TITLE:
Feasibility of Using Two versus Three Rapid Tests for HIV Diagnosis in India: Analysis of Public Health Program Data
AUTHORS:
Varsha Kale, Archana Beri, Madhuri Thakar, Veenita Sinha Dar, Shilpa Bembalkar, Naresh Goel, Arun Risbud, Ramesh Paranjape
KEYWORDS:
HIV Testing, Two Test Strategy, Rapid Diagnostic Tests (RDT), National AIDS Control Program (NACP)
JOURNAL NAME:
World Journal of AIDS,
Vol.7 No.1,
January
26,
2017
ABSTRACT: Under India’s National AIDS Control Program (NACP), WHO recommended strategy of using three rapid tests is adopted for diagnosis of HIV in an asymptomatic individual. Since the NACP has a stringent kit evaluation procedure and due to the availability of newer third generation Rapid Diagnostic Tests (RDTs) it may be possible to adapt two test strategy for HIV diagnosis instead of the recommended three test strategy. The authors reviewed programmatic data on HIV testing to explore whether use of two rapid tests for HIV diagnosis as against three tests could be a feasible approach without compromising the quality and readability of testing. Data on the HIV diagnosis performed on serum specimens collected for the period of one year (2011-12) from 82 Integrated Testing and Counselling Centers (ICTCs) or Prevention of Parent to Child Transmission Centers (PPTCTs) associated with State reference Laboratories (SRLs) was analysed. Out of 654258 results that were analysed 25168 (3.84%) specimens were positives. It was observed that both two and three test algorithms provided similar results for majority (>99%) of the HIV positive specimens. Discordance was observed in labelling of specimen with inconclusive status (N = 21), however the true status of these samples could not be obtained. The analysis indicated that the use of two test algorithm will have programmatic benefits in terms of reduced financial burden to the programme and ease of procurement, shipment and storage before and after distribution without comprising the quality of the testing. The prospective study would confirm this observation.