Ayoughi 2012.

Study characteristics
Methods	Study design: RCT  Duration of study: recruitment September to October 2009, intervention for 5 to 13 weeks, follow‐up at 3 months; study concluded in March 2010
Participants	Country: Afghanistan  Income classification: low income from 2009 to 2010  Geographical scope: in Balkh Province of Afghanistan, in its capital Mazar‐e‐Sharif, 320 km northwest of Kabul, urban  Healthcare setting: PC facility  Mental health condition: depression  Population   1. Age: ≥14 years  2. Gender: both (but all men dropped out, so only females in study sample)  3. Socioeconomic background: more than 30 years of war has disrupted the lives of 2 generations of people; most were unemployed ‐ intervention group 83.9%, control group 96.7%; most have no education  4. Inclusion criteria    a. Adults, aged 18 years or over with score ≥ 11 on Patient Health Questionnaire (PHQ‐9)  b. Confirmed 5th edition of Diagnostic and Statistical Manual of Mental Disorders (DSM‐V) diagnosis of depression using the Composite International Diagnostic Interview (CIDI)  c. Informed consent provided  5. Exclusion criteria   a. Schizophrenia  b. Mental retardation  c. Neurological disorder
Interventions	Stated purpose: to assess common treatment for mental health patients in Afghanistan  INTERVENTION (n = 31) Name: Psychosocial Counselling  Delivered by: lay PHWs Title/name of PW and number: psychosocial counsellors (N = 30 trained and deployed at local health care centres in North Afghanistan, with N = 3 at a counselling centre at Mazar‐e‐Sharif)  1. Selection: not mentioned  2. Educational background: not mentioned  3. Training: 3.5 months of intensive training ending with a final examination, ensuring the required quality standard of counsellors set by the Ministry of Public Health. Training provided by experienced local physicians who had been trained as psychosocial counsellors in an extensive, 2‐year training programme for psychosocial counselling in 2005/2006 and gathering considerable experience in counselling thereafter  4. Supervision: research team based in Kabul  5. Incentives/remuneration: employed at clinic  Intervention details  1. Duration/frequency: first 5 held over 5 weeks. Each session 45 to 60 minutes long; in case of clinical necessity, up to 8 additional counselling sessions could be added, following selected intervention modules of cognitive‐behaviour therapy. Supervising team needed to agree on whether patient required additional sessions  2. Content of intervention: approach has been developed and adapted to the sociocultural background of Afghanistan. Watzlawick's short‐term therapy and Antonowsky's salutogenetic approach lie at the core; selected intervention modules of cognitive‐behaviour therapy have additionally been included. Resource‐ and problem‐solving approach that aims at restoring self‐efficacy and developing resources, enabling patients to re‐participate in their daily life in a satisfying and responsible way. Approach is geared towards improving patients' general mental, physical, social, and spiritual health. Emphasis on a sense of coherence, covering comprehensibility, manageability, and meaningfulness; followed treatment guidelines in the manual "Professional Package for Psychosocial Counsellors working in the BPHS in Afghanistan"; no medication prescribed  CONTROL: routine medical treatment (n = 30) Usual medical treatment was carried out by 4 local physicians, who regularly examined patients in the control group and prescribed medication. We agreed with them on a weekly appointment and precise documentation on prescribed medication. This intervention can be described as usual treatment within the Basic Public Health Care System for patients reporting mental suffering and psychosocial problems. Routine medical treatment by local physicians started immediately after initial expert interview. Patients receiving medical care were treated at local health care centre for the following 3 months  CO‐INTERVENTIONS: nil
Outcomes	Patients   1. HSCL Depression Score  2. HSCL Anxiety Score  3. MINI diagnosis of major depression  4. Screening for depression change between pretest and follow‐up  Carers  Nil  Process/health workers Nil  Economic outcomes None Time points: baseline, 3 months post baseline (*intervention length for intervention /counselling group = 5 to 8 weeks, intervention length for control/medication group = 3 months)
Notes	Source of funding: EU Delegation in Kabul  Notes on validation of instruments (screening and outcomes): all validated  Additional information: declarations of interest ‐ none Handling the data: as per footnotes in data and analysis  Prospective trial registration number: ClinicalTrials.gov Identifier: NCT01155687
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Judgement comment: not completely random, as every other day "Our team allocated the participants to one of the treatment conditions based on a daily alternation routine, meaning that alternately, one day patients were allocated to the medication group, and the next day to the counselling group"
Allocation concealment (selection bias)	High risk	Judgement comment: randomisation was not concealed
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Patients and assessors were all aware of treatment allocation, but this may not have influenced results
Blinding of outcome assessment (detection bias) all outcomes	High risk	"The interviewers who carried out the follow‐up test were not fully blind to the treatment condition as the two types of intervention (psycho‐ vs pharmacotherapy) were very different and thus sometimes revealed through unsolicited information given by the patient. Moreover, although the knowledge about the treatment condition was not updated before follow‐up, we cannot rule out that the expert‐interviewer still remembered the treatment condition of some patients"
Baseline outcome measurements similar	Unclear risk	Yen ‐ LOW: similar baseline HSCL depression, HSCL anxiety, MINI, screening for depression, psychosocial stressors, and coping mechanisms results in both groups Nadja ‐ HIGH: OK apart from how many people were on medication (none in psychosocial counselling and 29 in control (medication only))
Baseline characteristics similar?	Low risk	Baseline gender, ethnicity, religion, marital status, education, employment, number of people living in the household, medication, and age all showed no statistically significant difference
Incomplete outcome data (attrition bias) Efficacy data	Unclear risk	Yen ‐ LOW: efficacy analysis; analysis was carried out with and without participants who dropped out; there was no difference in treatment effect Nadja ‐ UNCLEAR: no access to protocol to check
Protection against contamination	Unclear risk	Judgement comment: Yen ‐ LOW: psychosocial counselling (intervention) and usual medical care (control) delivered by different providers, and psychosocial counselling programme highly specialised, making contamination between groups unlikely, even though they may have lived in the same area and received treatment at centres close to one another. Nadja ‐ UNCLEAR: although in the same clinic, it would not be possible for control group to be contaminated by psychosocial counselling, unless hearing indirectly about it from friends who may be included in the intervention arm. Given the 2 arms are in the same clinic, contamination is possible
Selective reporting (reporting bias)	Low risk	All outcomes planned in clinical trial protocol were reported in this article
Other bias	Low risk	No other sources of bias were found