 RRs for all pairwise comparisons in a league table. We will also estimate the prediction intervals to 24 25 assess how much the common heterogeneity affects the relative effect with respect to the extra 26 uncertainty anticipated in a future study. To rank the various treatments for each outcome, we will 27 use the surface under the cumulative ranking curve (SUCRA) and the mean ranks. 28 29 30 Assessment of inconsistency 31 The strategical and conceptual evaluation of transitivity will be supplemented with a statistical 32 evaluation of consistency, the agreement between direct and indirect evidence. We will employ local http://bmjopen.bmj.com/ 33 34 as well as global methods to evaluate consistency (58). Local methods detect ‘hot spots’ of 35 inconsistency, evidence loops that are inconsistent or comparisons for which direct and indirect 36 evidence disagree. We will employ a method that separates direct evidence from indirect evidence 37 provided by the entire network and then evaluate the agreement of these two pieces of evidence 38 (59). We will also evaluate consistency in the entire network by calculating the design-by-treatment 39 40 interaction test and I-squared for network heterogeneity, inconsistency, and for both (60). Tests for on September 24, 2021 by guest. Protected copyright. 41 inconsistency are known to have low power, and empirical evidence has suggested that 10% of 42 evidence loops published in the medical literature are expected to be inconsistent (61). Therefore, 43 interpretation of the statistical inference about inconsistency will be carried out with caution and 44 possible sources of inconsistency will be explored even in the absence of evidence for inconsistency. 45 46 47 Investigation of heterogeneity and inconsistency 48 We expect small amounts of heterogeneity and inconsistency to be present given the variety of study 49 settings we plan to include. The following potential effect modifiers of the primary outcome will be 50 51 explored by subgroup analyses: 52 a) Number of sessions 53 b) Study duration 54 c) Setting: individual vs group 55 d) Expertise of the therapist 56 57 e) Baseline severity (PANSS or BPRS score at baseline) 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 f) Different types of patients, with a different clinical outline concerning symptoms (if identified). 4 5 Sensitivity analyses 6 We will explore the following sensitivity analyses by excluding: 7 8 a) studies in which the outcome assessor was not blind (open studies) 9 b) studies that presented only completer analyses 10 c) studies characterized as pertaining to high risk of bias 11 d) studies with high risk of bias in researchers’ allegiance 12 e) studies focused on treatment resistant patients (study defined) 13 14 f) studies with a non-active comparison group. 15 16 Publication bias For peer review only 17 We will first examine funnel plots of pairwise MAs if there are 10 or more studies included. We will 18 19 also explore the association between study size and effect size with a comparison-adjusted funnel 20 plot that has been adapted to network meta-analysis (62). 21 22 Evaluating the quality of the evidence 23 24 The quality of evidence contributing to each network estimate will be evaluated using an adaptation 25 of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework 26 specifically developed for network meta-analysis (58). We will characterize the credibility of a body 27 of evidence based on the study limitations, imprecision, heterogeneity/inconsistency, indirectness, 28 and publication bias. 29 30 31 Statistical software 32

