Predictive Factors for Positive HIV Test Results in a Hospital Setting

Abstract

Background: Hospital admission for any reason provides the situation for voluntary HIV testing and consultation. Identifying the predictors of positivity may lead to a cost-effective method while enhancing professionalism. Objectives: To find the predictors of HIV-positive test result in a general hospital in Shiraz compared to a control group. Methods: In this case-control study, the records of all patients who received HIV testing upon their hospitalization in a general hospital in Shiraz, south of Iran, from January 2017 to the end of December 2017 were reviewed. For each HIV-positive case, at least one control from the same ward in the hospital with negative HIV test result was randomly selected. Based on the best-fitted model of logistic regression, the probability of positive HIV test results was estimated for each participant according to the risk factors, and a receiver operating characteristic (ROC) curve was drawn. Results: Out of 7333 persons who accepted to be tested, 77 patients tested positive for HIV, of whom 55 (71.4%) were male with the mean age of 41.5 ± 9.5 years. None of the HIV-positive patients were intravenous drug users, nor had they a history of imprisonment. The odds ratio (OR) was 21 for hepatitis-positive patients (hepatitis B and/or C) compared to negative ones, which was seven times higher in opium addicts than non-opium addicts. We developed a model using age, sex, opium addiction, and HBV and HCV status to predict the probability of being positive for HIV with an AUC of 0.853 (95% confidence interval 0.797 to 0.909). Conclusions: Hospital admission could be an appropriate momentum for providing voluntary counseling and testing. Infection with HBV and HCV are important risk factors for HIV infection, and additional testing should be offered, especially to these patients.