Prevalence of family violence in adults and children: estimates using the capture-recapture method.

BACKGROUND Reliable prevalence estimates of family violence in adults and children are difficult to obtain. Most are based on surveys or registration counts, whose research designs and methods are often questionable, making the results difficult to compare. This article presents an alternative approach. METHODS The capture-recapture method (CRC), which makes it possible to estimate unknown numbers in a partly hidden population, was applied to data from eight collaborating organizations in Haarlem, The Netherlands. RESULTS Uniform data registration took place over a 7-month period. The 1-year prevalence rate for adult victims of family violence was estimated to be 2.0% of the adult population (95% CI: 1.3-3.1). For victims of child abuse, it was 1.5-2.5%, and for child witnesses of spouse-abuse, it was 1.2-2.1%, though small numbers made these results more uncertain. Only approximately 20% of all victims in the study were known to one or more of the participating organizations. Our results accorded quite well with results obtained by general health surveys in the Netherlands. CONCLUSIONS CRC appears to be a valid and feasible research method for estimating the prevalence of family violence and child abuse. It can be used to complement other methods, especially in young children, in whom valid results are otherwise difficult to obtain.