Polymerase chain reaction for diagnosis of human immunodeficiency virus infection in infancy in low resource settings

Background: Diagnosis of human immunodeficiency virus (HIV) is essential for accessing treatment. Current HIV diagnostic protocols for infants require adaptation and validation before they can be implemented in the developing world. The timing and type of HIV assays will be dictated by country-specific circumstances and experience from similar settings. The performance of an HIV-1 DNA polymerase chain reaction (PCR) test, and in particular a single test at 6 weeks of age, in diagnosing HIV subtype C infection acquired in utero or peripartum was assessed. Methods: A retrospective review of 1825 Amplicor HIV-1 DNA PCR version 1.5 tests performed between 2000 and 2004 in 2 laboratories in Johannesburg, South Africa on 769 effectively non-breast-fed infants from 3 clinically well characterized cohorts was undertaken. The HIV status of each infant was used as the standard against which the HIV PCR results were compared. Results: The overall sensitivity and specificity of the HIV PCR test were 99.3 and 99.5% respectively. A single test was 98.8% sensitive and 99.4% specific in the 627 infants tested at 6 weeks of age (58 HIV-infected and 569 HIV-uninfected). Repeat testing of all positive HIV PCR tests minimized false positive results. Conclusions: In resource-poor settings where HIV PCR testing in an environment of good laboratory practice is feasible, a single 6-week HIV DNA PCR test can increase identification of HIV-infected children substantially from current levels. Further operational research on how best to implement and monitor such a diagnostic protocol in specific local settings, especially in breast-fed infants, is necessary. Copyright © 2005 by Lippincott Williams & Wilkins.