At-scene investigations of road crashes for research purposes commenced in the 1950s in the United States and England. These investigations ranged from ones conducted by small multi-disciplinary teams of researchers to the use of specially trained police and medical officers, as exemplified by the Automotive Crash Injury Research program of Cornell University which laid the foundation for our understanding of the causes of injury in car crashes. The first attempt by a small team of researchers to investigate a representative sample of crashes at the scene was conducted in Adelaide in the early 1960s. The second Adelaide in-depth study, conducted in the mid-1970s, successfully investigated in great detail a sample of crashes that was representative of the population of crashes by time of day and day of week. Since then crash investigation at the scene has not been conducted on a representative basis but still with an ability to identify many factors that would otherwise be missed. At scene crash investigation has also formed the basis of case control studies such as those relating travelling speed with the relative risk of crash involvement.
This paper will provide a brief review of at-scene crash investigation over the past 50 years, noting methodological lessons that have been learnt and, in some instances, forgotten. It concentrates on research conducted in the United States and Australia.