February 24, 2012

Can We Skip the Autopsy?

AUTHOR: Tom Heston, MD

The postmortem autopsy is considered the gold standard in the determination of the cause of death. Newer imaging technologies, however, including high resolution computed tomography (CT) and magnetic resonance imaging (MRI), may allow in some cases a virtual autopsy instead, that utilizes medical imaging alone. The benefits of a virtual, imaging autopsy include the potential for conducting more autopsies which could lead to more accurate mortality statistics, and reduced costs. The virtual autopsy may also be more widely accepted by families and religions.

A study published in the January 14th, 2012 issue of the Lancet compared traditional autopsy results with virtual autopsy by both CT and MRI. They randomly enrolled 182 cases that underwent both virtual and full conventional autopsy. The CT and MRI scans were independently interpreted for cause of death, then a combined report was created from both imaging modalities. The radiologists also indicated how confident they were in their diagnosis, which was based entirely upon the scan images. The cases were then dividing into two groups: those with a definite imaging diagnosis, and those without a definite imaging diagnosis.
The researchers found that overall, about 1 in 3 virtual autopsies contained a major discrepancy when compared with the full, traditional autopsy. Radiologists considered the imaging diagnosis for cause of death to be definite in about half of the cases. In these cases where the imaging results were considered definite, the major discrepancy rate with full autopsy was about 1 in 6. The researchers also found that CT was more accurate than MRI when using a conventional autopsy as the gold standard. Major common sources of error were when the cause of death was coronary heart disease, pulmonary embolism, bronchopneumonia, and intestinal infarction. As the study progressed, the radiologists improved their interpretation accuracy, however, major discrepancies continued to exist.

The researchers concluded that when conducting a virtual autopsy, CT imaging was better than MRI scanning in providing an accurate cause of death. When the findings on virtual autopsy were considered definite, the major discrepancy rate with full autopsy was 16%.

COMMENT: This is a new, emerging application of medical imaging that has tremendous potential. The authors note that when the imaging diagnosis was considered definite, the error rate was comparable to the error rate of a conventional, full autopsy. As physician experience with this relatively new application of medical imaging improves, it is likely that the accuracy will significantly rise. Because of the relatively low cost and ease of conducting a virtual autopsy, it is likely to become fully integrated into and a routine part of postmortem investigation.


Roberts IS, Benamore RE, Benbow EW et al. Post-mortem imaging as an alternative to autopsy in the diagnosis of adult deaths: a validation study. Lancet. 2012 Jan 14;379(9811):136-42