SOME EFFECTS OF THE RISING CASE LOAD OF ADULT HIV-RELATED DISEASE ON A HOSPITAL IN NAIROBI

Increasing numbers of HIV-infected adults in Africa need hospital care . It remains unclear what impact this has on health care services or o n how hospitals respond. The aim of this study was to describe the eff ects of a rising case load of adult HIV-related disease by comparing r esults from a prospective cross-sectional study of acute adult medical admissions to a government hospital in Nairobi conducted in 1992 with results from a previous study done in 1988 and 1989 in the same hospi tal, using the same study design and protocol. Data on age, gender, nu mber admitted, length of stay, HIV status, clinical AIDS, final diagno sis, case mix, and outcome were compared. In 1992, 374 consecutive pat ients were admitted in 15 24-hour periods (24.9 patients/period) compa red with the 1988 to 1989 study, which enrolled 506 patients in 22 24- hour periods (23.0 patients/period). Patients' age, gender, and length of hospital stay were similar in both studies. In 1992, 39% of patien ts were HIV-positive compared with 19% in 1988 to 1989 (p < 10(-6)); w hereas seropositive admissions rose 123% between the two periods (p < .0001), HIV-negative admissions declined 18% (p < .05). Clinical surve illance for AIDS consistently identified <40% of HIV-positive patients . Irrespective of HIV status, tuberculosis and pneumococcal pneumonia were the leading diagnoses in both surveys. No change was found in the diagnoses recorded for HIV-positive patients, but in HIV-negative pat ients, reductions were significant in the case mix (p < .00001) and ra nge of diagnoses (p < .001) seen in 1992. Outcome remained unchanged f or HIV-positive patients with approximately 35% mortality in both surv eys. Outcome significantly worsened, in relative and absolute terms, f or HIV-negative patients: in 1992, mortality was 23%, compared with 13 .9% in 1988 to 1989 (p < .005), with 3.5 deaths per 24-hour period in 1992 compared with 2.6 deaths per 24-hour period in 1988 to 1989 (p < .05, one-tailed). These data suggest that increasing selection for adm ission is taking place as demand for care increases because of HIV/AID S. This process appears to favor HIV-positive patients at the expense of HIV-negative patients who seem to be crowded out and, once admitted , experience higher mortality rates. The true social costs of the HIV epidemic are underestimated by not including the effects on HIV-negati ve people.