Back to search

University of Crete (UOC)

A multicenter comparison of surgical risk prediction tools for emergency laparotomies in Greece

Abstract

dc:description

Introduction: Emergency Laparotomy (EL) encompasses a broad range of procedures that warrant urgent abdominal exploration and is associated with substantial morbidity and mortality. Interest in risk stratification for this patient population is rising, as prediction of perioperative risk optimizes shared decision-making and planning of the proper operative or non-operative strategy and has been shown to be associated with improved patient outcomes. In Greece, reports of the outcomes of patients undergoing EL are missing, and it is still unclear, which prognostic models might be suitable for prediction of mortality or major morbidity following emergency general surgery (EGS). The aim of this dissertation is to provide insight into the outcomes of patients undergoing EGS in Greece, to identify the best performing model for prediction of postoperative mortality and to develop a new prognostic model suited for the Greek population. Methods: The dissertation was conducted based on a prospective multicenter collaboration of 11 centers in Greece and Cyprus, under the coordination of the University Hospital of Heraklion in Greece. Patient inclusion and exclusion criteria were in accordance with the National Emergency Laparotomy Audit (NELA). Following data collection between January 2020 and May 2021, the postoperative outcomes of patients undergoing EL were reported and risk factors for 30-day mortality were identified through a mixed-effects logistic regression analysis. Postoperative pathways of the Greek patients were analyzed and compared with their respective counterparts in the UK. Subsequently, a comparative external validation of four prognostic models for 30-day mortality following EL, namely the National Emergency Laparotomy Audit (NELA) model, the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) Surgical Risk Calculator (SRC), the Portsmouth Physiological and Operative Severity Score for the Enumeration of Mortality and Morbidity (P-POSSUM), and the Predictive Optimal Trees in Emergency Surgery Risk tool (POTTER) was carried out. Measures of overall performance, discrimination, calibration and clinical utility were used to compare the models. Finally, a novel prognostic model for prediction of serious complications after EGS based on the Clavien-Dindo classification, was developed and underwent internal validation. Results: 633 EL patients were included in the analysis, most commonly undergoing surgery for gastrointestinal obstruction (39%), perforation (36%) and ischemia (15%). The 30-day mortality rate was 16.3% (95% CI, 13.5–19.4%). Age, American Society of Anesthesiologists (ASA) status, preoperative functional status, ascites, sepsis/septic shock and diabetes were independent factors associated with mortality. A significantly higher postoperative mortality was noted in the Greek cohort in comparison with the NELA cohort (16.3% vs 8.7%), despite a similar proportion of high-risk patients, while higher levels of postoperative care were underutilized (ICU admission rate: 15.8% vs 56.8%). The comparative external validation of prognostic models for 30-day mortality showed that all models had adequate discriminative performance, however the ACS-NSQIP model, especially when combined with the subjective surgeon’s assessment, was superior in terms of calibration (Hosmer-Lemeshow test, p = 0.742), had higher net benefit across a wider range of thresholds, properties that were retained following recalibration of the models’ intercept and slope. The ACS-NSQIP model and its surgeon-adjusted version showed minimal between-hospital heterogeneity and is expected to perform better in new centers. Lastly, the SErious complications After Laparotomy (SEAL) model was developed. This included 6 predictors, namely ASA score, presence of sepsis/septic shock, dependent functional status, ascites, blood urea nitrogen and serum albumin. The model displayed good performance for prediction of serious postoperative complications following internal validation, with favorable overall fit (scaled Brier score: 25.1%, 95% CI 24.1–26.1%), discriminative ability (optimism-corrected c-statistic: 0.80, 95% CI 0.79–0.81) and calibration (optimism-corrected calibration slope: 1.01, 95% CI 0.99–1.03). The SEAL model outperformed the ACS-NSQIP model, which was used as a benchmark in this study following its recalibration. Conclusions: EL in Greece is associated with significant postoperative mortality, partially explainable by differences in postoperative pathways, with concerning underutilization of higher levels of care. The ACS-NSQIP SRC’s performance for 30-day mortality was promising and could potentially be implemented as the tool of choice for risk stratification of Greek patients undergoing EGS. Finally, the novel SEAL model could predict the risk of developing serious complications following EL, serving as a new tool available for quantification of perioperative risk in Greek hospitals. Further exploration of new prognostic tools and updating of validated models is mandatory for proper prognostication, which translates to better patient outcomes.

Degree

thesis:*
Grantor dc:publisher
University of Crete (UOC)
Year dc:date
2025

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Kokkinakis, Stamatios
  • Κοκκινάκης, Σταμάτιος

Subjects

dc:subject × 16

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/58538

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Kokkinakis, Stamatios; Κοκκινάκης, Σταμάτιος. A multicenter comparison of surgical risk prediction tools for emergency laparotomies in Greece. University of Crete (UOC), 2025. http://hdl.handle.net/10442/hedi/58538