Performance of Virological Testing for Early Infant Diagnosis: A Systematic Review

Background: Improved access to both maternal antiretroviral therapy and infant prophylaxis may have an impact on the performance of virological assays for diagnosis of HIV infection in infants. This systematic review was performed to assess the diagnostic accuracy of virological testing at birth as well as the performance of virological testing on dried blood spots at 6 weeks among HIV- and antiretroviral (ARV)-exposed infants. Methods: A systematic review was performed for studies published between January 1, 2009 and January 30, 2015. The search strategy included studies related to HIV, nucleic acid amplification tests, and newborns/infants and queried PubMed, Embase, the Cochrane Library, LILACS as well as several conference proceedings. Two independent reviewers collected studies and extracted data. The final analysis includes summary estimates of the sensitivities and specificities of the virological assays assessed. The GRADE approach was used to assess the overall quality of evidence and Quality Assessment of Diagnostic Accuracy Studies was used to evaluate the risk of bias in the studies. Results: A total of 2243 records were screened with a final selection of 5 manuscripts. To assess the test characteristics of virological testing at birth, 2 studies were used to calculate a pooled sensitivity of 69.3% (95% confidence interval: 61.1 to 77.4) and a specificity of 99.9% (98.6-100%). The quality of evidence to support the sensitivity of assays at birth was low, whereas the quality of evidence for the specificity of such tests was intermediate-high. In terms of the performance of virological testing on dried blood spots for HIV- and ARV-exposed infants, 3 studies were used to calculate a pooled sensitivity of 99.4% (98.3-100.00%) and specificity of 99.6% (99.1-100.00%). The quality of evidence for both outcomes was low. Conclusion: The performance of polymerase chain reaction at birth demonstrates low sensitivity and high specificity, reflecting the difficulty of detecting intrapartum infections at birth and transmission dynamics. In addition, there is no evidence to suggest poor performance of virological testing on dried blood spots for ARV-exposed infants. Overall, given the very limited and low-quality evidence, further research is needed to assess the accuracy of polymerase chain reaction at different time points and in the context of more effective prevention of mother-to-child transmission interventions.