Postmortem Orbital Hemorrhage in the Investigation of Child Abuse

To the Editor: Iread with interest and some alarm the recent article of Gilliland et al. It is laudable to set protocols for ensuring complete autopsy examinations. However, the proposed selective application of this protocol and the proposed interpretation for the finding of orbital hemorrhage by these authors demand the presentation of a counter point of view. The title indicates that these guidelines are for the Binvestigation of unexplained infant death and suspected physical child abuse.[ It is unscientific to go looking for a finding only in those cases where one expects to find it, and this approach risks the gathering of biased data and seriously misrepresenting reality. How can one be sure that the same finding does not occur in many other situations where one did not expect it and therefore did not look? If this investigational protocol is to be of any utility, it must be applied to all autopsies not just to a preselected group. If, when applied to a large prospective study, it were shown to make some valid distinction, then it may be of value. If, however, the same finding occurs in other clinical settings across all age groups, it has failed to be a unique finding capable of distinguishing between different mechanisms of injury. In the opening background discussion, the authors attempt to present a case for the unique importance of eye findings in evaluating cases of shaken baby syndrome. They make statements that give an almost characteristic/diagnostic/pathognomonic significance to certain findings: