[How to measure disability in children? Different methodologies, values and applications].

Infantile mortality data are insufficient for perinatal care evaluation. Long-term morbidity, particularly child impairment prevalence, needs to be assessed regularly. In this paper different surveys on child impairment registration published in the literature are examined: cohort studies, follow-up studies on "at risk" children, cross sectional surveys and morbidity registers. Study designs and case ascertainment are analysed as well as case validation and reference population. The consistency of the results is discussed. In France, most of the studies on childhood impairment are follow-up studies of at risk children, not concerning a geographically defined population, and there is a need for larger epidemiological studies in order to better assess perinatal morbidity and the quality of perinatal cares.