The aim was to ascertain the extent of under-reporting of serious traffic crashes, as recorded by police in New Zealand, and to what extent this coverage was biased by demographic, crash and injury factors. Public hospital records (resulting from on-road traffic crashes) and traffic crash records were linked using probabilistic methods. This involved the use of computer based linkage, in addition to manual review of indeterminate links. Linkage rates were examined by a range of environmental, demographic and injury factors. For 1995, less than two-thirds of all hospitalised vehicle occupant traffic crash victims were recorded by the police. It was found that linkage rates were affected by demographic factors, specifically age and ethnicity. Crash factors also affected linkage rates. The number of vehicles involved and whether or not a collision occurred were found to be important explanatory factors. Injury severity and length of stay in hospital affected linkage rates, and they varied widely by geographic region, which could not be accounted for by regional variations in ethnicity and age distributions. Potential users of traffic crash records need to be aware of the extent and nature of the biases contained within this database. (A)
Abstract