Abstract

Infections with human immunodeficiency virus are common in areas of the world where laboratory testing and sophisticated diagnostic facilities are unavailable. A World Health Organization clinical case definition for acquired immunodeficiency syndrome was developed in 1985 for use in such areas. In 1987, we tested this definition on 1328 inpatients and outpatients in 15 hospitals throughout Uganda. Five hundred sixty-two patients (42%) were positive by enzyme-linked immunosorbent assay for human immunodeficiency virus antibody. The World Health Organization definition had a sensitivity of 55%, a specificity of 85%, and a positive predictive value of 73%. Modification of the case definition by excluding a known cough from tuberculosis as a minor criteria decreased sensitivity slightly to 52%, but specificity and positive predictive value increased to 92% and 83%, respectively. Amenorrhea, although not specifically asked about, was a symptom noted by many female patients (26% of females who were positive by enzyme-linked immunosorbent assay); as a symptom indicative of human immunodeficiency virus infection, amenorrhea had a specificity of 99%, with a positive predictive value of 89%. These findings support the generalizability of the World Health Organization clinical acquired immunodeficiency syndrome definition and its use (especially the modified version) in areas of Uganda without sophisticated facilities.