Impact of the 1993 revision of the AIDS case definition on the prevalence of AIDS in a clinical setting

Objective: To determine the impact of the 1993 revision of the Centers for Disease Control and Prevention (CDC) AIDS surveillance case definition on the prevalence of AIDS. Design: Review of prospectively collected baseline clinical and demographic data on HIV-infected patients presenting for care between December 1988 and May 1991. Setting: The HIV Clinic of the Johns Hopkins Hospital, an urban, primary care institution. Main outcome measure: Diagnosis of AIDS by the 1987 (specific indicator diseases) or the 1993 (indicator diseases, pulmonary tuberculosis, recurrent bacterial pneumonia, cervical carcinoma, or CD4 lymphocyte count < 200 × 10⁶/l) CDC case definition. Results: Of 955 patients evaluated, 122 (13%) had AIDS by the 1987 case definition at presentation. An additional 126 (13%) met the 1993 but not the 1987 case definition. Patients meeting only the 1993 case definition were more likely to be female [28 versus 14%; odds ratio (OR), 2.4; 95% confidence interval (Cl), 1.2-3.0; P = 0.01] and intravenous drug users (40 versus 26%; OR, 2.0; 95% Cl, 1.1-3.3; P = 0.02) than patients meeting the 1987 case definition. Fifty-five per cent of patients meeting only the 1993 case definition were asymptomatic, and 7% (nine patients) had new indicator diseases but CD4 counts > 200 × 10⁶/l. Median time to progression from a diagnosis of AIDS by the 1993 case definition to diagnosis by the 1987 case definition was 435 days. Patients with AIDS by the 1987 case definition had a median survival of 594 days from presentation (2-year survival, 42%), while median survival time for patients with AIDS by the 1993 case definition only was 947 days (2-year survival, 60%; P < 0.005). Conclusions: The proposed 1993 revision of the AIDS surveillance case definition would double the number of prevalent AIDS cases, with significant increases in the proportion of cases who are female, intravenous drug users, or asymptomatic. Survival of patients meeting the 1993 case definition is significantly longer than that of patients meeting the 1987 case definition. The new AIDS case definition will have a major impact both on AIDS surveillance and on medical and social service programs that use diagnosis of AIDS as a criterion for eligibility for services.