Background Harstad Hospital started in 1985 injury surveillance. Published articles on under-reporting of traffic injuries by police, inspired public funding of the surveillance.

Objective To show that hospital data on traffic injury can facilitate prevention and assess update police underreporting.

Methods The Nordic System variables were recorded in an Injury Data Bank (IDB) for all hospital treated injuries, in and outpatient. Geographical data on traffic injuries in Harstad (25000) were included. The IDB was used for analysis of the local traffic injury panorama. Data on where and when crashes occurred, facilitated tailored interventions. Active interventions were information to risk groups through media, schools and NGOs. Passive interventions, e.g. separate bicycle and pedestrian roads, lowering speed limits, photo box installation and converting crossings to roundabouts. From 1994 to 2015 all traffic injuries, reported by the police, were available for comparison with the IDB.

Results International publications reported effective traffic interventions. These were evaluated by a Cochrane group craving a reference population and hard endpoints. Police data on traffic injury showed underreporting compared with the IDB. The police percentage for registered road users were: bicyclists 3, motorcycles and mopeds 21, pedestrians 24 and automobiles 26.

Discussion Local IDB facilitated the effectivity of an injury prevention program in Harstad. The cost of funding a nurse for supervising the data collection, and the time used by the preventive coalition, has overwhelmingly been balanced by savings in terms of human suffering and treatment resources. Police under-reporting seems substantial.