Objective birth data and the prediction of child abuse.

The Cardiff Births Survey contains objective data that are easily collected and are reproducible between observers. We have compared survey records of 80 abused children with those of 80 control children. Of 44 factors that were examined 11 were appreciably more common in abused children. Such factors included the fact that the baby was preterm and had been admitted to a special care baby unit. Other factors were young maternal age, marital instability, low social class, and maternal smoking. 66% of abused children had at least 5 of these factors compared with only 16% of the control group.