Survival time has increased substantially for adult Brazilian AIDS patients. The timing of these gains and analysis of the predictors of survival both indicate antiretroviral treatment as the cause. These findings demonstrate that universal access to antiretroviral treatment in a developing country can produce benefits on the same scale as in richer countries.

Two decades of experience and research provide new insights into the role of condoms for AIDS prevention in the developing world. This literature review and synthesis is based on computerized searches of the scientific literature and review of conference presentations, publications of national and international organizations, and popular media. Condoms are about 90 percent effective for preventing HIV transmission, and their use has grown rapidly in many countries. Condoms have produced substantial benefit in countries like Thailand, where both transmission and condom promotion are concentrated in the area of commercial sex. The public health benefit of condom promotion in settings with widespread heterosexual transmission, however, remains unestablished. In countries like Uganda that have curbed generalized epidemics, reducing the number of individuals' sex partners appears to have been more important than promoting the use of condoms. Other countries continue to have high rates of HIV transmission despite high reported rates of condom use among the sexually active. The impact of condoms may be limited by inconsistent use, low use among those at highest risk, and negative interactions with other strategies. Recommendations include increased condom promotion for groups at high risk, more rigorous measurement of the impact of condom promotion, and more research on how best to integrate condom promotion with other prevention strategies.

The largest investments in AIDS prevention targeted to the general population are being made in interventions where the evidence for large-scale impact is uncertain.

OBJECTIVES. Little is known about the human immunodeficiency virus (HIV) high-risk sexual practices of gay and bisexual African-American men. These data are needed so that better interventions can be developed and implemented in this population. METHODS. The frequency and correlates of unprotected anal intercourse were examined among 250 gay and bisexual African-American men in the San Francisco Bay Area. The cohort was recruited in 1990 from bars, bathhouses, and erotic bookstores, and through African-American gay organizations, street outreach, advertisements in gay mainstream and African-American newspapers, health clinics, and personal referral from other participants. RESULTS. More than 50% of the men in our sample reported having unprotected anal intercourse in the past 6 months, a considerably higher percentage than that among gay White men in San Francisco through 1988 and 1989. Men who practiced unprotected anal intercourse were more likely to be poor, to have been paid for sex, or to have used injection drugs; to have a higher perceived risk of HIV infection; and to report less social support for concerns about risky sexual behavior. Condom norms, condom efficacy, and negative expectations about using condoms predicted these men's failure to use them. CONCLUSION. In the second decade of the acquired immunodeficiency syndrome epidemic, risk reduction programs are still needed for gay and bisexual African-American men.

A national probability survey of human immunodeficiency virus (HIV)-related risk factors among the general heterosexual population, the National AIDS (acquired immunodeficiency syndrome) Behavioral Surveys, has obtained data from 10,630 respondents. Data are presented on the prevalence of HIV-related risks in the general heterosexual population, on the distribution of the three largest risk groups across social strata, and on the prevalence and distribution of condom use among heterosexuals reporting a risk factor. Between 15 and 31 percent of heterosexuals nationally and 20 and 41 percent in cities with a high prevalence of AIDS reported an HIV risk factor. Condom use was relatively low. Only 17 percent of those with multiple sexual partners, 12.6 percent of those with risky sexual partners, and 10.8 percent of untested transfusion recipients used condoms all the time. Overall, the results suggest that current HIV prevention programs have, to a very limited extent, reached those heterosexuals with multiple sexual partners but have failed to reach many other groups of the heterosexual population at risk for HIV. New public health strategies may be needed for these specific risk groups.