PloS one 2017 01 0612(1) e0169497 doi 10.1371/journal.pone.0169497
The long delay in returning test results during early infant diagnosis of HIV (EID) often causes loss-to-follow-up prior to antiretroviral treatment (ART) initiation in resource-limited settings. A point-of-care (POC) test may help overcome these challenges. We evaluated the performance of the LYNX p24 Antigen POC test in Mozambique. 879 HIV-exposed infants under 18 months of age were enrolled consecutively at three primary healthcare clinics (PHC). Lancet heel-drawn blood was tested on-site by nurses using a prototype POC test for HIV Gag p24 antigen detection. Results of POC testing were compared to laboratory-based nucleic acid testing on dried blood spots. A comparison of the effect of sensitivity and timely test results return on successful diagnosis by POC and laboratory-based platforms was also calculated. The sensitivity and specificity of the LYNX p24 Ag test were 71.9%; (95% confidence interval [CI]: 58.5-83.0%) and 99.6% (95% CI: 98.9-99.9%), respectively. The predictive value of positive and negative tests were 93.2% (95% CI: 81.3-98.6%) and 97.9% (95% CI: 96.8-98.8%), respectively. Overall agreement was high (Cohen Kappa = 0.80; 95% CI: 0.71-0.89). Despite its lower sensitivity, the POC test had the potential to provide test results to up to 81% more patients compared to the laboratory-based test. This prototype POC p24 assay was feasible for use in PHCs but demonstrated low sensitivity for HIV detection. POC EID technologies that perform below standard recommendations may still be valuable diagnostic tools in settings with inefficient EID networks.