The analysis and presentation of results will be performed using the Stata packages network and http://bmjopen.bmj.com/ 33 network_graphs, the R package netmeta. 34 35 36 ETHICS AND DISSEMINATION 37 38 This review does not require ethical approval. Findings will be published in peer reviewed scientific 39 journals, granting open access, and the database will be made publicly available (in agreement with 40 European Research Council Guidelines on Implementation of Open Access to Scientific Publications on September 24, 2021 by guest. Protected copyright. 41 and Research Data (http://ec.europa.eu/research/participants/data/ref/h2020/other/hi/oa- 42 pilot/h2020-hi-erc-oa-guide_en.pdf). 43 44 45 46 REFERENCES 47 1. Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M et al. Years lived with disability 48 (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the Global 49 50 Burden of Disease Study 2010. Lancet 2012; 380(9859):2163–96. 51 2. Leucht S, Cipriani A, Spineli L, Mavridis D, Orey D, Richter F et al. Comparative efficacy and 52 53 tolerability of 15 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet 54 2013; 382(9896):951–62. 55 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 3. National Collaborating Centre for Mental Health. Core interventions in the treatment and 4 management of schizophrenia in adults in primary and secondary care (Clinical Guideline CG82). 5 London; 2009. 6 7 4. Buchanan RW, Kreyenbuhl J, Kelly DL, Noel JM, Boggs DL, Fischer BA et al. The 2009 schizophrenia 8 PORT psychopharmacological treatment recommendations and summary statements. Schizophr Bull 9 2010; 36(1):71–93. 10 11 5. Pinquart M, Oslejsek B, Teubert D. Efficacy of systemic therapy on adults with mental disorders: A 12 meta-analysis. Psychother Res 2016; 26(2):241–57. 13 14 6. Ruddy R, Milnes D. Art therapy for schizophrenia or schizophrenia-like illnesses. Cochrane 15 Database Syst Rev 2005; (4):CD003728. 16 For peer review only 17 7. Jones C, Hacker D, Cormac I, Meaden A, Irving CB. Cognitive behaviour therapy versus other 18 psychosocial treatments for schizophrenia. Cochrane Database Syst Rev 2012; (4):CD008712. 19 20 8. Lutgens D, Gariepy G, Malla A. Psychological and psychosocial interventions for negative 21 symptoms in psychosis: Systematic review and meta-analysis. Br J Psychiatry 2017; 210(5):324–32. 22 9. Zimmermann G, Favrod J, Trieu VH, Pomini V. The effect of cognitive behavioral treatment on the 23 24 positive symptoms of schizophrenia spectrum disorders: A meta-analysis. Schizophr Res 2005; 25 77(1):1–9. 26 27 10. Turner DT, van der Gaag M, Karyotaki E, Cuijpers P. Psychological interventions for psychosis: a 28 meta-analysis of comparative outcome studies. Am J Psychiatry 2014; 171(5):523–38. 29 11. Naeem F, Asmer MS, Khoury B, Kingdon D, Farooq S. Acceptance and Commitment Therapy for 30 31 schizophrenia and related disorders (Protocol). Cochrane Database of Systematic Reviews 2015; (6). 32 12. Gray R, Bressington D, Ivanecka A, Hardy S, Jones M, Schulz M et al. Is adherence therapy an http://bmjopen.bmj.com/ 33 34 effective adjunct treatment for patients with schizophrenia spectrum disorders? A systematic review 35 and meta-analysis. BMC Psychiatry 2016; 16:90. 36 13. Marshall M, Lockwood A. Assertive community treatment for people with severe mental 37 38 disorders. Cochrane Database Syst Rev 2000; (2):CD001089. 39 14. Dieterich M, Irving CB, Bergman H, Khokhar MA, Park B, Marshall M. Intensive case management 40 for severe mental illness. Cochrane Database Syst Rev 2017; 1:CD007906. on September 24, 2021 by guest. Protected copyright. 41 42 15. Jauhar S, McKenna PJ, Radua J, Fung E, Salvador R, Laws KR. Cognitive-behavioural therapy for 43 the symptoms of schizophrenia: systematic review and meta-analysis with examination of potential 44 45 bias. Br J Psychiatry 2014; 204(1):20–9. 46 16. van der Gaag M, Valmaggia LR, Smit F. The effects of individually tailored formulation-based 47 cognitive behavioural therapy in auditory hallucinations and delusions: A meta-analysis. Schizophr 48 49 Res 2014; 156(1):30–7. 50 17. Hazell CM, Hayward M, Cavanagh K, Strauss C. A systematic review and meta-analysis of low 51 52 intensity CBT for psychosis. Clin Psychol Rev 2016; 45:183–92. 53 18. Kennedy L, Xyrichis A. Cognitive Behavioral Therapy Compared with Non-specialized Therapy for 54 Alleviating the Effect of Auditory Hallucinations in People with Reoccurring Schizophrenia: A 55 56 Systematic Review and Meta-analysis. Community Ment Health J 2017; 53(2):127–33. 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 19. Cella M, Preti A, Edwards C, Dow T, Wykes T. Cognitive remediation for negative symptoms of 4 schizophrenia: A network meta-analysis. Clin Psychol Rev 2017; 52:43–51. 