Background: Perinatal exposure of infants in low resource settings generates the bulk of pediatric human immunodeficiency virus (HIV) disease globally. The HIV status of these infants is established by testing serum for anti-HIV antibodies at 12 months of age in Prevention of Mother to Child Transmission (PMTCT) programs because polymerase chain reaction testing is unavailable. The diagnostic accuracy of 2 oral fluid (OF) HIV tests has not been previously evaluated in children.
Methods: A serum and 2 OF HIV tests were performed at 12 months of age in a cohort of 321 vertically exposed children in a prospective, longitudinal study at a secondary level hospital in Johannesburg, South Africa during a 14-month period preceding October 2003. The 3 HIV tests were performed independently of each other by personnel blinded to the child's true HIV infection status, the reference standard used for comparison.
Results: HIV testing was performed at a median age of 12.1 months. The true HIV infection status of 310 of 321 (97%) children was determined. In comparison with serum testing results, OF HIV tests reduced the percentage of children requiring repeat HIV tests from 45% to 8-12%. The abilities of OF and serum to predict an HIV-uninfected status were comparable with negative predictive values >99%. Interpretation of HIV tests in conjunction with simple clinical assessment further improved the predictive value of the test.
Conclusions: OF HIV tests perform well in children and have the potential to increase accessibility and acceptability of HIV diagnosis for infants in the context of PMTCT programs in low resource settings.