Thetesting ofsaliva samples forHIV-1 antibodies: reliability inanon-clinic setting

Aims-Toassess thereliability ofsaliva samples asa meansoftesting forHIVantibodies outside clinic settings. Methods-Mentakingpartin a nonclinic longitudinal studyofhomosexually active men provided samplesofsaliva andblood.Serawerescreened usinga competitive ELISA (Welicozyme) and positive serawereconfirmed byanindirectELISA (Abbott). Salivasamples werescreened either using an IgGcaptiveradioimmunoassay or an amplified ELISA. total of534paired saliva and bloodsamples weretested. Overall sensitivity was96.2%andspecificity was100%. Noneofthesaliva tests werefalsely positiveforHIV-1antibodies. Conclusions-HIV-l saliva tests canreliablybeusedinanon-clinic orfield setting. However, ifresults aretobegiven torespondents, itisnecessary tooffer adequate counselling andconsider the mechanismsforreferral andfollow-up forthosethatarefoundtobeHIV-1antibodypositive.