5 6 20. Ren J, Xia J. Dance therapy for schizophrenia. Cochrane Database Syst Rev 2013; (10):CD006868. 7 21. Pitschel-Walz G, Leucht S, Bauml J, Kissling W, Engel RR. The effect of family interventions on 8 relapse and rehospitalization in schizophrenia--a meta-analysis. Schizophr Bull 2001; 27(1):73–92. 9 10 22. Pharoah F, Mari J, Rathbone J, Wong W. Family intervention for schizophrenia. Cochrane 11 Database Syst Rev 2010; (12):CD000088. 12 13 23. Orfanos S, Banks C, Priebe S. Are Group Psychotherapeutic Treatments Effective for Patients with 14 Schizophrenia? A Systematic Review and Meta-Analysis. Psychother Psychosom 2015; 84(4):241–9. 15 16 24. Roder V, MuellerFor DR, Mueser peer KT, Brenner review HD. Integrated psychological only therapy (IPT) for 17 schizophrenia: Is it effective? Schizophr Bull 2006; 32 Suppl 1:S81-93. 18 19 25. Eichner C, Berna F. Acceptance and Efficacy of Metacognitive Training (MCT) on Positive 20 Symptoms and Delusions in Patients With Schizophrenia: A Meta-analysis Taking Into Account 21 Important Moderators. Schizophr Bull 2016; 42(4):952–62. 22 23 26. Aust J, Bradshaw T. Mindfulness interventions for psychosis: A systematic review of the literature. 24 J Psychiatr Ment Health Nurs 2017; 24(1):69–83. 25 26 27. Geretsegger M, Mössler KA, Bieleninik Ł, Chen X-J, Heldal TO, Gold C. Music therapy for people 27 with schizophrenia and schizophrenia-like disorders. Cochrane Database Syst Rev 2017; 5:CD004025. 28 29 28. Malmberg L, Fenton M. Individual psychodynamic psychotherapy and psychoanalysis for 30 schizophrenia and severe mental illness. Cochrane Database Syst Rev 2001; (3):CD001360. 31 32 29. Pekkala E, Merinder L. Psychoeducation for schizophrenia. Cochrane Database Syst Rev 2000; http://bmjopen.bmj.com/ 33 (4):CD002831. 34 30. Almerie MQ, Okba Al Marhi M, Jawoosh M, Alsabbagh M, Matar HE, Maayan N et al. Social skills 35 36 programmes for schizophrenia. Cochrane Database Syst Rev 2015; (6):CD009006. 37 31. Buckley LA, Maayan N, Soares-Weiser K, Adams CE. Supportive therapy for schizophrenia. 38 39 Cochrane Database Syst Rev 2015; (4):CD004716. 40 32. Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C et al. The PRISMA on September 24, 2021 by guest. Protected copyright. 41 extension statement for reporting of systematic reviews incorporating network meta-analyses of 42 43 health care interventions: Checklist and explanations. Ann Intern Med 2015; 162(11):777–84. 44 33. Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M et al. Preferred reporting items 45 46 for systematic review and meta-analysis protocols (PRISMA-P) 2015: Elaboration and explanation. 47 BMJ 2015; 349:g7647. 48 34. Egger M, Zellweger-Zahner T, Schneider M, Junker C, Lengeler C, Antes G. Language bias in 49 50 randomised controlled trials published in English and German. Lancet 1997; 350(9074):326–9. 51 35. Wu T, Li Y, Liu G, Bian Z, Li J, Zhang J, Xie L, Ni J, editor. Investigation of authenticity of 'claimed' 52 53 randomized controlled trials (RCTs) and quality assessment of RCT reports published in China; 2006. 54 36. Woodhead M. 80% of China's clinical trial data are fraudulent, investigation finds. BMJ 2016; 55 355:i5396. 56 57 37. Carpenter WT, JR, Buchanan RW. Schizophrenia. N Engl J Med 1994; 330(10):681–90. 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 38. Adams,C.E., Coutinho,E., Davis,J.M., Duggan,L., Essali,A., Fenton,M., Li,C., Jayaram,M., Leucht,S., 4 Tharyan,P., Välimäki,M. Cochrane Schizophrenia Group. The Cochrane Library. Chichester, UK: John 5 Wiley & Sons Ltd; 2011. 6 7 39. Leucht S, Leucht C, Huhn M, Chaimani A, Mavridis D, Helfer B et al. Sixty Years of Placebo- 8 Controlled Antipsychotic Drug Trials in Acute Schizophrenia: Systematic Review, Bayesian Meta- 9 Analysis, and Meta-Regression of Efficacy Predictors. Am J Psychiatry 2017; 174(10):927–42. 10 11 40. Zhu Y, Li C, Huhn M, Rothe P, Krause M, Bighelli I et al. How well do patients with a first episode 12 of schizophrenia respond to antipsychotics: A systematic review and meta-analysis. Eur 13 Neuropsychopharmacol 2017; 27(9):835–44. 14 15 41. Furukawa TA, Noma H, Caldwell DM, Honyashiki M, Shinohara K, Imai H et al. Waiting list may be 16 a nocebo conditionFor in psychotherapy peer trials: areview contribution from networkonly meta-analysis. Acta 17 Psychiatr Scand 2014; 130(3):181–92. 18 19 42. Marshall M, Lockwood A, Bradley C, Adams C, Joy C, Fenton M. Unpublished rating scales: a 20 major source of bias in randomised controlled trials of treatments for schizophrenia. Br J Psychiatry 21 2000; 176:249–52. 22 23 43. Vaughan B, Goldstein MH, Alikakos M, Cohen LJ, Serby MJ. Frequency of reporting of adverse 24 events in randomized controlled trials of psychotherapy vs. psychopharmacotherapy. Compr 25 26 Psychiatry 2014; 55(4):849–55. 