Berkowitz 2011.

Methods	Block‐randomised trial of child and family traumatic stress intervention vs supportive therapy
Participants	Included (n = 106) Youth 7 to 17 years of age attending a paediatric emergency department and reporting ≥ 1 new distressing post‐traumatic stress symptom. Mean age: 12 years. Female: 52% Caucasian 32%, African American 37%, Hispanic 22%, multi‐ethnic 7%, other ethnicities 2%. Trauma exposure: motor vehicle accident 24%, sexual abuse 18%, witnessing violence 19%, physical assaults 21%, injuries 8%, threats of violence 5% Excluded Receiving counselling or mental health treatment, developmental delay, psychotic or bipolar disorder, caregiver or participant did not speak English Setting Trauma Section of the Yale Child Study Center, USA, 2006 to 2009
Interventions	Child and family traumatic stress intervention (n = 53) Four sessions involving child and caregiver used cognitive, behavioural and psychoeducational techniques. Identification of trauma responses, behavioural skills and communication between caregiver and child included Control (n = 53) Four sessions over 4 weeks involving child and caregiver included psychoeducation, relaxation training, coping strategies and supportive therapy Therapists Master's and doctoral level clinicians. Fidelity checked and weekly supervision provided
Outcomes	Trauma symptoms Scale: Trauma Symptom Checklist for Children (54‐item) Rater: child/adolescent Anxiety Scale: Trauma Symptom Checklist for Children Rater: child/adolescent When Post therapy and at 3‐month follow‐up
Notes	Standard deviations calculated from standard errors
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Block‐randomised (block size of 10) using number containers
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants (performance bias	Unclear risk	Participants could not be blinded, but both groups received psychological therapy
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Outcome assessment could not be blinded, as all measures were self report
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Analysis appears to be based on 'last observation carried forward', but loss to follow‐up was relatively high: post therapy 25%, 3 months 27%
Selective reporting (reporting bias)	Low risk	All outcomes appear to have been reported
Other bias	Low risk	No other bias was apparent