UBC Faculty Research and Publications

Injury severity in police collision reports correlates poorly with requirement for hospital admission Brubacher, Jeffrey; Chan, Herbert; Erdelyi, Shannon

Abstract

Introduction: This study linked driver records with hospital records to investigate the correlation of injury data in police collision reports with need for hospital admission. Methods: We studied hospital admissions and traffic collisions involving British Columbia residents between 2005 and 2015. Probabilistic linkage between driver licence and personal health number was successful for 95.5% of drivers. We first compared population level statistics on number of serious injuries following a motor vehicle crash according to police reports with the number hospital admissions for road trauma recorded in hospital records. Second, we determined how often drivers involved in police-reported crashes were admitted to hospital. Finally we studied drivers admitted to hospital for road trauma and determined how many of these collisions resulted in a police report, and when reports were available, how often police indicated no injury, minor injury, or major injury. Results: There was poor agreement between police indication of injury severity and requirement for hospital admission. Population statistics from police reports counted only 70% as many road users with major injuries as there were hospital admissions for road trauma survivors. At the individual level, police indication of major injury had even worse correlation with requirement for hospital admission. Only half (56%) of drivers for whom police indicated “major injury” required admission to hospital and only 36% of drivers who required admission to hospital after a crash had a police report indicating major injury. Part of this discrepancy is because police attended only 65% of crashes of hospitalized drivers. However, even when police attended the crash of a hospitalized driver, the police report indicated a major injury only 55% of the time. Implications: Our findings highlight a significant limitation of road trauma surveillance and traffic evaluations that rely on injury data from police collision reports. Recommendations for improvement are made.

Item Citations and Data

Rights

Attribution-NonCommercial-NoDerivatives 4.0 International