Patterns of mortality in public and private hospitals of Addis Ababa, Ethiopia

Background: Ethiopia is encountering a growing burden of non-communicable diseases along with infectious diseases, perinatal and nutritional problems that have long been considered major problems of public health importance. This retrospective analysis was carried out to examine the mortality patterns from communicable diseases and non communicable diseases in public and private hospitals of Addis Ababa. Methods: Approximately 47,153 deaths were captured over eight years (2002-2010) in forty three public and private hospitals of Addis Ababa, Ethiopia. Data collectors (43 hospital clerks) and coordinators (3 nurses) had been extensively trained on how to review hospital death records. Information obtained included: dates of admission and death, age, sex, address, and principal cause of death. Only the diseases responsible for deaths are taken as the cause of death. Cause of death was coded using International Classification of Diseases (ICD-10) and data were double entered. Diseases were classified into: Group I (communicable diseases, maternal conditions and nutritional deficiencies); Group II (non-communicable causes); and Group III (injuries). Percentages, proportional mortality ratios, 95% confidence intervals (CI) and Adjusted odd ratios (OR) were calculated. Results: Overall, 59% of the deaths were attributed to Group I diseases, and 31% to Group II diseases and 12% to injuries. Nearly 56% of the males and 68% of the females deaths were due to five leading causes (conditions arising during perinatal period, HIV/AIDS, tuberculosis, cardiovascular diseases and respiratory infections). Significantly larger proportions of females died from Group I (67%) and Group II diseases (32%) compared with males (where the respective proportions were 52% and 30%). Significantly higher proportion of males (17%) than females (6%) were dying from Group III diseases. Deaths due to Group I diseases decreased while those due to Group II diseases increased with age. Overall Group I diseases and HIV/AIDS, tuberculosis and still birth mortality in particular have showed decreasing trend while Group II and III increasing over time. Double burden in mortality was highly observed in the age groups of 15-64 years. Those aged > 45 years were dying more likely with non-communicable diseases compared with children. Children aged below 15 years were 16 times more likely to die from communicable, perinatal and nutritional conditions compared with elders. Mortality variation with age has been identified between public and private hospitals. Conclusions: The results of the present study shows that, in addition to the common Group I causes of death, emerging group II diseases are contributing to high proportions of mortality in the public and private hospitals of Addis Ababa, Ethiopia. Thus, priority should be given to the prevention and management of conditions arising during perinatal period such as low birth weight and still birth, HIV/AIDS; tuberculosis, respiratory infections, cardiovascular diseases, malignant neoplasm, chronic respiratory diseases and road traffic accident. The planning of health resources and activities should take into account the double burden in mortality due to Group I and Group II diseases. This calls for strengthening approaches towards the control and prevention of non-communicable diseases such as cardiovascular and malignant neoplasm.