27 44. Linden M, Schermuly-Haupt M-L. Definition, assessment and rate of psychotherapy side effects. 28 World Psychiatry 2014; 13(3):306–9. 29 30 45. Salanti G, Ades AE, Ioannidis JPA. Graphical methods and numerical summaries for presenting 31 results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol 2011; 32 http://bmjopen.bmj.com/ 33 64(2):163–71. 34 46. Furukawa TA, Barbui C, Cipriani A, Brambilla P, Watanabe N. Imputing missing standard 35 deviations in meta-analyses can provide accurate results. J Clin Epidemiol 2006; 59(1):7–10. 36 37 47. Higgins JPT, editor. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 38 [updated March 2011]; 2011. 39 40 48. Higgins JPT, Altman DG, Sterne JAC (editors), editor. Assessing risk of bias in included studies. In: on September 24, 2021 by guest. Protected copyright. 41 Higgins JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic 42 Reviews of Interventions version 5.2.0 (updated June 2017); Cochrane, 2017. Available from: URL: 43 44 http://training.cochrane.org/handbook. 45 49. Dragioti E, Dimoliatis I, Evangelou E. Disclosure of researcher allegiance in meta-analyses and 46 47 randomised controlled trials of psychotherapy: a systematic appraisal. BMJ Open 2015; 48 5(6):e007206. 49 50. Lieb K, Osten-Sacken J von der, Stoffers-Winterling J, Reiss N, Barth J. Conflicts of interest and 50 51 spin in reviews of psychological therapies: a systematic review. BMJ Open 2016; 6(4):e010606. 52 51. Munder T, Brutsch O, Leonhart R, Gerger H, Barth J. Researcher allegiance in psychotherapy 53 54 outcome research: an overview of reviews. Clin Psychol Rev 2013; 33(4):501–11. 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 23 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 52. Furukawa TA, Salanti G, Atkinson LZ, Leucht S, Ruhe HG, Turner EH et al. Comparative efficacy 4 and acceptability of first-generation and second-generation antidepressants in the acute treatment 5 of major depression: Protocol for a network meta-analysis. BMJ Open 2016; 6(7):e010919. 6 7 53. Turner RM, Davey J, Clarke MJ, Thompson SG, Higgins JP. Predicting the extent of heterogeneity 8 in meta-analysis, using empirical data from the Cochrane Database of Systematic Reviews. Int J 9 Epidemiol 2012; 41(3):818–27. 10 11 54. Rhodes KM, Turner RM, Higgins JPT. Empirical evidence about inconsistency among studies in a 12 pair-wise meta-analysis. Res Synth Methods 2016; 7(4):346–70. 13 14 55. Rhodes KM, Turner RM, White IR, Jackson D, Spiegelhalter DJ, Higgins JPT. Implementing 15 informative priors for heterogeneity in meta-analysis using meta-regression and pseudo data. Stat 16 Med 2016; 35(29):5495–511.For peer review only 17 18 56. Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta- 19 analysis: Many names, many benefits, many concerns for the next generation evidence synthesis 20 tool. Res Synth Methods 2012; 3(2):80–97. 21 22 57. Salanti G, Higgins JPT, Ades AE, Ioannidis JPA. Evaluation of networks of randomized trials. Stat 23 Methods Med Res 2008; 17(3):279–301. 24 25 58. Salanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JPT. Evaluating the quality of evidence 26 from a network meta-analysis. PLoS One 2014; 9(7):e99682. 27 28 59. Dias S, Welton NJ, Caldwell DM, Ades AE. Checking consistency in mixed treatment comparison 29 meta-analysis. Stat Med 2010; 29(7-8):932–44. 30 31 60. Higgins JPT, Jackson D, Barrett JK, Lu G, Ades AE, White IR. Consistency and inconsistency in 32 network meta-analysis: Concepts and models for multi-arm studies. Res Synth Methods 2012; http://bmjopen.bmj.com/ 33 3(2):98–110. 34 35 61. Veroniki AA, Vasiliadis HS, Higgins JPT, Salanti G. Evaluation of inconsistency in networks of 36 interventions. Int J Epidemiol 2013; 42(1):332–45. 37 38 62. Chaimani A, Salanti G. Using network meta-analysis to evaluate the existence of small-study 39 effects in a network of interventions. Res Synth Methods 2012; 3(2):161–76. 40 on September 24, 2021 by guest. Protected copyright. 41 42 Authors’ contributions: IB and SL designed this study, drafted and critically revised the 43 protocol. IB will screen search results for inclusion, conduct data extraction and data analysis, and 44 draft the final manuscript. SL will assist with data extraction and analysis, and revise the final 45 46 manuscript. CR and SW will screen search results for inclusion and conduct data extraction. GS 47 provided substantial methodological advice in planning the study, and will assist with data analysis. 48 CB and TF contributed with clinical and methodological input in planning the study. All authors 49 contributed to and have approved the final manuscript. 50

