Clinician judgement as a tool for targeting HIV counseling and testing in North Carolina state mental hospitals, 1994

HIV infection increasingly affects populations that may not appear at high risk based on the use of some traditional targeting strategies. To shed some light on how to more sensitively/effectively identify people who need routine HIV testing and counseling, the objective of this study is to determine the prevalence of HIV infection in North Carolina state mental hospitals and to evaluate clinician judgment as a tool for targeting HIV counseling and testing. The design used is a blinded seroprevalence study. The study population includes all patients admitted to North Carolina state mental hospitals between March 1st and May 31st, 1994. The main outcome measures are the HIV seroprevalence, demographic and diagnostic features, and clinician assessment of the likelihood of HIV infection. The results of the study find that of 2159 study subjects, 35 persons (1.6%) were infected with HIV; of these, 14 (40%) were not previously known to be infected. All 35 HIV infections occurred in persons aged 13-59 years. Within this age group, infection rates were significantly higher for Blacks, males, persons who had a diagnosis of organic brain disease, and persons who had multiple psychiatric diagnoses. However, testing strategies that targeted any of the higher risk groups were insensitive. The rate of HIV infection for persons judged by the admitting clinician to have a high or intermediate likelihood of HIV infection was 26.4 times higher than the rate for those judged to have a low likelihood of infection (2.1 vs. 0.1%, 95% confidence intervals: 3.5-201.3). Of the 14 previously undiagnosed HIV-infected persons, 13 were judged by clinicians to have a high or intermediate likelihood of HIV infection. Moreover, 1258 persons were correctly assessed to have a low likelihood of infection. Conclusions from this study are that an HIV counseling and testing strategy targeting persons (in this setting aged 13-59 years) who were judged by clinicians to have a high or intermediate likelihood of infection, would have identified more than 90% of previously undetected infections while substantially reducing the number of negative HIV tests performed.