thors found no association between frequency of NSE use and seroconversion.
As in Montreal, the Vancouver NSE originally operated with strict limits on the number of needles and syringes that users could exchange at any one time, and the program operated in only one location. The Vancouver program also made dramatic changes in response to early results. Specifically, the NSE switched from a limited exchange approach to a need-based approach—allowing unlimited distribution of needles/syringes—and greatly increased the number of access points. The program also began offering a variety of distribution methods, including fixed, mobile, and home delivery. HIV incidence among IDUs has since fallen by 30 percent (Personal communication, Chris Buchner, Vancouver Coastal Health Authority, May 5, 2006).
Several studies in Amsterdam found no significant relationship in either direction between NSE participation and HIV incidence (van Ameijden et al., 1992; Coutinho, 2005). Several other papers by these authors (van Ameijden and Coutinho, 1998, 2001) found initial reductions in risk behavior after NSE and other interventions began, but no further reductions over time. These studies also found that NSE was not associated with an increase in injecting drug use, and attributed declines in injecting to cultural and ecological factors. Krol (2006) reached the same conclusion.
Several ecological studies from the developed world found that early, comprehensive programs of outreach, prevention, education, and access to sterile injecting equipment may prevent the expansion of IDU-driven epidemics. Ecological studies, as well as serial cross-sectional studies, reflect community-level patterns of prevalence and risk behaviors rather than patterns or changes at the individual level. For example, Des Jarlais et al. (1995) examined five cities (Glasgow, Scotland; Lund, Sweden; Sydney, Australia; Tacoma, U.S.; and Toronto, Canada) where HIV was introduced into the IDU population but infection rates remained below 5 percent for at least 5 years. The authors found that all five cities had pursued early prevention activities, such as offering sterile injection equipment and community-based outreach. Such interventions may also help reduce HIV prevalence and incidence among IDUs in more mature HIV epidemics, such as in New York City (Des Jarlais et al., 2005a).
In a study of 81 cities, Hurley and colleagues (1997) found that annual HIV seroprevalence between 1988 and 1993 rose by 5.9 percent in 52 cities without NSEs, and fell by 5.8 percent in 29 cities with NSEs. In a similar analysis of 99 cities, MacDonald and colleagues (2003) found that annual HIV prevalence fell by 18.6 percent in cities that introduced NSEs, and rose by 8.1 percent in cities without NSEs. Critics objected that this study did not account for the stage of the epidemic in these cities, and that the researchers used different protocols to collect data on seroprevalence in