The epidemiology of school injuries: the problem of measuring injury severity.

In the few surveys of injuries occurring in schools, severity has been defined using a priori criteria based on the nature of injury and by the area injured. The validity of these methods of classifying injury severity has not been established. The association between two commonly used measures of injury severity (nature of injury and body area injured) with referral to medical assessment was explored, based on a simple model derived from the literature. Kendalls Tau-b was used to assess the association between the indicators of injury severity and referral; controlled for both school level (elementary school and secondary school) and the degree of behavioural control that could be exerted in the location of the injury. There was a very low association between indices of injury severity and referral for medical assessment regardless of level of school or degree of behavioural control. It is concluded that the simple model derived from the literature does not adequately describe the relationship between injury severity and referral in the school setting. The major issue facing school staff is not the measurement of injury severity per se, but the appropriate referral of the injured child for medical assessment and treatment.