Performance of HIV Rapid Tests Among Breastfeeding, Malawian Infants.

INTRODUCTION: Timely, accurate and affordable testing algorithms at point-of-care are critical for early infant HIV diagnosis and initiation of antiretroviral therapy in the postpartum period. We aimed to assess the utility of HIV rapid tests for young, breast-fed HIV-exposed infants in resource limited, high HIV burden settings. MATERIALS AND METHODS: We collected data on the performance of 2 commonly used rapid tests (Determine and Unigold) in Malawi between 2008 and 2012 or at the University of North Carolina between 2014 and 2015. For each 3-month interval between ages 3 and 18 months, we calculated the sensitivity, specificity, positive and negative predictive values of each test compared with the HIV DNA/RNA PCR gold standard. We also assessed the utility of each rapid test to diagnose incident HIV infection during the breastfeeding period. RESULTS: Among 121 HIV-exposed infants who were negative at age 6 weeks, 21 (17.2%) became infected by 18 months. At 3 months of age, both rapid tests had minimal clinical value with specificity values of 7.0% [95% confidence interval (CI): 2.3-15.7] for Determine and 19.4% (95% CI: 11.1-30.5) for Unigold. Starting at age 6 and 9 months, the Unigold test could be used as a screening tool in the follow-up of HIV-exposed infants with specificity values of 83.7% (95% CI: 74.4-89.9) and 97.7% (95% CI: 94.6-99.7), respectively. Starting at age 12 months, the type of test became less important as both tests performed well in identifying HIV-free children, although both tests failed to detect some incident HIV infections. CONCLUSIONS: Updated guidelines for the use of rapid tests in young HIV-exposed children that explicitly take type of test and infant age into account are urgently needed to ensure optimal care for the 1.5 million HIV-exposed infants born annually.

Duke Scholars

Author Emily R Smith Emergency Medicine