Clinical Relevance of Unexpected Findings of Post-Mortem Computed Tomography in Hospitalized Patients: An Observational Study

Abstract

Background and objective: The current literature describing the use of minimally invasive autopsy in clinical care is mainly focused on the cause of death. However, the identification of unexpected findings is equally important for the evaluation and improvement of daily clinical care. The purpose of this study was to analyze unexpected post-mortem computed tomography (PMCT) findings of hospitalized patients and assess their clinical relevance. Materials and methods: This observational study included patients admitted to the internal medicine ward. Consent for PMCT and autopsy was requested from the next of kin. Decedents were included when consent for at least PMCT was obtained. Consent for autopsy was not obtained for all decedents. All findings reported by PMCT were coded with an International Classification of Diseases (ICD) code. Unexpected findings were identified and subsequently categorized for their clinical relevance by the Goldman classification. Goldman class I and III were considered clinically relevant. Additionally, correlation with autopsy results and ante-mortem imaging was performed. Results: In total, 120 decedents were included and evaluated for unexpected findings on PMCT. Of them, 57 decedents also underwent an autopsy. A total of 1020 findings were identified; 111 correlated with the cause of death (10.9%), 508 were previously reported (49.8%), 99 were interpreted as post-mortem changes (9.7%), and 302 were classified as unexpected findings (29.6%). After correlation with autopsy (in 57 decedents), 24 clinically relevant unexpected findings remained. These findings were reported in 18 of 57 decedents (32%). Interestingly, 25% of all unexpected findings were not reported by autopsy. Conclusion: Many unexpected findings are reported by PMCT in hospitalized patients, a substantial portion of which is clinically relevant. Additionally, PMCT is able to identify pathology and injuries not reported by conventional autopsy. A combination of PMCT and autopsy can thus be considered a more comprehensive and complete post-mortem examination.

Keywords: autopsy; post-mortem computed tomography; radiology; unexpected findings.

Figure 2

Two cases that show a clinically relevant unexpected finding identified by PMCT and interpreted as false-negative of autopsy. (A) This example shows a hydropneumothorax in a 73-year-old woman (class I finding). An air-fluid level can be identified at the white arrow. The pleural drain that was placed for drainage of pleural fluid can also be seen (white arrowhead). The pneumothorax component was unknown, and the pneumothorax test during autopsy was negative. (B) A 74-year-old male with a periprosthetic fracture of a hip prosthesis. The autopsy report mentioned a normal position and mobility of the extremities with no fractures. The autopsy determined the cause of death as a pneumosepsis. The finding did not have a direct relationship to the cause of death and was subsequently classified as a class III finding.

Figure 3

This figure shows multiple dense intra-cranial masses in the right hemisphere identified by PMCT in a 67-year-old male recently diagnosed with a stage IV small-cell lung carcinoma. The patient experienced no neurological complaints and showed no abnormalities during the neurological examination. The intra-cranial masses were suspected to be cerebral metastases and subsequently scored as a class III unexpected finding. Brain autopsy was not performed.

Figure 1

Flowchart of post-mortem computed tomography (PMCT) findings in the autopsy subgroup of 57 decedents. The findings are classified according to Goldman.