Post-mortem computed tomography compared to medico-legal autopsy - pathologies in the torso and limbs

Introduction: Post-mortem computed tomography (PMCT) has become a routine part of post-mortem forensic evaluation in many institutes worldwide. Its added benefit to the traditional autopsy is indisputable, but its ability to detect different findings compared to those found in an autopsy has only been evaluated in a limited number of large scale studies. Methods: In this retrospective study, we assessed the agreement between autopsy and PMCT, and their ability to detect pathological findings (by "Kappa" and "McNemar" scores, respectively), using all finding of both methods as reference standard. We included findings in the torso and the limbs extracted from autopsy and PMCT reports of 105 consecutive cases. Results: The level of agreement between autopsy and PMCT depends on the tissues, locations, and type of pathologies examined. Autopsy much better demonstrates bullet tracks (68/76 vs. 18/76, p<0.01) and stab wounds (22/25 vs. 11/25, p=0.013), while PMCT is more sensitive to shrapnel (86/121 vs. 37/121, p<0.001). PMCT better demonstrates fractures (259/344 vs. 222/344, p=0.012), especially in bones that are hard to access in autopsy, but is less sensitive to rib fractures (137/177 vs. 115/177, p=0.037). Parenchymal organ pathologies are not well demonstrated by PMCT (194/257 vs. 117/257, p=<0.001). Conclusions: Shrapnel, foreign bodies, gas-related pathologies, pelvic fluid and fractures, excluding rib fractures, are detected more often by post-mortem CT. It is important to consider PMCT as a tool in the evaluation of specific tissues and organs, possibly providing solid answers, or at least directing the team in the performance of the autopsy.