Background It is cumbersome to ask patients for location of occurrence and irrelevant to the treatment upon arrival at hospitals. Since data on location of occurrence are found in ambulances and pre-hospital mobile emergency care units (MECU) information could easily be recorded. It is therefore of interest to explore the potential and possible biases of combining hospital- and pre-hospital data on traffic injury as an alternative to lengthy interviews once patients arrive at the hospital. The specific aim of this study was to assess complement and combined proportions of patients known from ambulances, MECU and trauma registries.

Methods Any patient record documented as “acute traffic injury” in the primary geographic catchment area of Kolding Hospital (level 2 trauma) and Odense University Hospital (level 1 trauma) in Denmark occurring in 2013 was included if identified in at least one of the three registries(hospital trauma, MECU, Ambulance Service). Records were merged on civil registration number and date of injury. Transfers between hospitals were excluded. No other service or hospitals provide service in the area.

Results Primary data consisted of 8895 records (patients) from the ambulance service, 669 from the medical emergency care unit and 564 from the trauma registries (excluding transfers). By matching on civil registration number and date 97 patients were found in all three registries, 140 only in hospitals trauma, 8683 only in pre-hospital and 8147 only in ambulance data. For 69% of the trauma registry patients location of occurrence where known.

Conclusions For the present geographical region in Denmark pooling data from these registries seems a promising way for identifying place of occurrence. However, immediate merge were only valid with manual cumbersome review. Further analysis must look into whether the recorded locations are valid indications of actual places and gives sufficient input to the planning of preventive measures in municipalities.