TPT 3.01 Correlation Between Observed Mortality and SORT-Predicted 30-Day Mortality Following the Five Most Common Emergency Laparotomy Procedures in Romania: A Retrospective Multicentre Observational Cohort Study

Abstract

Emergency laparotomy (EL) comprises heterogeneous high-risk procedures with substantial postoperative mortality, reaching 8–18% at 30 days. Although overall 30-day mortality (30-dm) is documented, procedure-specific outcomes remain limited. Defining intervention-level mortality may improve risk stratification and resource allocation. This study aimed to quantify actual 30-dm for the five most frequently performed EL procedures in Romania and examine its correlation with Surgical-Outcome-Risk-Tool (SORT) predicted mortality. This retrospective multicentre cohort study included adults undergoing EL over two years across four acute care hospitals in North-Eastern Romania, including one tertiary centre. The primary outcome was the correlation between actual and SORT predicted 30-dm for the 5 most common EL procedures. The study followed the TRIPOD checklist. Statistical analysis was performed using GraphPad-Prism v10.6.1. Among 793 EL patients, 448 (56.4%) underwent the five most frequent procedures: incisional hernia repair with adhesiolysis (n=100;12.61%); actual 30-dm 4.0%; SORT 3.8% (1.7,11.2); R²=0.33, AUC=0.9648, p<0.0001; adhesiolysis (n=107;13.495); actual 30-dm 15.9%; SORT 6.86% (3.1,14.0); R²=0.05, AUC=0.7497, p=0.0117; suture of perforated peptic ulcer (n=88;11.09%), actual 30-dm 20.5%; SORT 5.78% (1.4,14.0); R²=0.24, AUC=0.8516, p<0.0001; Hartmann’s procedure (n=88;11.09%), actual 30-dm 21.6%; SORT 23.06% (12.1,40.3); R²=0.16, AUC=0.7645, p=0.0001; and inguinal hernia repair with adhesiolysis (n=65;8.19%) actual 30-dm mortality 1.5%; SORT 3.21% (0.9,8.2); R²=0.13, p=0.0001. Data is presented as median ± IQR. SORT demonstrated limited reliability, weak correlation, and persistent misprediction of 30-day mortality across the five most common emergency laparotomy procedures in Romania.