Measuring the frequency of "severe" accidental injury in childhood.

STUDY OBJECTIVE The aim was to improve the epidemiological information concerning child accidental injuries which can be extracted from routine inpatient and coroners' inquest data. DESIGN This was a retrospective study of coroners' inquest reports and inpatient case notes to undertake objective severity scaling and to extract basic data. This material was related to denominators from OPCS mid-year population estimates, to 1981 census ward populations, and to sociodemographic data collected in a local census in 1986. SETTING The study population was derived from three north east health districts and their corresponding census enumeration districts. PARTICIPANTS A stratified sample of 500 children aged 0-16 years from among residents admitted to hospital with accidental injuries in 1986 was studied, together with all accidental injury child deaths between 1980 and 1986; 94% of the relevant case notes were localised and extracted. MAIN RESULTS When differentiated by injury severity there are major systematic differences in the basic epidemiology of child accidental injury by age and place of residence of victims as well as in the nature and causes of injuries sustained. CONCLUSIONS Injury severity scores can be used to define a "severity" threshold, within the spectrum of injuries, leading to hospital admission or death. By ensuring complete ascertainment this technique can provide a more accurate case definition than crude admission rates for estimating the frequency of injury in a population of children.