Prevalence and correlates of crack-cocaine injection among young injection drug users in the United States, 1997-1999.

OBJECTIVES: We estimated prevalence and identified correlates of crack-cocaine injection among young injection drug users in the United States.

METHODS: We analyzed data from the second Collaborative Injection Drug Users Study (CIDUS II), a 1997-1999 cohort study of 18-30-year-old, street-recruited injection drug users from six US cities.

RESULTS: Crack-cocaine injection was reported by 329 (15%) of 2198 participants. Prevalence varied considerably by site (range, 1.5-28.0%). No participants injected only crack-cocaine. At four sites where crack-cocaine injection prevalence was greater than 10%, recent (past 6 months) crack-cocaine injection was correlated with recent daily injection and sharing of syringes, equipment, and drug solution. Lifetime crack-cocaine injection was correlated with using shooting galleries, initiating others into drug injection, and having serologic evidence of hepatitis B virus and hepatitis C virus infection.

CONCLUSIONS: Crack-cocaine injection may be a marker for high-risk behaviors that can be used to direct efforts to prevent HIV and other blood-borne viral infections.