Thirty-six publications were included. The number of participants included was not stated.
Of the identified studies: 19 publications were evaluations of National Institute on Drug Abuse (NADR) interventions; 12 were from the ongoing Cooperative Agreement Program. Most evaluations were based on one-group pretest and posttest with no control group; a handful of studies used an experimental or quasi- experimental design to assess the behavioural effects of outreach interventions vs the sensitizing effects about HIV risks or the impact of secular events or trends; several studies compared the effects of basic vs enhanced outreach-based interventions; and only one longitudinal study used seroincidence as an outcome measure.
The majority of evaluations shows that IDUs in a variety of places and time periods changed their drug-related and sex- related risk behaviours following participation in outreach (reported below as risk behaviour: number of studies reporting significant reduction in risk behaviour 'out of' number of studies evaluating given risk behaviour). Stopped injection use: 10 out of 11 studies; Reduced injection frequency: 17 out of 18 studies; Stopped/reduced multi person reuse of needles/syringes: 16 out of 20 studies; Stopped/reduced reuse of cookers, cotton, rinse water: 8 out of 12 studies; Stopped/reduced crack use: 7 out of 7 studies; Increased needle disinfection: 10 out of 16 studies; Entered drug treatment: 6 out of 7 studies; Increased condom use or had less unprotected sex: 16 out of 17 studies.
Overall drug risk behaviours calculated by combining results from 4 or more studies:
Stopped injecting drugs at time of follow-up (5 studies, 6254 patients): median at baseline 100% injected; range of change at follow-up 24% to 31% IDUs stopped injecting; median change at follow-up 26% of IDUs stopped injecting.
Reduced drug injection frequency: (7 studies, 2447 patients): median at baseline 73 injections per month; range of change at follow-up 11 to 62 fewer injections per month; median change at follow-up 28 fewer injections per month.
Stopped/reduced reuse of needles/syringes (4 studies, 2830 patients): median at baseline 37% reused syringes; range of change at follow-up 14% to 43% fewer IDUs reused syringes; median change at follow-up 19% fewer IDUs reused syringes.
Stopped/reduced reuse of cookers, cotton, rinse water (4 studies, 2554 patients): median at baseline 67% reused equipment; range of change at follow-up 16% to 34% fewer IDUs reused equipment; median change at follow-up 27% fewer IDUs reused equipment.
Seroincidence (1 quasi experimental study followed up 6 waves of data between 1988 and 1992 and compared subjects exposed to an intensive outreach program with a non equivalent control of IDUs not exposed to outreach): injection risk declined from 54% at wave 1 to 14% at final follow-up; seroincidence among outreach participants declined from 8.4 to 2.4 per 100 person years. The only behavioural risk factor associated with reduction in HIV seroincidence was a reduction in injection risk. Risk behaviour in outreach participants was 14% compared to 50% in control group.