Several crash investigation studies recently lead to opposing conclusions on the importance of abdominal injury risk to children. This paper deals with an in-depth and synthetic analysis of new accidentological data. In particular, the risk is studied by isolating parameters such as age and crash type, as well as restraining systems.

The abdominal risk appears higher above 2 or 3 years old for children using poorly designed booster-cushions or adult belts alone.

The prevention of such a risk for protected children should 6e given by a pertinent restraint-system assessment in frontal impact simulation. Unfortunately, the lack of biofidelity of the pelvic-abdominal segment of the current child dummies does not allow this child restraint assessment. Data for the improvement of dummy pelvises and for the definition of an abdominal criterion are provided in this paper.

The authors propose the use of geometrical criteria to differentiate poor and acceptable booster cushions, as a temporary measure, until effective abdominal injury assessment becomes possible.