HIV-INFECTION AMONG NON-INJECTING DRUG-USERS ENTERING DRUG-TREATMENT,UNITED-STATES, 1989-1992

Objective: To describe HIV seroprevalence among non-injecting drug use rs (non-IDU) entering sentinel drug treatment centers in the United St ates. Design: Anonymous, blinded (unlinked) HIV seroprevalence surveys . Setting: Sixty-eight sentinel drug treatment centers in 37 United St ates metropolitan areas. Participants: Consecutive sample of clients a dmitted to sentinel drug treatment centers from January 1989 through D ecember 1992. Of 84 617 clients, 37 633 (44.5%) had used illicit drugs but reported no injecting drug use since 1978. Main outcome measures: Center-specific, metropolitan area-specific, and national median HIV seroprevalence rates. Results: National median center-specific HIV ser oprevalence among non-IDU was 3.2% (range, 0-15.2%). Rates varied wide ly by geographic area. Median rates were highest in the northeast (5.6 %; range, 0-15.2%), intermediate in the south (3.4%; range, 0.6-8.0%), and generally lower throughout the rest of the country: midwest (1.3% ; range, 0-3.1%) and west (1.8%; range, 0-14.5%). When stratified by t reatment center, there were few statistically significant differences in seroprevalence among African Americans, Hispanics and whites. The m edian rate was 3.4% among men and 2.7% among women. Rates among non-ID U were lower than among IDU attending the same drug treatment centers, but consistently higher than among heterosexual patients attending se xually transmitted disease clinics in the same metropolitan areas. Con clusions: HIV seroprevalence among non-IDU entering drug treatment is high in many metropolitan areas. HIV prevention and education efforts in drug treatment centers should target sexual as well as drug-use ris k reduction for all clients.