J. Bruneau et al., HIGH-RATES OF HIV-INFECTION AMONG INJECTION-DRUG USERS PARTICIPATING IN NEEDLE EXCHANGE PROGRAMS IN MONTREAL - RESULTS OF A COHORT STUDY, American journal of epidemiology, 146(12), 1997, pp. 994-1002
Needle exchange programs (NEPs) are designed to prevent human immunode
ficiency virus (HIV) transmission among injection drug users. Although
most studies report beneficial effects in terms of behavior modificat
ion, a direct assessment of the effectiveness of NEPs in preventing HI
V infection has been lacking. A cohort study was conducted to assess t
he association between risk behaviors and HIV seroprevalence and seroi
ncidence among injection drug users in Montreal, Canada. The associati
on between NEP use and HIV infection was examined in three risk assess
ment scenarios using intensive covariate adjustment for empirical conf
ounders: a cross-sectional analysis of NEP use at entry as a determina
nt: of seroprevalence, a cohort analysis of NEP use at entry as a pred
ictor of subsequent seroconversion, and a nested case-control analysis
of NEP participation during follow-up as a predictor of seroconversio
n. From September 1988 to January 1995, 1,599 subjects were enrolled w
ith a baseline seroprevalence of 10.7%. The mean follow-up period was
21.7 months. The adjusted odds ratio for HIV seroprevalence in injecti
on drug users reporting recent NEP use was 2.2 (95% confidence interva
l 1.5-3.2). In the cohort study, there were 89 incident cases of HIV i
nfection with a cumulative probability of HIV seroconversion of 33% fo
r NEP users and 13% for nonusers (p < 0.0001). In the nested case-cont
rol study, consistent NEP use was associated with HIV seroconversion d
uring follow-up (odds ratio = 10.5, 95% confidence interval 2.7-41.0).
Risk elevations for HIV infection associated with NEP attendance were
substantial and consistent in all three risk assessment scenarios in
our cohort of injection drug users, despite extensive adjustment for c
onfounders. In summary, in Montreal, NEP users appear to have higher s
eroconversion rates then NEP nonusers.