Background: More routine laboratories in South Africa are equipped to perform quantitative than qualitative HIV viral detection assays. The accessibility of early infant diagnosis would be improved if a quantitative viral load (VL) assay performed on dried blood spots (DBS) could accurately diagnose HIV-infection. The VL assay routinely used in the country has not previously been assessed on DBS for early infant diagnosis.
Objectives: To determine the accuracy of the NucliSens EasyQ assay (bioMerieux, Lyon, France) on DBS for early infant diagnosis of HIV in a subtype C-infected population.
Study design: Stored DBS samples collected from children presenting for HIV testing were analyzed. DBS EasyQ VL results were compared to the child's HIV status as determined by a whole blood HIV DNA PCR result.
Results: The EasyQ assay was performed on 235 stored DBS samples from 71 HIV-infected and 164 HIV-uninfected children. Of the 216 infants (children aged 12 months or less) tested, all 52 HIV-infected infants were detected (sensitivity of 100%). Six of 164 HIV-uninfected infants yielded false positive results (specificity 96.3%). All false positive and six of the true positive infants had VL<3.7 log IU/ml. These 12 (5.6%) infants would require repeat testing to differentiate true from false positives. Using a threshold above 3.7 log IU/ml (equivalent to 4 log copies/ml) to define a positive result would yield an accurate diagnosis in 204 (94.4%) infants.
Conclusions: DBS EasyQ VL assays provide an accurate option for early infant diagnosis in low resource settings.
Copyright 2010 Elsevier B.V. All rights reserved.