Abstract

The broad genetic diversity of HIV requires regular reassessment of adopted HIV diagnostic tests. The objective of this study was to reassess the performances of Determine^® HIV 1/2 and Uni-gold^® HIV1/2, 2 HIV rapid diagnostic tests of the national algorithm, adopted since 2005 in Central African Republic. A reference prospective panel of 150 plasma samples were tested in duplicate with Determine^® HIV 1/2 and Uni-gold^® HIV, according to reference serological immune-enzymatic method. Analytical and virological characteristics were calculated. Sensitivity, specificity, positive and negative predictive values (PPV, NPV) of Determine^® HIV 1/2 were 100%, as well as the specificity and NPV of Uni-gold^® HIV 1/2. Uni-gold^® HIV 1/2 sensitivity and PPV were 96 and 92%, respectively. The Cohen k coefficient was close to 1 for the 2 tests, 1 for Determine^® HIV 1/2 and 0.97 for Uni-gold^® HIV 1/2. Except for the sensitivity of Uni-gold^® HIV 1/2, the performances of 2 tests were acceptable and in perfect agreement with the reference tests. These results require a revision of the algorithm. In addition, an exploration in molecular biology is indicated to determine the subtypes of the HIV strains of the 4 samples that tested false negative with Uni-gold^® HIV 1/2.

Key words: HIV, rapid diagnosis test, sensitivity, specificity, Africa, Central African Republic.