[Accidents involving children and adolescents: from epidemiological findings to preventive action]

It is generally accepted that improving epidemiological knowledge of accidents in childhood and adolescence leads to a better implementation of more efficient prevention programs. Nevertheless, it is not yet generally admitted that any new preventive measure addressing these accidents has to be evaluated by competent epidemiologists. Obviously, there are exceptions regarding these two rules. Four examples of successful implementation are presented in this paper: safety packaging of aspirin; fences around private swimming pools; changes in minimal drinking and driving age in adolescence; bicycle accidents. The discussion is focused on four intervention levels suggested for the epidemiologists.