et al., 1991; Des Jarlais et al., 1992; Selwyn and Alcabes, 1994). The recognition of multiple pathogens that can be transmitted parenterally by injection drug users is important, because development and implementation of prevention programs directed at HIV infection can be viewed more broadly as prevention programs for blood-borne pathogens in general.
Sexual intercourse was implicated as a primary mode of transmission of the virus even before the etiologic agent (HIV) had been identified (Jaffe et al., 1983a, 1983b; Centers for Disease Control, 1981, 1982). However, the sexual transmission of the virus is not highly efficient, and the risk of acquiring the infection as a result of a single sexual exposure is relatively low (Friedland and Klein, 1987; Institute of Medicine, 1988; Holmberg et al., 1989). That does not mean that documented evidence of people becoming infected after one or only a few sexual contacts does not exist (Padian et al., 1988). Sexual transmission depends on the type and frequency of sexual encounters, as well as the prevalence of other risk factors (e.g., condom use). Receptive anal intercourse is particularly dangerous regardless of the sexual orientation of the individuals (Kingsley et al., 1987; Winkelstein et al., 1987), and frequent sexual exposures (vaginal or anal) to an infected partner also increase the likelihood of transmission (Padian et al., 1990; Lazzarin et al., 1991).
With respect to heterosexual transmission, as with other sexually transmitted diseases, women are at higher risk than men (Aral, 1993). Nonetheless, as is also the case with other sexually transmitted diseases, transmission occurs in both directions (male-to-female and female-to-male). The differential efficiency of transmission between the sexes has led to some debate about whether current estimates of female-to-male transmission rates are accurate (Redfield et al., 1985; Haverkos and Edelman, 1985, 1988; Polk, 1985; Handsfield, 1988; Padian et al., 1991; Haverkos and Battjes, 1992; Haverkos and Needle, 1994; Haverkos, 1994).
Transmission rates may also vary depending on the risk group of the originally infected partner. The risk of transmission has been shown to be lower for female partners of hemophiliacs and bisexual men and for partners of transfusion-infected persons than it is for female or male partners of injection drug users (Padian, 1987; Padian et al., 1987; Curran et al., 1988; De Gruttola and Mayer, 1988; Johnson, 1988).
Delineation of the precise biological mechanisms involved in the heterosexual transmission of HIV has been complicated by the difficulty of identifying a potential series of sexual encounters in which exposure to HIV is known to have occurred each time. We know that cell-free virus is infectious for blood product recipients and that cell-associated virus can infect