Background Availability of injury related mortality data can play a vital role in planning health services for injury prevention and monitoring progress towards Sustainable Development Goal 3.4. In the absence of a robust death registration system, this study explored the feasibility of a model to identify injury deaths occurring outside hospital in rural Nepal.

Methods Data were collected prospectively between February 2019 and January 2020, in two rural municipalities in Makwanpur district. Female Community Health Volunteers notified all deaths in their area to the local Health-Post. Trained data collectors identified the injury-related deaths and invited relatives of the deceased to take part in a structured face-to-face interview.

Results Over one year, 67/451 (14.9%) deaths were identified as secondary to an injury in the two study areas (injury mortality rate 104/100,000 population). The rate of community injury deaths was higher in males (128/100,000) than females (80/100,000) with a median age at death of 40 years (SD=20.8, range 2 to 93). Community injury deaths were most commonly secondary to suicide (65/100,000), followed by unintentional injury (36/100,000) and assault/violence (3/100,000).

Conclusion Large numbers of injury deaths occur outside hospital, and these appear more likely to be due to intentional than unintentional harm. These findings are a valuable additional source of information for engaging stakeholders, developing targeted injury prevention interventions and informing policies.

Learning Outcomes Community deaths secondary to injuries are common and will be missed in hospital-based data collection systems. The mechanism of injury may be different for community compared to hospital cases.