FACTORS ASSOCIATED WITH UNRECOGNIZED HIV-1 INFECTION IN AN INNER-CITYEMERGENCY DEPARTMENT

Study objective: To determine the prevalence of and risk factors assoc iated with unrecognized HIV-1 infection among medical patients present ing to an inner-city emergency department. Methods: We conducted anony mous HIV-1 testing in subjects interviewed for risk behaviors and know ledge of HIV status at an inner-city ED in the Bronx, New York. Our su bjects were consecutive adult medical patients in noncritical conditio n (N=1,744) who were evaluated by three physicians providing primary e mergency care. Each patient was given a structured interview for demog raphic characteristics, risk behaviors, and knowledge of HIV status. E xcess serum, drawn for clinical purposes, was linked without identifie rs to responses and tested for antibodies to HIV-1. In subjects who de nied HIV infection, we tested associations with seropositivity using u nivariate analyses and logistic-regression techniques (multivariate). Results: Of the 1,744 patients interviewed, 656 (37.6%) reported HIV r isk behaviors. Of 970 tested for HIV-1 antibodies, 125 (12.9%) were se ropositive. The prevalence of HIV-1 infection among those who denied k nown infection was 4.0% (35 of 875). In the multivariate model, indepe ndent predictors of unrecognized HIV-1 infection were age 35 to 44 yea rs, crack cocaine use, history of syphilis, and ED diagnosis of an inf ection not necessarily related to HIV infection. Unrecognized HIV-1 in fection was more likely among patients admitted to the hospital, but 2 1 of the 35 with unrecognized infection (60%) were not admitted and in 9 (25.7%) no risk factors were identified. Conclusion: More than one third of patients who visited one inner-city ED acknowledged HIV risk behaviors. One quarter of patients with unrecognized HIV-1 infection r eported no identifiable risk factors. Easily accessible HIV counseling and testing should be considered in EDs in areas serving persons at r isk for HIV infection.