As illustrated by the literature reviewed in Chapter 7, previous evaluation studies have provided valuable insights into issues to be considered for future research endeavors. Such issues include (but are not limited to): categorization of exposure to the intervention condition (ever/never, secondary exchangers); measurement of alternative sources of sterile syringes other than needle exchange programs (e.g., pharmacies, diabetics); opportunistic comparison groups; eligibility requirements; temporal associations of participation program and HIV seroconversion; accounting for alternative modes of HIV transmission; and consolidation and coordination of program effects. A discussion of these issues follows.
Many studies have simply compared ''ever" versus "never used" a needle exchange, with no provisions to allow the researcher to distinguish between different intensity (or frequency) of program exposure across participants. This creates a problem because most needle exchange reports to date identify a substantial proportion of needle exchange participants who have used the program only once. To consider the single-time users similar to frequent users, but different from never users, dilutes any potential effect due to program participation. At a minimum, analyses of the effects of needle exchange program use should stratify by the frequency with which the program is used, to allow for an assessment of a dose response.
At the other extreme, a small proportion (e.g., >25 percent) of program participants have been known to exchange extraordinarily large numbers of needles (e.g., in Tacoma, Chicago, and San Francisco), unlike the single-time users who are not interested in continuing participation for undetermined reasons. Regular low or moderate users who are engaged in personal risk reduction are unlike the high-frequency users who exchange large numbers and are likely to be secondary exchangers . This phenomenon raises serious concerns about the common practice of using means as a summary statistic in the analyses of program effect. Employing this method equates all exchangers and may lead to masking trends in subgroups. Therefore, it is important to delineate subgroups and define risks accordingly.
The role of secondary exchangers is controversial because a small portion of exchange users obtains a large portion of syringes distributed by the programs. The number of syringes exchanged by these participants typically exceeds the number of syringes that would be used by a single individual. Although this could lead to individuals making money to support