TABLE 3-3 Studies with HIV Incidence or Prevalence Outcomes



Bruneau et al., 1997, Montreal (prospective cohort)

Increased HIV seroconversion among NSE users.

Des Jarlais et al., 2005a, New York City (ecological)

From 1990–2001, HIV prevalence declined.

Des Jarlais et al., 2005b, New York City (serial cross-sectional)

Strong negative relationship between the number of syringes exchanged and estimated HIV incidence.

Hammett et al., 2006, Vietnam and China (serial cross-ectional)

HIV prevalence among IDUs declined in Vietnam and remained stable in China.

Hurley et al., 1997, worldwide (ecological)

Increased HIV prevalence in cities without NSE.

MacDonald et al., 2003, worldwide (ecological)

Increased HIV prevalence in cities without NSE.

Mansson et al., 2000, Sweden (prospective cohort)

No new HIV cases during a median of 31 months among NSE participants.

Patrick et al., 1997, Vancouver (case control)

No association with frequency of NSE use and HIV seroconversion.

Schechter et al., 1999, Vancouver (prospective cohort)

Cumulative HIV incidence was significantly elevated in frequent NSE attenders.

Strathdee et al., 1997, Vancouver (prospective cohort)

Increased HIV and HCV prevalence in the presence of NSE.

Van Ameijden et al., 1992, Amsterdam (case control)

No association between NSE participation and HIV seroconversion.

(1997) used three risk-assessment approaches to examine the association between NSE use and HIV infection. All three analytical approaches associated NSE attendance with a substantial and consistently higher risk of HIV infection. For example, in the cohort approach, in which there were 89 incident cases of HIV infection, the researchers found a 33 percent cumulative probability of HIV seroconversion for NSE users, compared with a 13 percent probability for non-users. In the nested case-control study, consistent NSE use was associated with HIV seroconversion during follow-up (OR=10.5; 95% CI: 2.7–41.0). The analyses employed methodologies to control for a range of confounders, including drug of choice and frequency of injecting drug use in the previous month. These findings persisted after controlling for confounders.

The authors and commentators on this research pointed out that the Montreal NSE appeared to have attracted high-risk cocaine injectors, who injected much more often than heroin users. Also, as shown by the seroprevalence data at baseline, Montreal NSE users had high baseline rates of HIV and hepatitis B infection (Bruneau et al., 1997). The NSE also originally strictly limited the number of needles and syringes users could

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement