Introduction Mortuary data are incomplete in many developing countries. Small investments could lead to substantial improvements.

Objective To evaluate recent trends in injury reporting systems. To identify ways to strengthen the reporting systems and make them more useful for injury control.

Methods We evaluated changes in data quality on fatally injured persons at KATH over a 12-year period.

Results During 1994–95, cases in KATH's mortuary logbooks had missing information on cause of death, and only 70 injury deaths/year were reported. This was felt to be a gross under-estimate of injury among Kumasi's 1 000 000 people. A pilot program hired part-time nurses to improve recording of injury deaths. During 1996–1999, 633 deaths/year were recorded, more accurately reflecting injury's toll in Kumasi. To assess the sustainability of this improvement after funding expired, we re-assessed mortuary injury data quality in 2006. Reporting remained high, with 865 fatally injured cases recorded. However, only 80% of these cases were found in mortuary logbooks. An additional 20% were found in supplemental sources in the hospital, principally the ICU. Data quality had diminished slightly.

Conclusion In many developing countries, important information is not recorded. Low-cost improvements could lead to improved reporting and data quality. This study has shown that such improvements are sustainable after the initial investments. Minor problems such as insufficient collation of data from multiple sources and coding of mechanism could be corrected by improved organization and training.