POLICE RECORDING OF ROAD ACCIDENT IN-PATIENTS. INVESTIGATION INTO THE COMPLETENESS, REPRESENTATIVITY, AND RELIABILITY OF POLICE RECORDS OF HOSPITALIZED TRAFFIC VICTIMS

Abstract Many road safety research projects make use of the official police road accident data. Their use is often restricted to the data of fatal accidents and fatalities because it is the only complete registration, and the extent of underreporting of injury accidents is unknown. The need to extend the use of data beyond fatalities is great for two reasons: (1) in a small country like the Netherlands the absolute numbers (less than 2000 fatalities per annum in recent years) are often too small for detailed analyses. (2) Fatal accidents are not typical road accidents but an extreme type. Data on surviving in-patients however, is not so extreme and there are more than 20,000 per annum. The incomplete police data on road accident in-patients was compared with the hospital discharge data to establish how representative it was. Hospital data was collected through the Medical Record Foundation, a national institution, which registers approx. 95% of all road accident in-patients. During the years 1977–1979 it was found that the extent of underreporting was constant and for the year 1979 the police data had a coverage of 83% of all road accident in-patients. According to the results of an Eckart-Young analysis, the general structures of the police and hospital data were similar but there were differences. The underreporting of users of motorized vehicles in the age group 15–34 yr was significantly smaller than for others; and of cyclists and pedestrians, particularly in the age group 0–14 yr, significantly greater. The police data is therefore reliable for time series and for period studies of most mode of transport/age group combinations.