Global childhood mortality rates remain high. Millennium Development Goal 4 focused efforts on reducing rates by two-thirds between 1990 and 2015. In Ethiopia, child mortality rates dropped 71 % from 1990 to 2015, however it is estimated that 184,000 Ethiopian children die each year. There is limited information about pediatric hospital admissions in Ethiopia. Our aims were to examine the temporal relationship of mortality to admission, describe the demographics, and identify cause mortality of children admitted to the Zewditu Memorial Hospital (ZMH).

A four-year retrospective review of pediatric admissions was conducted at the pediatric emergency room and pediatric hospital ward at ZMH in Addis Ababa, Ethiopia. Admission entries from 2011–2014 of children age 29 days-14 years were reviewed. Age, gender, admission date, disease classification, discharge status and date were obtained. Patient gender was compared using Chi-square analysis. A descriptive analysis was used for age and cause mortality.

A total of 6866 patient entries were reviewed. The proportion of admissions younger than age 5 was 0.747 (95 % CI 0.736-0.757). Overall mortality was 0.042 (95 % CI, 0.037-0.047). The proportion of recorded deaths occurring within 2 days of admission was 0.437 (95 % CI 0.380-0.494). The proportion of male admissions was significantly higher than female admissions in all age groups (male 0.575, p < 0.0001, 95 % CI 0.562-0.586). The main causes of mortality were pneumonia (0.253, 95 % CI, 0.203-0.303), severe acute malnutrition (0.222, 95 % CI 0.174-0.27), HIV/AIDS-related complications (0.056, 95 % CI 0.029-0.083), spina bifida (0.049, 95 % CI 0.024-0.074), and hydrocephalus (0.045, 95 % CI 0.021-0.069).

Our study revealed a lower mortality rate than previously reported in Ethiopia. Despite this, 44 % of pediatric hospital mortality occurred early during hospitalization, higher than reported at other Ethiopian hospitals. This adds further evidence that systematic efforts should be dedicated to improve pediatric emergency care. Admissions included 58 % male patients, similar to other reports in Ethiopia implying that this may be a nation-wide phenomenon. The observed disparity may be due to societal factors regarding care-seeking behaviors or male predilection for respiratory illness warranting further investigation. Cause mortality patterns were similar to reports in analogous settings.