Adult mortality in the cities of Bulawayo and Harare, Zimbabwe: 1979-2008

BACKGROUND: Zimbabwe has been severely affected by the HIV/AIDS and tuberculosis epidemics, with an estimated 80% of tuberculosis patients being HIV infected. We set out to use annual population-mortality records from the cities of Harare and Bulawayo to describe trends and possible causes of mortality from 1979 to 2008. The specific objectives were to document overall, sex and age-specific mortality, proportion of deaths attributed to AIDS and tuberculosis, and changes in death rates since the start of antiretroviral therapy in 2004. METHODS: This retrospective descriptive study used existing mortality records of the Health Services departments in Harare and Bulawayo. Data points included: estimated yearly total population; groupings by sex and age; deaths (total and by sex and age groups for each year of the study period); and most frequently reported causes of death (for age groups <15 years, 15-44 years and ≥45 years). Data on deaths were aggregated by year, and crude, sex- and age-specific death rates were calculated per 1000 population. Tuberculosis and HIV-related disease-specific death rates and proportion of deaths attributed to these conditions were computed. RESULTS: In both cities, crude death rates were lowest in the late 1980s, increased three- to five-fold by the early 2000s, and began a slow and, in the case of Bulawayo, intermittent decline from 2004. Sex-specific death rates followed a similar trend, being higher in males than in females. The death rates in the age groups <5 years, 15-44 years and ≥45 years showed significant increases, with a gradual levelling off and decline from 2002 onwards; death rates in those aged 5-14 years were relatively unaffected. Tuberculosis and HIV caused 70% of deaths in the age group of 15-44 years from the early 1990s. CONCLUSIONS: This study used routinely collected population-mortality data that are rare in resource-limited settings, and it described, for the first time in Zimbabwe, the effects of the HIV/AIDS epidemic and the introduction of antiretroviral therapy on death rates in two large cities. After a substantial rise in crude mortality rates, there has been a decline associated with the introduction of ART. Such routine population data must continue to be collected, collated and analyzed.