51 52 Collaborators: Samantha Roberts helped us to conduct the literature searches. Maximilian Huhn, 53 Johannes Schneider-Thoma, Marc Krause and Costanza Carmi provided help and suggestions. 54

55 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 23 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from

1 2 3 Funding: This project has received funding from the European Union’s Horizon 2020 Research and 4 Innovation Programme under the Marie Skłodowska-Curie grant agreement N° 701717. This work 5 was also supported by the German Research Foundation (DFG) and the Technical University of 6 7 Munich within the Open Access Publishing Funding Programme. The funder had no role in developing 8 the protocol. 9 10 SL in the past 3 years has received honoraria for consulting from Roche, 11 Competing interests: 12 TEVA, Otsuka, Lundbeck, and LB Pharma; for lectures from Otsuka, Lundbeck, Janssen, ICON, Lilly, 13 Sanofi Aventis, AOP Orphan, Roche, and Servier; and for a publication from Roche. TAF has received 14 lecture fees from Eli Lilly, Janssen, Meiji, Mitsubishi-Tanabe, MSD and Pfizer and consultancy fees 15 from Takeda Science Foundation. He has received royalties from Igaku-Shoin and Nihon Bunka 16 For peer review only 17 Kagaku-sha publishers. He has received research support from Mochida and Mitsubishi-Tanabe. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 40 on September 24, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 23 BMJ Open

1 2 3

Search terms for Pubmed BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from 4 5 6 ("Schizophrenia"[Mesh] OR "Paranoid Disorders"[Mesh] OR schizo*[Title/Abstract] OR 7 psychotic*[Title/Abstract] OR psychosis[Title/Abstract] OR psychoses[Title/Abstract]) AND 8 9 ("Psychotherapy"[Mesh] or "Behavior Therapy"[Mesh] or "Cognitive Therapy"[Mesh] or 10 "Complementary Therapies"[Mesh] or "Psychoanalysis"[Mesh] or "Counseling"[Mesh] or 11 "Hypnosis"[Mesh] or "Association"[Mesh] or "Association Learning"[Mesh] OR 12 abreaction[Title/Abstract] OR "acceptance[Title/Abstract] AND commitment therapy"[Title/Abstract] 13 14 OR "acting out"[Title/Abstract] OR adlerian[Title/Abstract] OR "analytical 15 psychotherapy"[Title/Abstract] OR "analytical psychotherapies"[Title/Abstract] OR "anger 16 control"[Title/Abstract] OR "anger management"[Title/Abstract] OR "animal therapy"[Title/Abstract] 17 18 OR "animal therapies"[Title/Abstract]For peer review OR "art therapy"[Title/Abstract]only OR "art 19 therapies"[Title/Abstract] OR "assertive training"[Title/Abstract] OR "assertiveness 20 training"[Title/Abstract] OR "attention training technique"[Title/Abstract] OR "autogenic 21 22 training"[Title/Abstract] OR autosuggestion[Title/Abstract] OR "aversion therapy"[Title/Abstract] OR 23 "aversion therapies"[Title/Abstract] OR "balint group"[Title/Abstract] OR befriending[Title/Abstract] 24 OR "behavior contracting"[Title/Abstract] OR "behavior modification"[Title/Abstract] OR "behavior 25 regulation"[Title/Abstract] OR "behavior therapy"[Title/Abstract] OR "behavior 26 27 therapies"[Title/Abstract] OR "behaviour contracting"[Title/Abstract] OR "behaviour 28 modification"[Title/Abstract] OR "behaviour regulation"[Title/Abstract] OR "behaviour 29 therapy"[Title/Abstract] OR "behaviour therapies"[Title/Abstract] OR bibliotherapy[Title/Abstract] 30 31 OR bibliotherapies[Title/Abstract] OR biofeedback[Title/Abstract] OR "body 32 psychotherapy"[Title/Abstract] OR "body psychotherapies"[Title/Abstract] OR "brief 33 psychotherapy"[Title/Abstract] OR "brief psychotherapies"[Title/Abstract] OR "caregiver 34 support"[Title/Abstract] OR cbt[Title/Abstract] OR "client centre"[Title/Abstract] OR "client 35 36 center"[Title/Abstract] OR "cognitive behavior"[Title/Abstract] OR "cognitive 37 behaviorial"[Title/Abstract] OR "cognitive intervention"[Title/Abstract] OR "cognitive http://bmjopen.bmj.com/ 38 interventions"[Title/Abstract] OR "cognitive rehabilitation"[Title/Abstract] OR "cognitive 39 40 remediation"[Title/Abstract] OR "cognitive technique"[Title/Abstract] OR "cognitive 41 techniques"[Title/Abstract] OR "cognitive therapy"[Title/Abstract] OR "cognitive 42 therapies"[Title/Abstract] OR "cognitive treatment"[Title/Abstract] OR "cognitive 43 treatments"[Title/Abstract] OR "color therapy"[Title/Abstract] OR "color therapies"[Title/Abstract] 44 45 OR "colour therapy"[Title/Abstract] OR "colour therapies"[Title/Abstract] OR "compassionate mind on September 24, 2021 by guest. Protected copyright. 46 training"[Title/Abstract] OR "conjoint therapy"[Title/Abstract] OR "conjoint therapies"[Title/Abstract] 47 OR "contingency management"[Title/Abstract] OR "conversational therapy"[Title/Abstract] OR 48 49 "conversational therapies"[Title/Abstract] OR "conversion therapy"[Title/Abstract] OR "conversion 50 therapies"[Title/Abstract] OR "coping skills"[Title/Abstract] OR counseling[Title/Abstract] OR 51 counselling[Title/Abstract] OR countertransference[Title/Abstract] OR "couples 52 therapy"[Title/Abstract] OR "couples therapies"[Title/Abstract] OR "covert 53 54 sensitization"[Title/Abstract] OR "covert sensitisation"[Title/Abstract] OR "crisis 55 intervention"[Title/Abstract] OR "dance therapy"[Title/Abstract] OR "dance 56 therapies"[Title/Abstract] OR dialectic[Title/Abstract] OR dialectical[Title/Abstract] OR "dream 57 58 analysis"[Title/Abstract] OR eclectic[Title/Abstract] OR "emotion focused"[Title/Abstract] OR 59 "emotionally focused"[Title/Abstract] OR "emotional freedom technique"[Title/Abstract] OR 60 "encounter group therapy"[Title/Abstract] OR "encounter group therapies"[Title/Abstract] OR "existential therapy"[Title/Abstract] OR "existential therapies"[Title/Abstract] OR "experiential

