Laboratory evaluation of the Alere q point-of-care system for early infant HIV diagnosis

Hsiao, Nei-yuan; Dunning, Lorna; Kroon, Max; Myer, Landon

Laboratory evaluation of the Alere q point-of-care system for early infant HIV diagnosis

Journal Article

2016

Journal Title

PLoS One

Link to Journal

http://journals.plos.org/plosone

Publisher

Public Library of Science

Publisher

University of Cape Town

Department

Division of Virology

Faculty

Faculty of Health Sciences

License

http://creativecommons.org/licenses/by/4.0

Abstract

Introduction Early infant diagnosis (EID) and prompt linkage to care are critical to minimise the high morbidity and mortality associated with infant HIV infection. Attrition in the "EID cascade" is common; however, point-of-care (POC) EID assays with same-day result could facilitate prompt linkage of HIV-infected infant to treatment. Despite a number of POC EID assays in development, few have been independently evaluated and data on new technologies are urgently needed to inform policy. METHODS: We compared Alere q 1/2 Detect POC system laboratory test characteristics with the local standard of care (SOC), Roche CAP/CTM HIV-1 qualitative PCR in an independent laboratory-based evaluation in Cape Town, South Africa. Routinely EID samples collected between November 2013 and September 2014 were each tested by both SOC and POC systems. Repeat testing was done to troubleshoot any discrepancy between POC and SOC results. RESULTS: Overall, 1098 children with a median age of 47 days (IQR, 42-117) were included. Birth PCR (age <7 days) comprised of 8% (n = 92) tests while 56% (n = 620) of children tested as part of routine EID (ages 6-14 weeks). In the overall direct comparison, Alere q Detect achieved sensitivity of 95.5% (95% CI, 91.7-97.9%) and a specificity of 99.8% (95% CI, 99.1-100%). Following repeat testing of discordant samples and exclusion of any inconclusive results, the POC assay sensitivity and specificity were 96.9% (95% CI 93.4-98.9%) and 100% (lower 95% CI 98%) respectively. Among birth PCR tests the POC assay had slightly lower sensitivity (93.3% vs 96.5% in routine EID) and higher assay error rate (10% vs 5% in samples of older children, p = 0.04). CONCLUSION: Our results indicate this POC assay performs well for EID in the laboratory. The high specificity and thus high positive predictive value would suggest a positive POC result may be adequate for immediate infant ART initiation. While POC testing for EID may have particular utility for birth testing at delivery facilities, the lower sensitivity and error rate requires further attention, as does field implementation of POC EID technologies in other clinical care settings.