Rahman 2019.

Study characteristics
Methods	Study design: single‐blind, cluster‐RCT. Unit of allocation: lady health worker (LHW) community cluster ‐ 17 in each arm, 34 total Duration of study: January to August 2016 (enrolled between 11 January 2016 and 21 August 2016; 1 week after intervention, follow‐up results; primary outcomes 3 months after intervention)
Participants	Country: Pakistan Income classification: lower‐middle income between 2016 and 2017 Geographical scope: 2 rural union councils of Swat, Odigram, and Ghalegay, in community clusters (defined as a neighbourhood of about 150 households covered by an LHW) (estimated population of about 2 million) Healthcare setting: community‐based (delivered through community‐based groups) Mental health condition: adults (women) with anxiety, depression, PTSD Population 1. Age: 18 to 60 years 2. Gender: female  3. Socioeconomic background: more than 85% live in rural areas; 2% no education; 85% housewives 4. Inclusion criteria a. Scored ≥ 3 on General Health Questionnaire‐12 (GHQ‐12) ‐ a screening questionnaire for common mental health disorders ‐ and ≥ 17 on the WHO Disability Assessment Schedule 2·0 (WHODAS) ‐ a questionnaire for functional impairments 5. Exclusion criteria a. Women with severe mental health disorders (e.g. psychotic disorders, substance dependence) b. Severe cognitive impairment (e.g. severe intellectual disability) c. At risk of imminent suicide
Interventions	Stated purpose: to establish the effectiveness of WHO's Problem Management Plus group intervention in a conflict‐affected setting INTERVENTION Name: WHO’s Problem Management Plus (PM+) intervention Delivered by: LHW Title/name of PW and number: for treatment arms of trials, PM+ treatments were delivered by 17 lady health workers (LHWs) and 6 facilitators (therapists) 1. Selection: these LHWs were community health workers, each responsible for a community cluster of approximately 1000 people or 150 homes, visiting 5 to 7 homes per day. Facilitators were members of the local community 2. Educational background: facilitators were local graduates with bachelor’s degrees without mental healthcare experience 3. Training: facilitators received 7 days of intervention training by a master trainer (KSD). Intervention training included education on adversity and its effects on mental health, basic helping skills, delivery of intervention strategies, skills in group facilitation, and facilitator self‐care. To assess competency, all facilitators delivered 1 practice group each at an accelerated rate (5 sessions in 2 weeks) with participants living outside the trial area and under intensive supervision (10 hours' supervision over 2 weeks), who were assessed for their competency using a specially developed checklist that evaluated basic counselling skills and their use of intervention strategies through direct observation 4. Supervision: supervision of facilitators was done through 2 hours of weekly group session by experienced Islamabad‐based supervisors via telecommunication software (Skype Technologies, Palo Alto, CA, USA). In turn, supervisors received 1·5 hours' fortnightly supervision via Skype by the master trainer in Sydney, NSW, Australia. Supervision included review of participants’ progress and individual case management, refresher training on strategies, and rehearsal of skills through role‐play 5. Incentives/remuneration: facilitators received a small honorarium of USD100 per month Intervention details 1. Duration/frequency: 5 group sessions per week, with approximately 6 to 8 participants per group, each session lasting for approximately 2 hours (excluding breaks) 2. Content of intervention: first session included psychoeducation, goal‐setting, and brief motivational interviewing. Sessions 1 to 4 introduced strategies for stress management, problem‐solving, behavioural activation, and strengthening of social support. Each strategy was reviewed in every subsequent session, and the final session involved revision of learning, education on preventing relapse, and a closing ceremony. Given that many participants were non‐literate, the intervention included locally relevant pictorial materials and adopted a narrative format. The groups, facilitated by local women, gave participants a safe space to share their feelings and to learn from one another’s experiences, allowing a degree of empowerment and control over their lives as they problem‐solved together CONTROL: EUC for all participants comprised feedback about assessment results, the offer of psychoeducation for themselves and accompanying family members, the opportunity to talk about their health with their LHW (who received a half‐day training programme in psychoeducation and supportive communication), and information about options for seeking care for distress (i.e. through the PHC centre or the tertiary healthcare centre). PHC providers received a half‐day training in detection and management of mental health problems and referral pathways for care  CO‐INTERVENTIONS: none
Outcomes	Patients 1. Hospital Anxiety and Depression Scale (HADS) 2. 14‐item scale consisting of 2 subscales—HADS‐A (anxiety, 7 items, range 0 to 21) and HADS‐D (depression, 7 items, range 0 to 21) 3. PTSD symptoms recorded using the 20‐item PTSD checklist of the Diagnostic and Statistical Manual of Mental Disorders‐5 (PTSD checklist‐5) 4. WHODAS (WHO Disability Assessment Schedule) 5. Patient Health Questionnaire‐9 (PHQ‐9) Carers None Process/health workers None Economic outcomes Direct cost: cost of acute treatment/care (* study's primary and secondary outcomes) Time points: baseline, 1 week, 3 months
Notes	Source of funding: Office of Foreign Disaster Assistance to WHO and a small travel grant from the University of Liverpool Overseas Development Agency Seed Fund Notes on validation of instruments (screening and outcomes): all instruments validated, not mentioned in WHODAS validation Additional information: declaration of interests ‐ study authors declared no competing interests Handling the data: as per footnotes in data and analysis Prospective trial registration number: Australian New Zealand Clinical Trials Registry: 12616000037404
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "permuted‐block randomisation method was used to generate the randomisation code, with a block size of six. Allocation of clusters was done by an independent statistician based at the University of Liverpool, Liverpool, UK, using a computerised randomisation sequence"
Allocation concealment (selection bias)	Low risk	Quote: "however, the researchers responsible for identifying participants, obtaining consent, enrolment, and outcome assessments were masked to allocation. All assessors resided outside the study area and had no interaction with the intervention team" Judgement comment: assessors and researchers are masked and are not part of the community
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Not possible to blind participants and personnel to the intervention; unlikely to affect outcomes
Blinding of outcome assessment (detection bias) all outcomes	Low risk	Researchers responsible for identification of participants, receipt of consent, enrolment, and outcome assessments were masked to allocation
Baseline outcome measurements similar	Low risk	See Table 2 (baseline outcome measurements are similar)
Baseline characteristics similar?	Low risk	Table 1: similar baseline characteristics
Incomplete outcome data (attrition bias) Efficacy data	Low risk	Attrition rate (about 5% to 6% in both arms) was lower than what was provided for (20%) in the sample size calculation
Incomplete outcome data (attrition bias) Safety data (e.g. adverse events)	Low risk	Safety data collected; no adverse events. Attrition rates do not differ significantly
Protection against contamination	Low risk	Judgement comment: LHW community clusters in 2 union councils were block‐randomised; intervention is 1‐to‐1, so contamination unlikely
Selective reporting (reporting bias)	Low risk	Same outcomes as in protocol, other than costs, which are not published yet
Other bias	Low risk	No comments