The UK Department for Transport commissioned a review of injury severity classification in order to assess whether the police could be helped to distinguish more serious injuries from the less serious consistently and how this might be achieved. Accurate measurement of 'severe' road casualties is essential to guide the development and evaluation of national and local road safety strategies and to target interventions at high-risk locations. The rationale behind this study is to determine whether the collection of a small number of additional data items would assist in distinguishing slight from severe injuries and provide more reliable statistics. First, injuries sustained by people admitted to hospital were linked to mortality records, which provided the annual number of cases for each specific injury(or group of injuries)1 and the number who subsequently died. This was used to identify those injuries most likely to result in death. Secondly, data from Hospital Episodes Statistics (HES) was linked to the police dataset (STATS19) to look at the frequency of injuries resulting from road traffic collisions. This provided the distribution of injury types and body locations occurring in the more serious (i.e. hospitalised) cases of road traffic injury recorded in STATS19 and showed that the 23 most common injuries recorded (out of a total of some 855) accounted for nearly three-quarters of all cases in the linked dataset. Using the analysis described above, supplemented with data from other sources and expertise in clinical medicine, disease classification and road traffic epidemiology, two alternative approaches were devised with the aim of improving the measurement of severity within STATS19. The first approach consisted of a series of observations police officers could make or questions they could ask at the scene of a collision which do not require direct clinical training or expertise. The second approach was to ask police officers, with the help of paramedics if present, to assess the actual injury or injury category using a pre-defined list developed from analysis of injuries sustained in road traffic collisions. In terms of further work to validate the list of injuries for inclusion in CRASH or other data gathering systems, a large scale STATS19/hospital data matching exercise would be needed to assess the levels of actual injuries (as recorded by hospitals) compared with suspected injuries. The sub-categorisation of severity within the overall serious category may need to be revised following this exercise. This report may be accessed by Internet users at http://www.dft.gov.uk/pgr/roadsafety/research/rsrr
Abstract