It is almost 40 years since Dr Henry Kempe and his colleagues in Denver shocked the world with their paper “The battered child syndrome” (Journal of the American Medical Association 1962;181:17–24) but still paediatricians there are able to comment on the failure of doctors to make the diagnosis of child abuse. They report (Journal of the American Medical Association1999;281:621–6) that of 173 children under 3 years seen at the Denver Children’s Hospital during 1990–95 with inflicted head injury, 54 (31%) had been seen by doctors who failed to make the diagnosis. In these cases the diagnosis of abuse was delayed by up to 6 months (mean 7 days, range < 1–189 days). The children in whom the diagnosis was not made seem to reflect doctors’ preconceptions; they were more often young infants, white, living with both parents, and without breathing difficulties or seizures. Missing or misinterpreting x ray abnormalities contributed to the difficulty in seven cases.

Of the 54 children whose diagnosis was missed 15 (28%) were injured again, 22 (41%) suffered complications related to the missed diagnosis, five died, and four of the deaths were thought to have been preventable. Nevertheless, the mortality in recognised cases (17%) was greater than in missed cases (9%), clearly because milder cases are more likely to be missed. Even in Denver, where doctors must have been exposed to as much publicity about child abuse as anywhere in the world, the diagnosis is still missed and a JAMAeditorial writer can still comment “physicians can do better” (Ibid: 657–9)