The south african national non-natural mortality surveillance system - Rationale, pilot results and evaluation

Background. While individual mortuaries have recorded data for non-natural deaths in time-limited studies, there have been no systematic efforts to dr aw forensic-medical services and state mortuaries into a nationwide fatal i njury surveillance system. Beginning in June 1998, the National Non-Natural Mortality Surveillance System (NMSS) commenced pilot operation. Objective. To evaluate the NMSS and illustrate its utility from sample find ings. Design. Data entered into the system by mortuary staff were checked against a random sample of cases for which separate forms were completed by an ind ependent researcher. Process observations and follow-up with data users wer e used to assess the system's acceptability, timeliness and data usefulness . Setting. Eighteen mortuaries in six provinces representing approximately 35 000 cases per year, or around 50% of all non-natural deaths. Participants. The National Departments of Health; Safety and Security; and Arts, Culture, Science and Technology; national and provincial forensic med ico-legal services; the South African Police Services; universities and sci ence research councils. Main outcome measures. Surveillance system simplicity, flexibility, accepta bility sensitivity, positive predictive value, representativeness, timeline ss, data usefulness and resources. Results. The NMSS was established at 10 target sites. Lack of equipment, personnel resistance, and closure of some mortuaries prevented implementation in the remaining eight mortuaries. Sens itivity was internally assessed and ranged from 65% to 95% for manner of de ath. Positive predictive value was also internally measured, and ranged fro m 74% to 80% for manner of death and from 71% to 82% for mechanism of death . Timeliness was good, and basic reports covering most items were available 6 weeks after a case had been examined. While staff found the system simpl e, acceptability depended on the individuals involved at different mortuari es, and the system was compromised to some extent by bureaucratic barriers. End users found the data to be of great value. NMSS set-up costs totalled approximately R26 000 per mortuary, and it is estimated that maintenance co sts will be R8.00 per case registered. Conclusions. With minimal resources, the NMSS uses existing investigative p rocedures to describe and report the epidemiology of fatal injuries. The pi lot study demonstrates the feasibility of the system, and identifies the ne ed to remove organisational constraints and individual barriers if it is to be sustained and expanded beyond the pilot sites.