1 Department of Family Medicine, Research Division, Jefferson Medical College of Thomas Jefferson University, Philadelphia, PA 19107
To the Editor.
Rosenberg et al1 report on a prospective, "random" study of 476 patients up to 2 years of age seen in an emergency room. The paper is concerned with a problem of importance to health care providers and of great interest to parents and society. The paper reports statistically significant differences in rates of observed or reported child abuse between children grouped according to the presence of certain abnormal features. Our concern is that the reported results may be easily misinterpreted and, therefore, may be misleading.