The changing impact of HIV/AIDS on Kenyatta National Hospital, Nairobi from 1988/89 through 1992 to 1997

Objective: Consequences of the growing HIV/AIDS epidemic for health service s in sub-Saharan Africa remain poorly defined. Longitudinal data from the s ame centre are scarce. We aimed to describe the impact of a rapidly rising HIV/AIDS disease burden on an urban hospital over the last decade. Design and setting: Cross-sectional observational study in 1997, compared t o similar data from 1988/89 and 1992. The study was carried out in the Keny atta National Hospital, Nairobi, Kenya. Method: Consecutive adult medical patients were enrolled on admission and t hen followed up until death or discharge. The main outcome measures were cl inical stage, HIV status, bacteraemia, length of stay, bed occupancy, final diagnosis and outcome of hospital admission. Results: In 1997, 518 patients, 493 with HIV serology, were enrolled: HIV p revalence was 40.0%, bed occupancy 190%, the mean length of stay 9.5 days ( SD 12) and overall mortality 18.5%. The mean number of HIV-positive admissi ons per day steadily rose from 4.3 [95% confidence interval (CI), 0.6] pati ents in 1988/89, through 9.6 (95% CI, 1.4) in 1992, to 13.1 (95% CI, 2.8) o r 13.9 adjusted for those enrolled without HIV serology in 1997. In contras t the mean number admitted with clinical AIDS, 1.7 in 1988/89 and 3.3 in 19 92, fell to 2.6 cases per day in 1997. With HIV-negative admissions increas ing by 37% and bed occupancy nearly doubling in 1997, HIV prevalence appear ed to be stabilizing (19 then 39 and 40% respectively). Over lime fewer HIV -infected patients were bacteraemic (26, 24 and 14%; P < 0.01); had clinica l AIDS (39, 34 and 24% respectively; P < 0.01); or died (36, 35 and 22.6%; P < 0.02). HIV-negative mortality, 14% in 1988/89, rose to 23% in 1992 but fell to 15% in 1997. The mean length of hospital stay (9.5-10 days) did not differ according to HIV status nor did it change across the decade. Conclusion: The HIV/AIDS disease burden in Kenyatta National Hospital medic al wards has risen inexorably over the last decade. Most recently, the numb er of HIV-uninfected patients has also risen, leading to bed occupancy figu res of 190%. Despite overcrowding and irrespective of HIV status, in-patien t mortality has fallen. Time trends suggest fewer clinical AIDS patients ar e presenting for hospital care, implying a rising community burden of chron ic HIV/AIDS disease. Although widely predicted, it is not inevitable that m edical services in urban African hospitals dealing with large volumes of HI V/AIDS disease, will collapse or become overwhelmed with chronic, end-stage disease and death. (C) 2000 Lippincott Williams & Wilkins.