SHOULD SEVERE AND MILD CHILDHOOD ACCIDENTS BE STUDIED SEPARATELY?

Abstract

The distinction between accidents of varying severity is rarely made in studies of the accident process. If different social and behavioral processes lead to severe and mild accidents, or if there is more error in the report of mild accidents, inference will be biased. Social and behavioral characteristics of 11,966 British children from birth to five and mother's reports of accidents requiring medical attention were analyzed. Severe accidents were those resulting in hospitalization; mild accidents were those treated on an out-patient basis. The report of severe accidents was more reliable than mild accidents: 56% of the severe accidents reported by the mother were also in the medical record, versus 14% of the mild accidents. Simultaneous multiple regressions revealed different patterns of association: mild accidents were most significantly associated with male sex, followed by mother's depression, young age, poor housing, liberal attitudes, non-neurotic and antisocial behavior. Severe accidents were most strongly associated with older siblings, mother's youth, child's antisocial behavior, mother's depression, breath holding, hearing problems, male sex and number of moves. The patterns of associations suggest that there is both differential reporting of accidents of varying severity and different processes involved. We conclude that inference about accident processes will be more valid if severe and mild accidents are analyzed separately.