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 23

1 2 3 psychotherapy"[Title/Abstract] OR "experiential psychotherapies"[Title/Abstract] OR "exposure 4 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from therapy"[Title/Abstract] OR "exposure therapies"[Title/Abstract] OR "expressive 5 6 psychotherapy"[Title/Abstract] OR "expressive psychotherapies"[Title/Abstract] OR "eye movement 7 desensitization"[Title/Abstract] OR "eye movement desensitisation"[Title/Abstract] OR "family 8 intervention"[Title/Abstract] OR "family interventions"[Title/Abstract] OR "family 9 10 therapy"[Title/Abstract] OR "family therapies"[Title/Abstract] OR "feminist therapy"[Title/Abstract] 11 OR "feminist therapies"[Title/Abstract] OR "free association"[Title/Abstract] OR 12 freudian[Title/Abstract] OR "geriatric psychotherapy"[Title/Abstract] OR "geriatric 13 psychotherapies"[Title/Abstract] OR "gestalt therapy"[Title/Abstract] OR "gestalt 14 15 therapies"[Title/Abstract] OR griefwork[Title/Abstract] OR "group intervention"[Title/Abstract] OR 16 "group interventions"[Title/Abstract] OR "group psychotherapy"[Title/Abstract] OR "group 17 psychotherapies"[Title/Abstract] OR "group therapy"[Title/Abstract] OR "group 18 For peer review only 19 therapies"[Title/Abstract] OR "guided imagery"[Title/Abstract] OR "holistic 20 psychotherapy"[Title/Abstract] OR "holistic psychotherapies"[Title/Abstract] OR "humanistic 21 psychotherapy"[Title/Abstract] OR "humanistic psychotherapies"[Title/Abstract] OR 22 hypnosis[Title/Abstract] OR hypnotherapy[Title/Abstract] OR hypnotherapies[Title/Abstract] OR 23 24 hypnotizability[Title/Abstract] OR hypnotisability[Title/Abstract] OR imagery[Title/Abstract] OR 25 "implosive therapy"[Title/Abstract] OR "implosive therapies"[Title/Abstract] OR "individual 26 psychotherapy"[Title/Abstract] OR "individual psychotherapies"[Title/Abstract] OR "insight 27 28 therapy"[Title/Abstract] OR "insight therapies"[Title/Abstract] OR "integrated psychological 29 therapy"[Title/Abstract] OR "integrative psychotherapy"[Title/Abstract] OR "integrative 30 psychotherapies"[Title/Abstract] OR "integrative therapy"[Title/Abstract] OR "integrative 31 therapies"[Title/Abstract] OR interpersonal[Title/Abstract] OR jungian[Title/Abstract] OR 32 33 kleinian[Title/Abstract] OR logotherapy[Title/Abstract] OR logotherapies[Title/Abstract] OR 34 "marathon group therapy"[Title/Abstract] OR "marathon group therapies"[Title/Abstract] OR 35 "marital therapy"[Title/Abstract] OR "marital therapies"[Title/Abstract] OR meditation[Title/Abstract] 36 37 OR "mental healing"[Title/Abstract] OR "metacognitive therapy"[Title/Abstract] OR "metacognitive http://bmjopen.bmj.com/ 38 therapies"[Title/Abstract] OR "metacognitive training"[Title/Abstract] OR "milieu 39 therapy"[Title/Abstract] OR "milieu therapies"[Title/Abstract] OR mindfulness[Title/Abstract] OR 40 "morita therapy"[Title/Abstract] OR "morita therapies"[Title/Abstract] OR multimodal[Title/Abstract] 41 42 OR "music therapy"[Title/Abstract] OR "music therapies"[Title/Abstract] OR "narrative 43 therapy"[Title/Abstract] OR "narrative therapies"[Title/Abstract] OR "nondirective 44

