Clinicopathological Concordance Across Etiologic Categories in Emergency Department Deaths: A Secondary Analysis

Background and Objectives: Clinicopathological concordance may vary according to the underlying disease context in rapidly fatal emergency department (ED) presentations. Previous analysis of the non-surgical subset of this autopsy cohort showed that ante-mortem recognition of the dominant clinical syndrome did not always translate into complete etiologic resolution. The present study extended this analysis to the completed autopsy cohort, including surgical deaths, to compare concordance distributions across major etiologic categories and secondarily compare surgical and non-surgical deaths. Materialsand Methods: We performed a secondary retrospective analysis of a single-center five-year autopsy cohort comprising 76 patients who died in the ED between 1 January 2019 and 31 December 2023 and underwent complete autopsy. The cohort included 45 non-surgical deaths previously evaluated and 31 surgical deaths. Clinicopathological concordance was classified as complete when both the principal disease and fatal mechanism were identified ante-mortem, partial when the dominant clinical syndrome or mechanism was recognized but the underlying disease or anatomical source remained unresolved, and discordant when the ante-mortem diagnosis differed substantially from the principal diagnosis established at autopsy. Concordance distributions across five etiologic categories were compared using the Fisher–Freeman–Halton exact test with Monte Carlo estimation, whereas the surgical versus non-surgical comparison was performed using direct exact computation. Results: Across the complete cohort, complete and partial concordance each occurred in 36/76 cases (47.4%), whereas discordance was identified in 4/76 cases (5.3%; 95% CI, 1.5–12.9%). Concordance distributions differed significantly across etiologic categories (p = 0.0014). Partial concordance predominated in respiratory (14/17, 82.4%) and infectious deaths (8/10, 80.0%), whereas complete concordance was more frequent in surgical-source intra-abdominal (15/21, 71.4%), neoplastic (6/9, 66.7%), and cardiac deaths (10/19, 52.6%). Complete concordance was observed in 61.3% of surgical versus 37.8% of non-surgical deaths, while partial concordance occurred in 32.3% versus 57.8%, respectively; however, the overall concordance distribution did not differ significantly between these groups (p = 0.086). Conclusions: Clinicopathological concordance patterns differed across etiologic categories when surgical and non-surgical ED deaths were considered together. Partial concordance predominated in respiratory and infectious deaths, whereas complete concordance was more frequent in surgical-source intra-abdominal, neoplastic, and cardiac deaths. These findings extend previous observations from the non-surgical subset and suggest that the level of ante-mortem diagnostic resolution may differ across autopsy-defined etiologic contexts. The surgical versus non-surgical differences remained descriptive, and these findings apply specifically to autopsied ED deaths and should not be generalized to all ED deaths.