The clinical utility of the perinatal autopsy.

Abstract

To evaluate the clinical utility of the perinatal autopsy in determining the cause of a perinatal death. Retrospective observational survey. University-affiliated, private, tertiary care hospital, limited to obstetrics, gynecology, and neonatology. All fetal deaths and neonatal deaths from 1990 and 1991 at Women and Infants Hospital, Providence, RI, were reviewed. Fetal deaths with a gestational age of less than 20 weeks and neonatal deaths occurring more than 48 hours after birth were excluded. A clinical medical record review assessed the clinical diagnosis. Pathology records were reviewed independently. The clinical and autopsy diagnoses were compared and categorized as follows: (1) confirm (clinical and autopsy diagnoses concordant); (2) change (clinical and autopsy diagnoses discordant); (3) add (significant unexpected findings noted on the autopsy although the clinical diagnosis was not altered); (4) autopsy inconclusive; (5) autopsy not done or not available. Of 168 perinatal deaths, an autopsy was not obtained in 26.2% and was inconclusive in 24.2% of cases with an autopsy. Of 94 patients with conclusive autopsies, in 55.3%, the pathologic diagnosis confirmed the clinical diagnosis, and in 44.7%, it changed or significantly added to the clinical diagnosis. These findings support the clinical relevance of the perinatal autopsy. As few published reports directly address the specific yield of the autopsy among fetal and neonatal deaths, these results may be useful in counseling patients who are considering a perinatal autopsy.