In-patient surgical mortality in Tikur Anbessa hospital: a five-year review.

BACKGROUND: To our knowledge neither hospital nor population-based data is available about mortality patterns in Ethiopia. OBJECTIVE: This review summarizes the patterns of deaths that occurred in hospital under the care of surgical team, during the 5-year period. PATIENTS AND METHODS: This is a hospital based retrospective cross sectional analysis of mortality among surgical patients admitted to Tikur Anbessa Specialized Teaching Hospital (TASTH), Department of Surgery, Addis Ababa, Ethiopia in the period between January 2002 and December 2006. The monthly activity reports, death registry sheets and death certificate of deceased patients from all surgical wards were analyzed using computer based statistical software SPSS version 11.0. Difference in proportions were examined using Chi-square test with its conventional criterion for statistical significance (p < 0.05). RESULTS: The study reviews 694 deaths from 9991 surgical admissions, and 9860 procedures at TASTH over the 5-year period. There were 6.9% overall deaths among the admissions and 7.0% per procedure. The postoperative mortality rate was 4.5% (443/9860). The male-to-female ratio was 2.9:1. The mean age was 41 [Standard Deviation, 17.38] years. Four hundred fifty-two (15.7%) patients admitted on emergency basis and 242 (3.4%) of elective admissions died while in hospital surgical care (p < 0.0001). About 56% of deaths resulted from non-traumatic causes whilst 44% from traumatic causes. Similarly operative deaths were observed in 443 (63.8%) while nonoperative mortality in 252 (32.2%). Among patients with trauma, isolated head injury (59.2%) was major cause of death (p < 0.0001), while, of the non-traumatic admissions, malignancy accounted for 164 (42.1%) of the deaths, When specialties are compared, significant number, 234 (34.4%) of the deaths belong to neurosurgery unit (p < 0.0001) followed by general surgery. CONCLUSION: In conclusion, more men died than women did, more patients admitted on emergency basis died than patients admitted on elective basis (p < 0.0001). Although, the overall mortality rate of 7% and postoperative death rate of 4.5% is acceptable, it could have been reduced significantly had there been appropriate setting to manage trauma cases. Trauma and Neurosurgical cares are yet to develop and need special attention.