Background Estimating injury burden in low income countries is hampered by lack of structured nationally representative data collection systems.

Objective To demonstrate the construction of estimates of the national burden of injuries using a systematic assessment of multiple data sources. The approach harnesses the strength of each source and accounts for biases in data.

Methods We conducted an environmental scan of all existing data sources that can inform estimates of the incidence of fatal and non-fatal injuries in Uganda. These included:

Mortuary data from Kampala city and Mulago hospital mortuaries.

Verbal autopsy data from two district health and demographic surveillance systems.

Community Injury Survey: conducted in one rural district and one urban district,

Uganda National Household Survey.

2004 Northern Uganda Baseline Survey.

Hospital based trauma surveillance at Mulago hospital.

Outputs and Building Estimates We estimated national injury mortality from urban rates estimated from mortuary data from Kampala and rural rates based on the two HDSS sites. We estimated the population incidence of non-fatal injuries using the household surveys. Total injury incidence was obtained from the two national health surveys and the external cause disaggregation was computed from the community injury survey. Finally, we used the hospital surveillance data to estimate the distribution of nature of injuries.

Conclusion Despite paucity of information, we derived national injury estimates from diverse data sources. Such an approach may have relevance to other low income settings. Development of a comprehensive national health information system that captures injuries is needed in Uganda.