This study aimed at compiling from existing hospital casualty records and police accident records a dataset containing information on the nature of injuries to road accident casualties and on the circumstances of the accident in which each casualty was injured. Hospital in-patient enquiry data and national police accident data, which comprise the source data, are described and their accuracy discussed. These records are anonymous. Consequently, matching casualty records in the two datasets was a matter of selecting the most likely match. The method of selecting matches is explained, and some of the difficulties involved discussed. It is estimated that about 50 per cent of the hospital records can be readily matched and that less than 10 per cent of the matches are incorrect. The matched dataset derived from one year's data for almost the whole of great Britain contains 3800 accident records in which the circumstances of the accident, the vehicle involved, and the casualties from the accident are recorded. For each accident the location of the major injury and length of stay in hospital of one casualty are known. Some examples are given of the many analyses carried out to study the usefulness and limitations of data derived in this way in investigating the factors which influence the nature of injuries to casualties. (Author/publisher)
Abstract