crack as a sexual stimulant (Grinspoon and Bakalar, 1985; Weiss and Mirin, 1987). However, the relative magnitude of the pharmacologic causal effect of crack (or cocaine) on increased sexual activity is still not well understood. Although there may be increased sexual activity associated with early stages of use, as use of the drug increases, sexual dysfunction follows. Even then, there may be heightened sexual activity—sex for crack exchanges—but at this more advanced stage of drug abuse, the increase in sexual activity appears to be driven by the compulsion to use the drug and tends to be devoid of pleasurable sensation. Hence, if infected by their injection drug-using partners, these noninjection drug-using sexual partners also constitute a conduit into the heterosexual community. Crack may also amplify the spread of HIV because of its strong association with syphilis. Trading sex for crack has resulted in a large increase in syphilis (Centers for Disease Control and Prevention, 1992; Greenberg et al., 1992), which may in turn facilitate HIV transmission by increasing either infectiousness or susceptibility.
Some of the increased rates of transmission from injection drug users to their sexual partners, compared with transmission rates from infected people from other risk groups to their sexual partners, may also be attributed to high-risk sexual practices. In one survey of injection drug users who were not in treatment but who were recruited and interviewed on the street in San Francisco (Lewis and Watters, 1991), 67 percent of the sample reported never using condoms, 15 percent had more than 10 partners, and approximately 35 percent engaged in prostitution or practiced anal sex (or both). In a similar nationwide survey, 70 percent of injection drug users reported never using condoms, and more than 25 percent practiced anal sex (Centers for Disease Control and Prevention, 1990b). In a survey of injection drug users in treatment taken in the Northeast, Texas, and California, only 14 percent reported using condoms (Battjes and Pickens, 1988). In another similar study in New York City, only 5 percent reported condom use at all (Primm et al., 1988); among those who did use condoms, fewer than half used them for all sexual encounters. In the San Francisco survey described above (Lewis and Watters, 1991), more than a third of both bisexual and heterosexual male injection drug users reported that they never used condoms.
In a variety of studies that attempted risk reduction programs among injection drug users, success was greater in influencing participants to change risky injection behavior than sexual behavior (Des Jarlais and Friedman, 1988b). Again, for women this may be particularly difficult because the change to safe sex practices requires the cooperation of the male partner, which may not always be a feasible proposition (Worth, 1988). An additional concern related to power dynamics in sexual partnerships is the disclosure of HIV status. In particular, women may fear disclosing their HIV