Shetgiri 2011.

Methods	Study design: RCT Intervention arm(s): school‐based curriculum intervention Comparator arm(s): standard education Sample size calculation performed: yes; a priori power analyses undertaken Subgroups prespecified: none presented Subgroup analyses: N/A Start date: August 2008 Duration of follow‐up: immediately post intervention Number of follow‐ups: 1 Follow‐up time points: 1 immediately post intervention ICC (if reported): not reported
Participants	Number of schools randomised: N/A; students from 1 school only were randomised Number of participants randomised (total and by arm): 108 overall (intervention 53, control 55) Age (range or mean (SD)) or grade at the start: intervention 14.4, control 13.9 Gender: intervention: 51% male, 49% female; control: 33% male, 67% female Ethnicity: intervention: Latino (81%), African American (8%), white (4%), American Indian/Alaskan Native (2%), Asian/Pacific Islander (0%); control: Latino (75%), African American (9%), white (4%), American Indian/Alaskan Native (2%), Asian/Pacific Islander (4%), other (6%) SES: not reported Inclusion criteria: enrolment of high‐risk students only, with high‐risk status based on rate of absence ≥ 80%, 2 or more disciplinary actions in grade 8, failing 2 or more classes in grade 8, or high levels of family dysfunction (identified by grade teacher, using proxies such as multiple family moves in grade 8, perceived lack of parental involvement, or family conflict) Exclusion criteria: N/S
Interventions	Randomisation before or after baseline survey: before Duration of the intervention (excluding follow‐up): 7 to 9 months Description of the intervention: 28 × weekly peer group sessions, each lasting around 45 minutes, focusing on improving student resilience through activities and counselling to build social skills, communication skills, anger management, conflict resolution, and healthy relationships. Student‐led discussions in groups of 6 to 9 students also took place, with violence exposure, alcohol, and smoking the topics of discussion. Students also participated in field trips, community service activities after school, during weekends, and throughout the summer. Brief description of the theoretical model: N/S Description of the comparator: control group; not stated what was delivered
Outcomes	Primary outcomes: fighting, smoking, alcohol use, marijuana use, other illegal drug use, grade point average scores Secondary outcomes: N/S
Setting	Country: USA; State: California Setting: school Focus: targeted
Process measures	Process data reported: no Method (qualitative or quantitative): N/A Description: N/A Acceptability of the intervention: N/S Adherence to the intervention: N/A Intensity of the intervention: N/A
Statistics	Sample size: N/S Unit of randomisation: individual Unit of analysis: individual Method to promote equivalence between groups: N/S Statistical models: ANOVA and GEE Baseline differences adjustment: yes Repeated measures methods in analysis: N/A; 1 follow‐up only
Notes	Equity: study baseline data reported Funding: study supported in part by Robert Wood Johnson Foundation Clinical Scholars Program and the Clinical Research Scholars Program at the University of Texas Southwestern Medical Center Randomisation method: rolled die and assigned odd numbers to intervention and even numbers to control Clustering accounted for in sample size calculation (if relevant): N/A Cluster randomisation methods to account for clustering in analysis: N/A
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Randomly assigned...by rolling a die and assigning odd numbers to the intervention group and even numbers to the control group"
Allocation concealment (selection bias)	Low risk	"Staff members who enrolled participants were unaware of the allocation of subjects to intervention or control groups".
Blinding of participants and personnel (performance bias) All outcomes	High risk	"After allocation, neither subjects nor staff members conducting the intervention were blinded to group assignment".
Blinding of outcome assessment (detection bias) All outcomes	High risk	Self‐reported outcome measures used
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attrition did not differ between study arms (and attrition was < 10% between arms).
Selective reporting (reporting bias)	Unclear risk	No protocol published
Other bias	Unclear risk	Only 1 school took part in the intervention, and randomisation was carried out at the individual (student) level. Risk of contamination was high between students in intervention and control groups at the same school.