The clinical case definition for AIDS proposed by WHO for use in Africa was evaluated against HIV antibody status in 72 patients in rural Zaire. Twenty-one (29%) of the patients were antibody-positive. For diagnosing anti-HIV seropositivity, the case definition had a sensitivity of 52%, a specificity of 78%, a positive predictive value of 50% and a negative predictive value of 80%. Calculation of the positive predictive value at different levels of prevalence of HIV infection suggests that the case definition operates at maximum reliability in selected high-risk groups. Modifications to the case definition should be evaluated to try and improve its sensitivity and positive predictive value.
PIP: In many areas of Africa where AIDS is endemic, facilities for laboratory diagnosis are too limited to reliably diagnose opportunistic infections. Therefore, the World Health Organization defined a clinical case definition of AIDS in which 2 major signs and at least 1 minor sign must be present to diagnose AIDS. The major signs are: weight loss greater than 10%, diarrhea for more than 1 month, and prolonged fever for more than 1 month. The minor signs are: persistent cough for more than 1 month, generalized pruritic dermatitis, recurrent herpes zoster, oropharyngeal candidiasis, chronic disseminated herpes simplex, and generalized lymphadenopathy. (The presence of Kaposi's sarcoma or cryptococcal meningitis are sufficient by themselves for a diagnosis of AIDS.) 72 patients in 4 hospitals in Equateur Province of Zaire were used to test the reliability of the clinical case definition. 21 (29%) of the patients were HIV seropositive, and 22 (32%) fulfilled the clinical criteria. From these data the sensitivity of the case definition was 52%, specificity was 78%, positive predictive value was 50%, and negative predictive value was 80%. Since positive predictive value rises with prevalence and HIV infection is maximal in the 20-40 age group, restricting the case definition to this age group would increase its predictive value. Exclusion of patients with tuberculosis would reduce the number of false positive results.