The goal of this study was to assess which factors are associated with access to antiretroviral therapy, antiviral therapy, antifungal therapy, and PCP prophylaxis in a large, multicenter AIDS cohort of gay and bisexual men (MACS).

MACS is a cohort established in 1984-1985 to study the natural history of HIV-1 infection in gay and bisexual men from four large U.S. cities. Approximately one-third of the sample was HIV-negative (36.5% of the sample). From this paper, it is uncertain how this cohort represents the larger U.S. population of gay and bisexual men with regard to demographic or illness characteristics. One of the main limitations of this study is that the cohort is a self-selected group of gay and bisexual men with relatively high SES. With regard to representativeness, it is important to note that between 1987 and 1991 an additional sub-sample of men was recruited with the intention of increasing minority representation.

This analysis focuses on the 1,415 HIV seropositive men who returned for at least one of the three study interviews from October 1990 to March 1992.

Overall, the use of antiretroviral therapy (zidovudine, didanosine and dideoxycytidine) was most common among patients with AIDS and among HIV-positive patients without AIDS who had low (200-500) CD4 counts. Zidovudine was used by more than 60% of patients in these groups at all follow-up three visits. Didanosine use and, particularly dideoxcytidine use, increased over the three visits to include more than 20% of the patients with AIDS and almost 10% of the HIV-positive patients without AIDS who had low CD4 counts. Combination antiretroviral therapy (two or more of these medication taken concurrently) increased in these groups over the study period as well. Approximately 88% of the men with AIDS took at least one PCP prophylaxis medication by the last visit, as did 75% of the HIV-positive men without AIDS who had low CD4 counts.

After adjusting for severity of illness and overall health services utilization, having health insurance (OR=1.32), a college education (OR=1.42) and being white (OR=1.58) were associated with increased odds of using antiretrovirals among patients who did not have AIDS but were HIV-positive. The sociodemographic variables were not associated with antiretroviral use among patients with AIDS. With regard to PCP prophylaxis, after adjustment for severity and health services utilization, only having a college education was significantly associated with medication use; none of the sociodemographic variables were associated with PCP prophylaxis use among the patients with AIDS.

It is important to note that after adjustment for access to care and insurance status, non-whites were less likely to use antiretroviral drugs. The authors did not suggest explanations, but noted that "factors such as personal choice, social and cultural norms or discriminatory practices were not ascertained in the MACS cohort," and are presumably potential explanatory factors.