Background To implement a randomised trial of a risk-reduction intervention consisting of small group sessions, individual counselling and support groups for African- and Mexican-American adolescent women with STI and abuse history and evaluate the effects of the intervention model vs enhanced counselling for this group on STI at 6 and 12 month follow-up.

Methods Mexican- and African-American adolescent women (n=409, 13–18 years) with STI and a history of abuse (physical, psychological, sexual) were recruited from public-health clinics in a metropolitan area of the USA. Following enrolment participants received targeted physical exams including contraception counselling and interviews including assessments for abuse, sexual risk behaviour, substance use, contraceptive use and STI. Participants were encouraged to return for unscheduled visits as needed for counselling or suspected STI or pregnancy. Scheduled follow-up rates at 6 (93%) and 12 (93%) months were high Participation in risk-reduction interventions was also high (92%). Surveillance for STI and pregnancy at off-site clinic visits was conducted and documented throughout the study.

Results At study entry, participants reported sexual (58.9%), physical (76.8%) and psychological (82.4%) abuse histories, ever pregnant (45%), currently pregnant (18.7%), ever used birth control (29.9%), currently in school (63.2%), previous arrest (56.2%), home runaway (54.5%), any substance use (94.1%) including marijuana (82.6%), heroin (13.4%), cigarettes (74.6%), alcohol (77.5%), cocaine (41.1%) and benzodiazepine (45%); forced sex (29.1%), sex when high/out of control (13.9%), sex with friends for benefits (36.2%), steady partner in last 3 months (77.4%) however off and on (31.4%). More workshop vs control group participants reported physical abuse (82.9% vs 71.0%), home runaway (63.1% vs 46.4%), cigarettes (78.9% vs 70.5%) and alcohol use (81.4% vs 73.8%). More control group participants were currently in school (68.1% vs 58.1%) and had used condoms (95.3% vs 86.4%). These variables plus age/ethnicity were initially included in the regression model for analysis of data from 0 to 6 and 0–12 month follow-ups. Fewer STI re-infections for intervention vs control group participants at each study interval were identified; physical abuse and substance use remained in the model.

Conclusions Behavioral interventions may reduce of STI among African-and Mexican-American adolescent women with STI and a history of abuse.