The quality of the injury severity classification by the police: An important step for a reliable assessment

This study aims to assess the under-reporting and misclassification of the traffic injury severity reported by the police for the first time in Portugal. The non-fatality traffic injuries classified by the police are compared with the information recorded by the hospitals using linked data. The underreporting in the police data was found to be of 29%. Therefore, a significant number of road traffic casualties admitted in the hospitals were not known by the police. Taking advantage of the linked information on accident injuries, the misclassification in the police reports is assessed considering two criteria: the length of hospital stay (LS) and the maximum abbreviated injury scale (MAIS). The latter criterion corresponds to the common definition recently established by the European Commission, which has the advantage of representing the medical conditions of the casualty. The comparison between police classification and LS indicates that a discrepancy between the police reports and the established police definition exists. On the other hand, the police classification shows a tendency to overclassify the injury severity regarding the MAIS. A remarkable proportion of severe injuries reported by the police are, in fact, slight injuries. Additionally, using univariate and multivariate analyses, factors contributing to the misclassification of casualties by the police are identified. Finally, similarly to the fatality adjustment coefficient used in Portugal and in other countries in the past, non-fatality adjustment coefficients were computed to estimate the total casualties taking into account the under-reporting and misclassification phenomena. These adjustment factors are compared with other countries.