it cannot be stated what portion of the reduction in risk behaviors is due to the needle exchange program.
An evaluation of the Portland needle exchange program (Oliver et al., 1994) assessed change in risk behaviors among 753 needle exchange program participants by comparing data collected on each study participant during an intake interview and a 6-month follow-up interview. Significant declines in sharing behaviors (rent needles and syringes: intake = 9 percent compared with 3 percent at follow-up; borrowed needles and syringes: intake = 20 percent compared with 7 percent at follow-up) and increased use of bleach (intake = 51 percent compared with 65 percent at follow-up) were observed. When drug-use risk behaviors of frequent attenders (attended four or more times) were compared with risk behaviors of those who attended three or fewer times, frequent attenders reported greater risk reduction on borrowing and returning used needles to the program. To supplement the pre-post analysis, needle exchange participants were also compared with another group of drug injectors, clients of Portland's National AIDS Demonstration Research (NADR) program (i.e., all received bleach and HIV education and a subset also received counseling). Frequent needle exchange participants were found to be less likely to reuse needles without cleaning or to improperly dispose of used needles than were the NADR clients. The two groups were not found to differ on other risk behaviors assessed. It is worth noting that there was little overlap between the two groups (11 percent). The two interventions apparently are recruiting different participants.
In sum, from the earliest studies of needle exchanges, there has been a dominant trend in the data showing significant and meaningful associations between participation in needle exchange programs and lower levels of drug-use risk behaviors, and small or no change in sexual risk behaviors. The most recent data continue to reflect this trend. Moreover, this pattern of findings has also been observed in foreign cities (Davoli et al., 1995; Hunter et al., 1995).
Two recent ecological studies examined trends in HIV seroprevalence rates among injection drug users, one in New York City and the other in San Francisco (Des Jarlais et al., 1994a; Watters, 1994). Both reported a stabilization of HIV seroprevalence rates that coincided with reductions in high-risk behaviors and the implementation of various prevention programs (including outreach, education, testing and counseling, bleach and condom distribution, and needle exchange programs). Although these ecological studies do not provide direct causal evidence of the effect of such programs, they nonetheless document a pattern in behavioral risk reduction that corresponds