Postmortem pulmonary CT in hypothermia

Abstract

Fatal hypothermia has been associated with pulmonary edema. With postmortem full body computed tomography scanning (PMCT), the lungs can also be examined for CT attenuation. In fatal hypothermia cases low CT attenuation appeared to prevail in the lungs. We compared 14 cases of fatal hypothermia with an age-sex matched control group. Additionally, 4 cases of carbon monoxide (CO) poisoning were examined. Furthermore, 10 test cases were examined to test predictability based on PMCT. Two readers measured CT attenuation on four different axial slices across the lungs (blinded to case group and other reader's results). Hypothermia was associated with statistically significantly lower lung PMCT attenuation and lower lung weights than controls, and there was a dose-effect relationship at an environmental temperature cutoff of 2 °C. CO poisoning yielded low pulmonary attenuation but higher lung weights. General model based prediction yielded a 94% probability for fatal hypothermia deaths and a 21% probability for non-hypothermia deaths in the test group. Increased breathing rate is known to accompany both CO poisoning and hypothermia, so this could partly explain the low PMCT lung attenuation due to an oxygen dissociation curve left shift. A more marked distension in fatal hypothermia, compared to CO poisoning, indicates that further, possibly different mechanisms, are involved in these cases. Increased dead space and increased stiffness to deflation (but not inflation) appear to be effects of inhaling cold air (but not CO) that may explain the difference in low PMCT attenuation seen in hypothermia cases.