IDDF2024-ABS-0011 Development and validation of a predictive model for acute bowel injury in patients with moderately severe acute pancreatitis and severe acute pancreatitis

<h3>Background</h3> In this study, we aimed to create and validate a nomogram prediction model for estimating the likelihood of acute bowel injury (ABI) in patients diagnosed with moderately severe acute pancreatitis (MSAP) and severe acute pancreatitis (SAP). <h3>Methods</h3> This retrospective study enrolled patients diagnosed with moderately severe acute pancreatitis (MSAP) and severe acute pancreatitis (SAP) at the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database from January 2008 to August 2019. Subsequently, a multivariate analysis was conducted. The patients were randomly assigned to a training group and a validation group in a 2:1 ratio. Candidate predictors were identified using the least absolute shrinkage and selection operator (LASSO) method in combination with machine learning techniques. Logistic regression analysis was utilized to identify the risk factors associated with acute bowel injury (ABI), and a predictive model was developed. Calibration plots were employed to evaluate the model’s consistency, while the receiver operating characteristic (ROC) curve was utilized to assess its predictive accuracy. Additionally, decision-curve analysis (DCA) was performed to evaluate the clinical utility of the model. <h3>Results</h3> This study included a total of 1,063 patients, of whom 711 were assigned to the training group and 352 to the validation group. Multiple logistic regression analysis identified CRP, IAP, and CysC as significant risk factors. The calibration curve validated the excellent concordance between predicted and observed outcomes. The ROC curve exhibited the strong predictive performance of the nomogram. DCA confirmed the favorable clinical utility of the nomogram. <h3>Conclusions</h3> The prognostic performance of the prediction model for ABI in MSAP and SAP patients is satisfactory. Employing this model can assist clinicians in categorizing patients for primary prevention and early therapeutic intervention, thereby enhancing prognosis.