OBJECTIVE: To determine the prevalence and risk factors for HIV and Tb infection (PPD+) in homeless adults, and to investigate the relationship between these two infections in this high risk population. METHODS: This population-based cross-sectional study has enrolled 916 subjects from public shelters and free meal lines throughout SF. Since January 1991, 697 subjects have had a PPD skin test in addition to an interview and a serum HIV test. RESULTS: Tb and HIV are common infections in this population: 28% were PPD+ and 10% were HIV+. There is substantial overlap between these two infections, 20% of the HIV+'s are also PPD+. The actual degree of overlap may be much higher than this given the problem of HIV-induced anergy. Most subjects were unaware of either their TB infection (80%) or HIV infection (85%). Even fewer had received prior medical evaluation for these infections. The traditional risk factors of intravenous drug use (IVDU) and male homosexual contact accounted for a majority of the HIV infection in this population: 88% in males and 75% in females. Homeless IVDU gay men were at particular high risk, 70% were HIV+. Even in the low risk homeless (i.e., non-IVDU heterosexuals), the HIV rate was relatively high at 2-3%. The HIV prevalence rates by race were unexpected: whites 11%, blacks 9%, Hispanics 9%. This racial pattern was reversed for PPD+: whites 16%, blacks 35%, Hispanics 48%. Time homeless did not distinguish HIV status but did predict PPD status: median time homeless was 12 months in both HIV+ & HIV-, whereas, PPD+'s had a median time homeless of 16 months compared to eight months for the PPD-'S. Persons living in public shelters had over twice the rate of PPD+ compared to persons living on the streets( 32% vs. 14%). Psychiatric conditions, crack cocaine use and alcoholism, although common features of this population, did not appear to be risk factors for either HIV or TB infection. CONCLUSIONS: The homeless population of SF has relatively high rates of HIV and TB infection with many persons doubly infected. Given the combination of poverty, crowded living conditions, high risk behaviors and lack of medical services, this population is likely to suffer increasingly from these communicable diseases in the 1990's. These data call for a comprehensive approach which addresses all of these features of homelessness. Piecemeal programs, for example providing public shelters without adequate medical services, may in fact worsen the problem of communicable disease's in this population.
Journal