Koniak‐Griffin 2008.

Methods	Study design: control before‐after study Sampling frame: adolescent Latino couples in Los Angeles, CA Sampling method: non‐random assignment Collection method: self report questionnaire (n = 49) Description of the community coalition: University of California Los Angeles School of Nursing, in collaboration with a community–based organization (CBO) already successful in providing innovative services to adolescent fathers (the Bienvenidos Family Services National Latino Fatherhood and Family Institute (NLFFI)), developed and pilot–tested a culturally based HIV prevention program for young couples; funded by the California Collaborative Research Initiative of the University‐wide AIDS Research Program
Participants	Communities: Los Angeles Country: USA Ages included in assessment: 14 to 23 years old Reasons provided for selection of intervention community: high proportion of Latino couples at risk for HIV/AIDS Intervention community (population size): unknown Comparison community (population size): unknown
Interventions	Name of intervention: HIV Risk Reduction for Latino Adolescents Theory: Healing the Wounded Spirit (Tello 1998) and Gender and Power (Amaro 1995) Aim: to reduce unprotected sex among adolescent Latino couples Description of costs and resources: unknown Components of the intervention: small group sessions with HIV/AIDS education and counseling Start date: unknown Duration: 12 hours of content provided in 6 sessions
Outcomes	Outcomes and measures: self report of rate of unprotected sex (questionnaire) Time points: baseline, follow‐up at 3 months and 6 months post intervention
Notes	Funding source: University of California
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Non‐random assignment
Allocation concealment (selection bias)	High risk	Non‐random assignment
Baseline outcome measurement similar	High risk	Baseline risk among females lower in control group
Baseline characteristics similar	High risk	Differences in sociodemographic characteristics at baseline
Blinding of outcome assessment (detection bias) All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias) All outcomes	High risk	High number of participants not included in follow‐up data analysis
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding
Protection against contamination	Low risk	Intervention offered in small groups
Selective reporting (reporting bias)	Low risk	Outcomes of interest described