Aim: to examine among high-risk populations or patients with signs or symptoms suggesting HIV-infection, two tests or even one single test might be sufficiently accurate for diagnosis of HIV in a hospital setting in Indonesia.
Methods: we retrospectively examined the rate of false-positive results of initial HIV-tests for all subjects tested in the referral hospital for HIV in West-Java, Indonesia, between 2006 and 2008. We also calculated the positive and negative predictive value of single test results and dual-testing, based on sensitivity and specificity of commonly used methods and prevalence data from Indonesia.
Results: among 3121 subjects, 803 were tested positive (25.7%). The initial rapid HIV-tests did not show a single false positive result, and no discrepancy was found between the second and third supplemental tests. Based on their high accuracy, most rapid tests carry a low risk of false-positive results among risk groups. Dual testing algorithms almost eliminate the risk of false-positive HIV-results, and are probably as accurate as three tests, even in low prevalence settings.
Conclusion: based on expected prevalence rates and the accuracy of methods used in Indonesia, one or two tests are usually accurate for HIV-diagnosis, especially for high risk populations. The possible implications and optimal conditions for more simple testing algorithms warrant further investigation.