Traffic accidents have been analysed systematically in Göteborg since 1966. For this a computer-based information system has been used since 1971. This system processes accident and environment data from police reports and traffrc measurements and facilitates continuous traffic regulations and road improvements. The number of traflic accidents per year in Göteborg decreased from 6,000 to 3,900 during 1971-1980, mainly as a result of successful traffic safety investments.

A complementary registration of traffrc accident casualties was started at the emergency hospitals in 1979. The purpose was improved injury classiflrcation and decreased number of not-reported accidents. This hospital-based primary registration was expanded to include the medical and social consequences for patients injured in traffic accidents during three months in 1979. The time in hospital, the hospital costs, the time on sick-leave and compensations from the health insurance indicated the accident outcome. The time in hospital and the hospital costs were found to be the most useful accident outcome variables. Pedestrian accidents were by far the most important category according to this analysis.

Improved traffic safety depends among other things on reliable traffrc and accident data. A continued decrease of the accident number should be a goal for the road traffic authorities. A more differentiated injury classifircation should improve the accident analysis and justify a continued use of this system. Medical information can be used in a more precise way to achieve this as indicated in this study.