Background:  The HIV epidemic has significantly affected the death rates in southern Africa. South African hospitals have reported a large increase in patient admissions and deaths due to HIV. However, detailed descriptions thereof are lacking.

Methods:  Retrospective record review of a mortuary death register at a Johannesburg hospital. The recorded causes of death over 4 years (2000 to 2003) of 3 selected months (March, July and September) were analysed for demographic and clinical data.

Results:  A total of 2318 deaths were analysed, 59 of which had ‘unknown’ as the diagnosis. From 2000 to 2003 the overall death rate increased by 24.3%. Of all deaths recorded, 47% (n = 1047/2228) occurred in the age group 15 to 45 years and this trend was reflected in all months analysed.  The winter month (July) claimed 31% more deaths than the other months; 53% of deaths occurred in males and 47% occurred in females. Average age at death amongst both males and females was 50 years. There was no difference in age at death between males and females below and above the age of 45 years.  47% of all patients (females 23.5%, males 23.5%) died before the age of 45 years. The 4 main causes of natural death were:  respiratory causes (36.5%) of which 82.2% were infections (bacterial pneumonia, pulmonary TB, PCP); central nervous system pathologies (16.5%) of which 37.3% were meningitis; cardiac causes (14.2%); and gastrointestinal causes (10.2%) of which 38% were gastroenteritis. Infectious causes of death accounted for 44% (n = 1025) of all deaths. 68.2% (n = 699) occurred in the age group 15 to 45 years with a 32.4% increase over the 4 years. Clinical AIDS diagnoses accounted for 34% (n = 348) of the infectious causes. HIV serologies and CD4 counts are being analysed, and will be presented.

INSERT GRAPHS

Conclusions:  In just 4 years, there was a significant increase in the death rate in a South African urban hospital. The deaths showed a strong seasonal variation. The predominant causes of death were infectious in nature, and this proportion increased over the 4 year period. There were no gender differences regarding age at death, which differs with South African seroprevalence studies showing high infection rates amongst young women. Recorded AIDS diagnoses accounted for 34% of the infectious causes. It is suspected that this is an underestimation as death records frequently omit this diagnosis.

Keywords: death; hospital; Africa