therapy"[Title/Abstract] OR "nondirective therapies"[Title/Abstract] OR "object on September 24, 2021 by guest. Protected copyright. 45 46 relations"[Title/Abstract] OR "person centred therapy"[Title/Abstract] OR "person centred 47 therapies"[Title/Abstract] OR "person centered therapy"[Title/Abstract] OR "person centered 48 therapies"[Title/Abstract] OR "personal construct therapy"[Title/Abstract] OR "personal construct 49 therapies"[Title/Abstract] OR "persuasion therapy"[Title/Abstract] OR "persuasion 50 51 therapies"[Title/Abstract] OR "pet therapy"[Title/Abstract] OR "pet therapies"[Title/Abstract] OR 52 "play therapy"[Title/Abstract] OR "play therapies"[Title/Abstract] OR "primal therapy"[Title/Abstract] 53 OR "primal therapies"[Title/Abstract] OR "problem solving"[Title/Abstract] OR 54 55 psychoanalyse[Title/Abstract] OR psychoanalysed[Title/Abstract] OR psychoanalysis[Title/Abstract] 56 OR psychoanalytic[Title/Abstract] OR psychodrama[Title/Abstract] OR psychodynamic[Title/Abstract] 57 OR psychoeducate[Title/Abstract] OR psychoeducation[Title/Abstract] OR 58 psychoeducating[Title/Abstract] OR psychologic[Title/Abstract] OR psychological[Title/Abstract] OR 59 60 psychologically[Title/Abstract] OR "psychological therapy"[Title/Abstract] OR "psychological therapies"[Title/Abstract] OR "psychosocial treatment"[Title/Abstract] OR "psychosocial

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 23 BMJ Open

1 2 3 treatments"[Title/Abstract] OR psychotherapy[Title/Abstract] OR psychotherapies[Title/Abstract] OR 4 BMJ Open: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from "psychotherapeutic counsel"[Title/Abstract] OR "psychotherapeutic counseling"[Title/Abstract] OR 5 6 "psychotherapeutic counselling"[Title/Abstract] OR "psychotherapeutic processes"[Title/Abstract] OR 7 "psychotherapeutic training"[Title/Abstract] OR "psychotherapeutic treatment"[Title/Abstract] OR 8 "psychotherapeutic treatments"[Title/Abstract] OR "rational emotive"[Title/Abstract] OR "reality 9 10 therapy"[Title/Abstract] OR "reality therapies"[Title/Abstract] OR "reciprocal 11 inhibition"[Title/Abstract] OR rehabilitation[Title/Abstract] OR rehabilitating[Title/Abstract] OR 12 "relationship therapy"[Title/Abstract] OR "relationship therapies"[Title/Abstract] OR 13 relaxation[Title/Abstract] OR "reminiscence therapy"[Title/Abstract] OR "reminiscence 14 15 therapies"[Title/Abstract] OR rogerian[Title/Abstract] OR "role play"[Title/Abstract] OR "role 16 plays"[Title/Abstract] OR "role playing"[Title/Abstract] OR "self analysis"[Title/Abstract] OR "self 17 analysing"[Title/Abstract] OR "self esteem"[Title/Abstract] OR "sensitivity training"[Title/Abstract] OR 18 For peer review only 19 "sex therapy"[Title/Abstract] OR "sex therapies"[Title/Abstract] OR "sleep phase 20 chronotherapy"[Title/Abstract] OR "sleep phase chronotherapies"[Title/Abstract] OR "social skills 21 education"[Title/Abstract] OR "social skills training"[Title/Abstract] OR "socioenvironmental 22 therapy"[Title/Abstract] OR "socioenvironmental therapies"[Title/Abstract] OR 23 24 sociotherapy[Title/Abstract] OR sociotherapies[Title/Abstract] OR "solution focused"[Title/Abstract] 25 OR "stress management"[Title/Abstract] OR "support group"[Title/Abstract] OR "support 26 groups"[Title/Abstract] OR "supportive therapy"[Title/Abstract] OR "supportive 27 28 therapies"[Title/Abstract] OR "systematic desensitization"[Title/Abstract] OR "systematic 29 desensitisation"[Title/Abstract] OR "systemic therapy"[Title/Abstract] OR "systemic 30 therapies"[Title/Abstract] OR "therapeutic community"[Title/Abstract] OR "therapeutic 31 communities"[Title/Abstract] OR "transactional analysis"[Title/Abstract] OR 32 33 transference[Title/Abstract] OR transtheoretical[Title/Abstract] OR "validation 34 therapy"[Title/Abstract] OR "validation therapies"[Title/Abstract]) AND (((randomized controlled 35 trial[pt]) OR (controlled clinical trial[pt]) OR (randomized[tiab]) OR (placebo[tiab]) OR (drug 36 37 therapy[sh]) OR (randomly[tiab]) OR (trial[tiab]) OR (groups[tiab])) NOT (animals[mh] NOT http://bmjopen.bmj.com/ 38 humans[mh])) 39 40 41 42 43 44 45 on September 24, 2021 by guest. Protected copyright. 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: first published as 10.1136/bmjopen-2017-019280 on 14 March 2018. Downloaded from Page 22 of 23

2 8 5 3 1 5 16 16 58 58 1617 1617 1617 1617 1 (title) 1 Reported on Reported pageon Supplementary Supplementary file BMJ Open Checklist item http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected copyright.

