Dybdahl 2001.

Study characteristics
Methods	Study design: randomised 2‐sided parallel‐group open‐label assessor‐blinded controlled trial (unit of randomisation: mother‐child dyads; unit of analysis: individuals) Duration of study: 1995 to 1996
Participants	Country: Bosnia Income classification: low income Geographical scope: urban (town of Tuzla, a multi‐ethnic industrial town in northeastern Bosnia) Healthcare setting: home (1 refugee settlement; private accommodation for refugees) Mental health condition: child mental health (PTSD, mental health, behavioural problems, scholastic difficulties) Population: mother‐child dyads (internally displaced refugees) 1. Age: mothers: mean 30.7 years (SD 4.9), range 20 to 44 years; children: mean 5.5 years (SD 0.7) 2. Gender: both (children: 48 girls, 39 boys) 3. Socioeconomic background: mothers: 85% urban origin, education 14% illiterate (mean 5.3 years, SD 2.8; range 0 to 14 years), married 63%, widowed 36%, divorced 1%, living in private accommodation 60%, living in refugee camp 40% 4. Inclusion criteria: internally displaced Bosnian mothers with a child aged 5 to 6 years 5. Exclusion criteria a. Not participating in any other intervention programme b. Unlikely to move out of the area before November 1996
Interventions	Stated purpose: to provide early childhood care and education as well as psychosocial support to traumatised children by working with their mothers to help them resolve grief and improve parenting and by providing a well‐functioning family environment utilising non‐medical professionals in a post‐conflict situation INTERVENTION Name: Psychosocial Intervention (+ basic medical care) ‐ 42 people Delivered by: CP Title/name of PW and number: group leaders ‐ preschool teachers trained for the study ‐ 5 1. Selection: not specified in this report 2. Educational background: as above 3. Training (contents, duration, by whom): to a group of 3 to 8 group leaders, provided by mental health professional a. Duration: 5‐day workshop. Before arrival, participants received basic information about the programme and its background and aims b. Content: participants were introduced to one another and received written material and introductory training on some of the key issues such as trauma, child development, and the importance of interaction and communication (mother‐child) in two 3‐hour seminars. Then 3 days of more detailed description of the programme and reinforcement through group work, demonstrations, role‐plays, and discussion of the above topics (roles of caretaker, trauma and its effects on adults and children, groups and group dynamics, supervision, logbook) 4. Supervision: weekly group meetings (with 6 to 8 group leaders along with a supervisor (a mental health professional) (later twice a month) 5. Incentives/remuneration: as above Intervention details 1. Duration/frequency: group leader met weekly with 2 groups of mothers (5 per group) for 5 months; 1 additional visit to each mother at her home at start of programme 2. Content of intervention: group work using a manual‐based approach derived from therapeutic discussions with war‐traumatised women at the Psychological Centre in Tuzla (1993‐1996) and the ICDP; semi‐structured group discussions introduced by group leaders dedicated to providing information about trauma and trauma reactions in adults and children, as well as suggestions for how to meet common post‐traumatic needs and problems, with emphasis on strengthening participants' own coping strategies and reinforcing existing normal basic communication and interaction skills. Direct attention was given to mothers and their mental health, to their beliefs and knowledge about children, and to the reactions and needs of adults and children following traumatic events. Mothers were also visited once at home to establish rapport and to express support CONTROL: non‐intervention group; participated in evaluations and received free basic medical care (45 people) CO‐INTERVENTIONS: free basic medical care by local physicians provided for both groups; vitamins or iron was given to 52 children (66% in intervention group; 81% in control group)
Outcomes	Children 1. Description of child (rated by mothers; 11 characteristics; 7‐point differential) 2. Mothers' ratings of children's problems § (10 problems; 4‐point scale; total 30 points) 3. Mothers' ratings of concentration problems § (yes/no) 4. Raven's Coloured Progressive Matrices § 5. Children's interview (modified Birleson Depressive Inventory; modified by removing 2 of 13 items; scored 0 to 32; 11 used as cutoff for depression) 6. Well‐being § 7. Psychologists' observations § (video‐rated; 14 items; 4‐point scale; scored on 2 factors ‐ problems 0 to 32; resources 0 to 16) 8. Anthropometrics: haemoglobin § Mothers 1. Perceived Social Support 2. IES (reported in adult PTSD) 3. Well‐being § Process/health workers Not reported Economic outcomes Not reported (*: primary outcomes of the study; §: outcomes that we have not reported in this review) Outcomes not used in quantitative synthesis War Trauma Questionnaire (given at baseline) Time points: baseline, 5 to 6 months after recruitment (0 to 1 month post intervention)
Notes	Source of funding: UNICEF; University of Tromso Notes on validation of instruments (screening and outcomes): mothers' ratings of child's concentration and concentration problems; perceived social support: not validated separately; IES scores: not diagnostic of PTSD but some literature suggests IES score above 33 suggestive of PTSD Additional information: group work is described in Dybdahl 1996 and Dybdahl 1999. Declaration of interests ‐ none Also included in prevention review, as some mothers seem to have a mental disorder, and others minimal to moderate psychological distress  Handling the data: as per footnotes in data and analysis Prospective trial registration number: not registered Also included in prevention review, as unsure about the population (roughly half the intervention group has mental distress or a mental disorder at baseline; thus the intervention may be a treatment, whereas for the other half, it could be a prevention strategy)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote from report: "the assignment was random. All the names of the mother–child dyads were written on pieces of paper, which were folded, mixed together, and then separated into two piles at random so that one pile formed the intervention group and the other pile formed the control group"
Allocation concealment (selection bias)	Unclear risk	Comment: not stated
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Comment: participants and intervention personnel were not blinded to allocation, but no evidence of impact on outcome is provided
Blinding of outcome assessment (detection bias) all outcomes	Low risk	Comment objective outcomes: physical and psychosocial outcomes were measured by teams of physicians and experienced health worker assistants not involved in delivering interventions and blind to interventions Comment subjective outcomes: physical and psychosocial outcomes were conducted by teams of physicians and experienced health worker assistants not involved in delivering interventions and blind to interventions
Baseline outcome measurements similar	Low risk	Comment: baseline imbalances in prognostic variables noted for psychosocial support for mothers and for well‐being (but not statistically significant) and for children's haemoglobin (P = 0.3); however, analyses included differences between groups in changes from baseline
Baseline characteristics similar?	Unclear risk	Comment: mothers in refugee camps reported more war trauma and were more likely to be widowed during the conflict
Incomplete outcome data (attrition bias) Efficacy data	High risk	Quote: "twelve of the families dropped out of the study and did not participate in scheduled interventions: 7 from the intervention group, and 5 from the control group. Several of the mothers and children did not complete all tests at both test periods for a variety of reasons; thus the number of participants varied from test to test" Comment: denominators for each of the tests are not provided by intervention or control
Protection against contamination	Unclear risk	Comment: mothers in refugee camps could have discussed contents of the intervention while supporting mothers in control group
Selective reporting (reporting bias)	Unclear risk	Comment: protocol is not available, but all measures stated in methods are reported
Other bias	High risk	Comment: multiple statistical analyses were used without pre‐specified primary or secondary outcomes; analyses corrected for multiple comparisons yielded non‐significant results