Development of mortality statistics at governmental hospitals, Egypt.

Mortality statistics are of crucial importance to epidemiological research. Sources of mortality data in Egypt are many. The first records are hospital records, which submit its data to health office records and lastly civil register office records. Hospital records include hospital death notification certificate and hospital mortician register. Health office records include mortality register, death certificate and death notification report to civil register office. The latter is designed to have a coded data and cause of death according to ICD-10. It is supposed that the death notification report that originates from the hospital is the source of data for the rest of records and registers. So, all these data for the same dead person must be the same. Quality of the original data will reflect itself on the rest of data. This study aims at establishment and assessment of automated system for management of mortality data at (hospital 1, affiliated to Ministry of Scientific Research) and comparing this system and its output with another one (hospital 2, affiliated to the Ministry of Health). The tools of this study are review of the above mentioned records and registers and personal interviews. Death certificate itself is not reviewed, as it is not found in the health office. One-year data sample was chosen for the study. Development of the system at hospital 1 is based on system analysis, design, implementation, and evaluation. Results showed that automated system of hospital1 have the values of automation as timely retrieval and decreasing the error of transcription. Epi-Info software package gives easy statistical analysis and the potential for record linkage. Comparison between hospital death notification certificate that originated from the hospital and health office register showed 97.4% matching for words or expressions in hospital 1 and 50.0% in hospital 2. Comparing hospital death notification report with that originated from health office to related civil office, showed 100.0% matching for hospital1 and 86.7% for hospital 2. As regards correctness of reporting the cause of death, 97.4% of deaths in hospital1, and 29.9% in hospital 2 showed non-specific cause of death reported as circulatory and respiratory distress, that is the actual cause of death is not reported. It is concluded that, unification and automation of the systems is a requirement. Record linkage between different locations of mortality data is recommended. Training on reporting and coding the cause of death is also recommended.