Field evaluation of near point of care Cepheid GeneXpert HIV-1 Qual for early infant diagnosis

PLoS One. 2018 Dec 27;13(12):e0209778. doi: 10.1371/journal.pone.0209778. eCollection 2018.

Abstract

Background: Access to point-of-care HIV testing shortens turn-around times, time to diagnosis and reduces loss to follow-up hence minimizing barriers to early linkage to care and treatment among HIV infected infants. Currently samples for early infant HIV diagnosis are sent to centralized testing facilities which are few and located only at specific regions in Kenya. However, there are Point of Care (POC) early infant diagnosis [EID] technologies elsewhere such as SAMBA and ALERE-Q that are yet to be evaluated in Kenya despite the urgent need for data to inform policy formulation regarding EID. The Cepheid GeneXpert HIV-1 Qual (GeneXpert) technology for POC EID offers a great opportunity to minimize HIV associated morbidity, mortality and loss to follow-up through decentralization of early infant HIV testing to the clinics. This technology also allows for same-day results thus facilitating prompt linkage to care.

Methods: We evaluated the GeneXpert HIV Qual EID POC in Homabay County against the standard of care platform, Roche CAP/CTM HIV-1 qualitative PCR, using dried blood spots (DBS). Between February-July 2016, DBS samples were collected from HIV exposed children <18 months of age enrolled in a cross-sectional study. Samples were collected by qualified nurse counselors, and were tested by trained technicians using field based GeneXpert and conventional laboratory based Roche CAP/CTM HIV-1 qualitative PCR. Sensitivity and specificity were determined.

Results: Overall, 3,814 mother/infant pairs were included in the study, out of which 921 infants were HIV exposed as per the mothers' HIV status and based on the infant's HIV rapid test. A total of 969 PCR tests were performed, out of which 30 (3.3%) infants were concordantly positive using both platforms. GeneXpert HIV-1 Qual yielded a sensitivity of 94.1% and specificity of 99.8% with an overall error rate of 0.7%.

Conclusion: Our findings show that GeneXpert HIV-1 Qual performs well compared to CAP/CTM using DBS samples, suggesting that this technology may be adopted in decentralized laboratories as a near POC device. It may contribute to prompt diagnosis of HIV exposed infants hence enabling early linkage to care, thus advancing further gains in EID.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Diagnostic Tests, Routine / methods*
  • Dried Blood Spot Testing
  • Early Diagnosis
  • Female
  • HIV Infections / diagnosis*
  • HIV Infections / genetics
  • Humans
  • Infant, Newborn
  • Kenya
  • Male
  • Mothers / statistics & numerical data
  • Point-of-Care Systems / statistics & numerical data*
  • Polymerase Chain Reaction
  • Sensitivity and Specificity

Grants and funding

This work was funded and sponsored by Medecins Sans Frontières, 8 rue Saint Sabin 75010 Paris France to DM. The study was also co-funded by UNITAID who provided the Cepheid GeneXpert instruments and cartridges.