Prevalence and predictors of HIV testing among a probability sample of homeless women in Los Angeles county

Objectives. To describe the prevalence and predictors of HIV testing in a probability cluster sample of urban homeless women. Methods. Analysis of data from the University of California Los Angeles-RAND Access to Health Care for Homeless Women of Reproductive Age Study, a survey conducted in six waves from January 1997 through November 1997 at shelters and soup kitchens in Los Angeles (LA) County, California. The sampling unit consists of homeless woman-visits, and data were collected using structured face-to-face interviews for which respondents were paid $10. Each sampling unit was weighted to take into account the frequency with which the respondent used shelters or meal programs. The main outcome measure was receipt of HIV test in the past year. Results. The response rate was 83%, and the final sample size was N=970. Sixty-eight percent of our sample reported receiving an HIV test in the past year, and 1.6% reported ever being diagnosed with HIV. HIV testing in the past year was most strongly associated with pregnancy in the past year (OR 2.99; p<.001) and having a regular source of care (OR 2.13; p<.001). Approximately 25% of homeless women with indications for HIV testing had not been tested in the past year. Conclusions. The reported HIV seroprevalence of greater than 1% suggests that providers should offer and encourage HIV testing for all homeless women in LA County. Our data, which show a high rate of testing and few statistically significant independent predictors, indicate that this may be what is happening in practice.