Patel 2017.

Study characteristics
Methods	Study design: RCT Duration of study: 28 October 2013 and 30 August 2016
Participants	Country:  India  Income classification: lower‐middle income Geographical scope: Goa, India  Healthcare setting: PC facility  Mental health condition: depression  Population   1. Age: 18 to 65 years  2. Gender: both  3. Socioeconomic background: lower SEC  4. Inclusion criteria    a. Above the age of 18 but below the age of 65  b. Reside within the geographic area selected for the PHC  c. Plan to stay at the same address for at least 12 months  d. Able to speak one of the following languages: Konkani/Hindi/Marathi/English  e. Must not have been screened in the previous 3 months  f. PHQ‐9 score > 14 5. Exclusion criteria   a. Pregnant women  b. Patients with drinking problems  c. Patients who need urgent medical attention (defined as needing emergency treatment and/or in‐patient admission)  d. Patients unable to communicate clearly (e.g. due to speech or hearing disability)  e. In receipt of PREMIUM counselling treatment  f. Patient lives together in the same household with previously recruited patient or is in regular contact with previously recruited patient
Interventions	Stated purpose: to assess the effectiveness and cost‐effectiveness of a brief psychological treatment (Healthy Activity Program (HAP)) for delivery by lay counsellors to patients with moderately severe to severe depression in primary healthcare settings  INTERVENTION (n = 245) Name: Healthy Activity Program (HAP)  Delivered by: LHW Title/name of PW and number: lay counsellors ‐ 11  1. Selection: trained and met competency standards  2. Educational background: not mentioned  3. Training: an international expert in behavioural activation (SD) trained and provided ongoing supervision for 5 local specialists, who in turn provided onsite training and supervision for lay counsellors. Training of lay counsellors involved a 3‐week participatory workshop covering both HAP and CAP treatments, followed by an internship phase of 6 months, in which trainee counsellors delivered treatment to eligible patients in primary healthcare clinics, combined with peer‐led group supervision as trainees gained experience in delivery of treatment. 11 counsellors who met competency standards as assessed by standardised role‐plays and therapy quality measures participated in the trial   4. Supervision: weekly peer‐led supervision in groups of 4 to 6 that involved rating of a randomly selected 10% of recorded sessions on the HAP Therapy Quality Scale (TQS)‐26 and individual supervision twice monthly   5. Incentives/remuneration: not mentioned.  Intervention details  1. Duration/frequency: 6 to 8 individual 30– to 40‐minute sessions, with initial sessions at weekly intervals 2. Content of intervention: EUC plus HAP. HAP is a manualised psychological treatment based on behavioural activation that includes the following strategies: psychoeducation, behavioural assessment, activity monitoring, activity structuring and scheduling, activation of social networks, and problem‐solving. Additional strategies used in response to specific needs consisted of behavioural strategies to improve interpersonal communication skills and decrease rumination, advice regarding sleep problems and tobacco cessation, and relaxation training  CONTROL: enhanced usual care (n = 248) EUC comprised routine consultation with the PHC physician, enhanced by providing PHQ‐9 screening results to both the PHC physician and the patient, and providing copies of a contextualised version of the WHO Mental Health Gap Action Programme (mhGAP) guidelines to the PHC physician, which included information on when and where to refer for psychiatric care. EUC was available to all trial participants   CO‐INTERVENTIONS: nil
Outcomes	Patients    Primary outcomes 1. Depression symptom severity on Beck Depression Inventory version II 2. Remission from depression (PHQ‐9 score < 10) at 3 months in the intention‐to‐treat population  Secondary outcomes 1. Disability 2. Days unable to work  3. Behavioural activation 4. Suicidal thoughts or attempts 5. Intimate partner violence 6. Resource use and costs of illness   Adverse events 1. Deaths 2. Suicide attempts 3. Unplanned admissions to hospital from any cause  Carers  Nil  Process/health workers Nil  Economic outcomes  Cost per QALY gained (from health system and societal perspectives)  Time points: baseline, 3 months post intervention
Notes	Source of funding: Wellcome Trust Senior Research Fellowship   Notes on validation of instruments (screening and outcomes): all validated  Additional information: Patel V, Weobong B, Nadkarni A, et al. The effectiveness and cost effectiveness of lay counsellor‐delivered psychological treatments for harmful and dependent drinking and moderate to severe depression in primary care in India: PREMIUM study protocol for randomized controlled trials. Trials 2014;15:101. Declarations of interest ‐  DM received honoraria for lectures not related to this work from Otsuka Pharmaceuticals, Janssen‐Cilag, and H Lundbeck. CGF holds a Principal Research Fellowship from the Wellcome Trust. All other study authors declare no competing interests Handling the data: as per footnotes in data and analysis  Prospective trial registration number: International Society for the Registration of Clinical Trials (ISRCTN95149997)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "an independent statistician generated a randomisation list in randomly sized blocks (block size four to six [two to four for men because we anticipated relatively fewer men on the basis of the epidemiology of the prevalence of depression and did not want imbalance between groups]), stratified by PHC and sex. Assignments..." Judgement comment: random sequence generation done
Allocation concealment (selection bias)	Low risk	Judgement comment: cluster‐RCT but appropriate block randomisation
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Cluster allocation but different PHCs, so low risk of contamination? Protocol: quality indicators will be assessed through ratings of 10% of audio‐recording transcripts of all sessions by independent experts blind to outcome data using respective PT quality assessment scales. Participants will be purposively recruited after completion of 12‐month outcome assessments by independent statistician (to maintain blinding) to ensure balance of arms, recovery status, and PHCs. Baseline assessments will be carried out by health assistants in the PHC before randomisation
Blinding of outcome assessment (detection bias) all outcomes	Low risk	Outcomes were assessed blindly: "the three‐ and 12‐month outcome assessments will be carried out by an independent team of field workers who have no contact with the PHCs and who will be entirely community based (that is, assessments will be done at home, to minimize the risk of unmasking)"
Baseline outcome measurements similar	Low risk	Table 1: baseline outcome data similar in terms of PHQ‐9 score
Baseline characteristics similar?	Low risk	Baseline characteristics similar
Incomplete outcome data (attrition bias) Efficacy data	Low risk	All outcome data from protocol reported
Incomplete outcome data (attrition bias) Safety data (e.g. adverse events)	Low risk	All adverse events reported (suicide attempts and relapse and suicidal behaviour)
Protection against contamination	Low risk	Contamination unlikely between PHC groups as patients attending practice usually will be from the locality
Selective reporting (reporting bias)	Low risk	No evidence of selective outcome reporting. All outcomes planned in published trial protocol were reported
Other bias	Unclear risk	Some scores reported at 12 months are not in original trial published; however they were in the 'analysis plan' (e.g. PHQ‐9 score at 12 months (mean); relapse scores, recovery scores). This has little impact potentially on outcomes (other than PHQ‐9 is not validated as a good score in India (Patel 2008)