Comparison of testing saliva and serum for detection of antibody to human immunodeficiency virus in Jamaica, West Indies

Background: In Jamaica the reported incidence of AIDS increased from 0.1/10 0 000 in 1985 to 20.2/100 000 in 1995. Here there is great reluctance to ha ve voluntary blood testing and, indeed, any blood testing. Since only enzym e-linked immunoassay (ETA) was available for screening serum HIV-1 and 2 an tibody, it was considered that a non-invasive saliva screening EIA could be an advantageous alternative. Objective: this study was designed to evaluat e the OraScreen(TM) HIV Rapid Test, a new, simple saliva screening ETA for anti-HIV-1&2 and to compare its sensitivity and specificity with a standard serum anti-HIV screening EIA in current use in Jamaica. Study design: spec ificity and sensitivity of HIV antibody assays were compared in matched ser um and saliva samples obtained from 257 volunteers from a family planning c linic and from visa applicants, representing a low risk population (Group I ), and from 52 volunteers known to be HIV infected (Group II). Results: in Group I, 357 volunteers of unknown HIV status, one was positive for anti-HI V-1 in both serum and saliva. One other was seropositive but negative on sa liva testing; confirmatory Western Blot (WB) testing on this serum was nega tive and this subject was tabulated as blood HIV negative. Fifty-one of the known seropositive volunteers (Group II) were saliva antibody positive. On e saliva sample was inadequate and this individual was excluded from the st udy. Serum samples from three others in Group were grossly haemolysed but w hose saliva samples were antibody positive. Conclusion: With the exclusion of one subject whose saliva sample was inadequate, the OraScreen(TM) HIV Ra pid Test showed 100% specificity identifying 256/256 HIV antibody negative individuals, and 100% sensitivity by identifying 52/52 infected individuals as HIV antibody positive. (C) 2000 Elsevier Science B.V. All rights reserv ed.