Abstract
OBJECTIVE: This study aimed to develop and validate a prognostic nomogram integrating clinical and laboratory variables to predict prolonged hospital stay in patients undergoing surgery for gastrointestinal (GI) perforation, facilitating early risk stratification and informed clinical decision-making. PATIENTS AND METHODS: A retrospective retrospective single-center study included 164 surgical patients with GI perforation from 2022-2024. Variables encompassed demographics, perforation characteristics, and preoperative/postoperative laboratory markers. The least absolute shrinkage and selection operator (LASSO) regression identified key predictors, followed by multivariate logistic regression to construct a nomogram. Model performance was evaluated using the receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). RESULTS: Upper GI perforation (OR=2.93, 95% CI:1.23-6.98), smaller perforation diameter (OR=0.48, 95% CI:0.28-0.82), and lower preoperative albumin (OR=1.10 per unit increase, 95% CI:1.03-1.17) independently predicted prolonged hospitalization. The nomogram demonstrated good discrimination (training AUC=0.75; validation AUC=0.79) and calibration. DCA confirmed clinical utility, with net benefit surpassing "treat all" or "treat none" strategies across risk thresholds. CONCLUSION: In summary, we developed and validated a nomogram that effectively identifies patients at high risk for prolonged hospitalization after GI perforation surgery by integrating three routinely available clinical parameters. This tool aids in optimizing resource allocation and personalized perioperative management. Further multicenter validation is warranted to enhance generalizability and incorporate dynamic biomarkers.