trends in risk behaviors and HIV seroprevalence over a 6.5-year period among heterosexual injection drug users over 13 cross-sectional surveys between 1986 and 1992. Interviews (5,956) were conducted with injectors in street settings and drug detoxification clinics. During that time period, multiple prevention efforts targeting injection drug users had been implemented (including outreach, education, voluntary HIV testing and counseling, bleach and condom distribution, and needle exchange programs). Among injection drug users who reported sharing needles, the proportion of those who reported ever using bleach increased from 3 percent in 1986 to 89 percent by 1988 and remained relatively constant at that level through fall 1992.
Sexually active heterosexual male injectors also reported significant changes in condom use (i.e., injection drug users reported using a condom 4.5 percent of the time in 1986, compared with 31 percent of the time in late 1992). However, Lewis and Watters (1994) found that a substantial proportion of sexually active male drug injectors, including heterosexuals, bisexuals, and homosexuals, reported frequently engaging in unprotected sex (i.e., reported condom use was low in all three groups). That is, 56 percent of the heterosexuals who had vaginal sex in the prior 6 months reported no condom use; one-third of the homosexuals who engaged in anal intercourse with male partners reported no condom use; and 41 percent of the bisexual men reported no condom use while engaging in vaginal sex and approximately half (52 percent) reported no condom use while engaging in anal or oral sex with females and/or males.
In another ecological report from New York City, Des Jarlais and colleagues (Des Jarlais et al., 1994a) examined trends in reported risk behaviors and seroprevalence among injection drug users for 1984 and 1990 to 1992 by comparing results of two surveys of injection drug users entering a drug detoxification program. Several trends in drug-use risk behaviors were reported. In 1984, for example, 65 percent of injection drug users reported having used shooting galleries in the preceding 2 years; in the 1990 to 1992 survey, only 3 percent reported injecting in shooting galleries in the preceding 6 months. Substantial reductions in sharing behavior were also observed. Use of potentially contaminated needles declined from 51 to 7 percent of injections. Moreover, an increasing proportion of injection drug users entering the detoxification program reported using the needle exchange programs since they opened in 1990. For the 1990 to 1992 period, results also show that needle exchange participation was associated with a downward trend in the proportion of subjects reporting any injection with needles that had been used by someone else and a reduction in the percentage of study participants reporting having passed on used needles to others. The extent to which the reductions in risk behaviors reported in the two surveys can be attributed to the needle exchange program itself is limited by the fact