different populations (Käll, 2005). To address the possibility that the effect of NSEs can vary with the stage of epidemic, both Hurley et al. and MacDonald et al. analyzed cities where the initial measured seroprevalence among IDUs was less than 10 percent. In the Hurley et al. study, the authors did not find a significant association between NSE presence and the trajectory of the epidemic. However, MacDonald and colleagues did find a significant relationship: the mean annual weighted increase in HIV prevalence was 32.1 percent in cities that did not introduce NSEs, compared with a mean annual decrease of 7.8 percent in cities with NSEs (p=0.03).

Multiple studies show that NSEs do not reduce transmission of HCV, which has been attributed to the apparent failure of NSEs to provide enough ancillary injecting equipment such as sterile cotton, water, and alcohol wipes. While NSEs do reduce the frequency of reported needle and syringe sharing, they do not appear to reduce the sharing of other injecting equipment, such as cookers, cotton, rinse water, and drug solution (Hagan and Thiede, 2000; Sarkar et al., 2003; Taylor et al., 2000; Mansson et al., 2000). In contrast, a case-control study by Hagan et al. (1995) in Seattle found that NSE attendance was associated with a six-fold decrease in acquisition of hepatitis B virus (HBV), and a seven-fold decline in HCV acquisition. Given the high prevalence of HCV among IDUs, this represents an important area for future research.

Based on this evidence, the Committee concludes: