Detection of salivary HIV-1-specific IgG antibodies in high-risk populations in Zaire.

Saliva and blood samples were tested for human immunodeficiency virus-1 (HIV-1) antibodies in two high-risk populations in Kinshasa, Zaire. In a seroprevalence study of 458 sexually transmitted disease (STD) clinic attendees, 142 of 145 seropositive individuals had enzyme-linked immunosorbent assay (ELISA)-positive saliva samples (97.9% sensitivity). All saliva samples from seronegative patients were ELISA-negative (100% specificity). Of the 142 ELISA-positive saliva specimens, 137 were also Western blot-positive (94.5% sensitivity). In a subsequent seroincidence study of 315 initially seronegative female prostitutes followed during 183 woman-years of observation, 9 of 14 women who seroconverted (7.7% seroincidence) had ELISA-positive saliva samples at the time seroconversion was detected. Only three of these saliva specimens could be confirmed by Western blot. Although salivary testing for HIV-1 antibodies using conventional assays was not sensitive in detecting recent seroconversions, screening of salivary samples for HIV-1 antibody provides a convenient alternative method for conducting seroprevalence surveys in populations in whom venipuncture is not possible or convenient.

Successful use of pooled sera to determine HIV-1 seroprevalence in Zaire with development of cost-efficiency models.

To determine the accuracy and cost efficiency of pooling sera prior to HIV-1 testing, sera from 8,000 Kinshasa factory workers and their spouses were screened individually (2.44% seropositive) and in 800 pools of 10 sera each. There were no false-negative or false-positive pools, resulting in a calculated seroprevalence estimate of 2.42%. Further testing of all sera in positive pools can identify HIV-positive individuals. These applications were modeled to compare the cost-efficiency of pooling with individual testing under different conditions. The results suggest that pooling provides an alternative test format for use in both developing and industrialized countries when the seroprevalence and/or the marginal cost of obtaining a sample are sufficiently low. For our cohort, testing only the pools for seroprevalence estimation resulted in a 78% cost saving compared with individual testing; pooling with subsequent identification of individual seropositives represented a 56% cost reduction.