Lettrt To The Editor

Abstract

In our report of battered children admitted to San Francisco General Hospital we used as controls children hospitalized for other causes. This methodology was adopted so that the abused and control children would both be drawn from the same racially mixed, indigent part of the population that traditionally utilizes our hospital and therefore would be likely to have in common important socioeconomic features, thereby minimizing bias. However, as Dr. Stover correctly points out, all potential bias was not eliminated.