For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml registers registers or grey literatureother sources) with planned coveragedates of Provide anexplicit Provide statement the review question(s)of address will the with referenceto participants, Provide name, Provide institutional affiliation, email address authors; of all protocol provide physical address mailing Describe all intended Describe information (suchsources as electronic contact databases, with study trial authors, Specify study characteristicsSpecify the as (such PICO, design,study setting, time frame) report and characteristics (such as years considered, as (such language, years publication to status) be as used criteria for eligibilityreview for the that it could it repeatedthat be interventions, interventions, comparators, and (PICO) outcomes of corresponding authorof If the the protocolIf anrepresents amendment previously a completed of or published protocol, and as identify such changes; list otherwise, state fordocumentingplan important amendments protocol

6 6 the rationale Describe for theincontext the review of what is already known 2 2 registered, If provide namethe the registry (such as PROSPERO)of and registration number 7 7 9 9 5c rolesfunder(s), Describe of and/orsponsor(s), institution(s), if in any, developing the protocol

Role of Role of sponsor

Sponsor 5b name Provide review for the funder and/or sponsor Sources funderor 5a Indicate financialsources of support reviewor forthe other Contact 3a Update 1b the foranprotocol If update is a systematicprevious of identifyreview, as such Identification 1a report the as Identify protocol systematicaa of review Contributions 3b contributions Describe authorsof protocol and identify the the guarantor reviewof Study records:Study Search Search strategy 10 search draft Present forto usedatbe strategy leastof electronic one database, including planned limits, such Information Information sources Eligibility criteriaEligibility 8 METHODS Objectives Rationale Rationale Support: Authors: Title: ADMINISTRATIVE INFORMATION PRISMA-P (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols) 2015 checklist: recommended items to to items recommended checklist: 2015 Protocols) Meta-Analysis review for and Systematic Items (PreferredReporting PRISMA-P reviewprotocol* a systematic in address Section topic and No Item Registration INTRODUCTION Amendments 4 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-2017-019280 on 14 March 2018. Downloaded from 9 9 9 13 13 10 10 12 12 89 89 89 89 78 78 11, 12 12, 12, 13 9,

BMJ Open http://bmjopen.bmj.com/ bias(es) (suchbias(es) as publication bias across studies, selective reporting ti ti A,Petticrew M, P, Shekelle Stewart L, PRISMA-P Preferred Group. reportingitems for systematic andreview d ind conjunctionwith Explanationthe PRISMA-P and Elaboration (citewhen available) for important on September 24, 2021 by guest. Protected copyright.

For peer review only For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml , τ)Kendall’s 2 Describe Describe planned extractingmethod of from (suchdata reports as piloting forms, independently,in done State the the State that process be (suchwill for used selecting studies independenttwo reviewers)as eachthrough Describe anticipatedDescribe methods assessing for risk of bias of studies, individual including whether this be will the at done level,or study outcome both; state this or informationhow will in used databe synthesis Specify any planned Specify assessment of meta within studies)within List and all define List forvariables which data (such be soughtwill asPICO items, funding sources), any pre List and all define List for outcomes whichdata sought, be including will prioritization of mainadditional and rationaleoutcomes, with duplicate), any duplicate), processes for and obtaining confirming data from investigators data and and combiningdata methods of data from studies, including planned explorationany of consistency(such as I phase of the phase review of is,(that screening, andeligibility inclusionin metaanalysis) planned data assumptionsandplanned simplifications 12 12 14 14 17 17 strength Describe how the the body evidenceof of willassessed be as(such GRADE) 13 13 11c 11c 11a 11a the mechanism(s) Describe that be will to used recordsmanage and thedata throughout review 15c 15c any Describe proposed additional analyses as (such or subgroupsensitivity analyses, metaregression) 11b 15b data are for If appropriate synthesis,quantitative describe planned summary methods handling measures, of 15d quantitative If synthesis is not appropriate, describe typethe of summary planned

collection Data Selection Data management process process It is It stronglyrecommendedchecklist this that be rea

clarification theclarification items. on to Amendments a review protocol should be tracked copyrightand dated. The for (includingPRISMA-P checklist) byheld the is Group PRISMA-P and is distributedunder a Creative Commons Attribution Licence 4.0.

* Confidence inConfidence evidencecumulative Data Data synthesis 15a criteria under Describe data which study be synthesisedwill quantitatively Risk Risk in of bias studiesindividual Metabias(es) 16 Outcomes and Outcomes meta-analysis protocols(PRISMA-P) elaboration2015: explanation. andBMJ. Jan 2015 2;349(jan02 1):g7647. Data itemsData From: Shamseer From: L, Moher D, Clarke M, Ghersi Libera D, prioritization 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 Page 23 of