Development and validation of a nomogram to predict massive bleeding requiring intervention in severe acute pancreatitis.

Background: Hemorrhage remains a formidable complication of severe acute pancreatitis (SAP), with a high mortality rate. However, there is currently no effective method for identifying SAP patients who are at high risk for massive bleeding. The present study aimed to explore risk factors for predicting massive bleeding in SAP patients and to develop a predictive nomogram, which could facilitate early prediction, and timely appropriate interventions.

Methods: We conducted a multivariate logistic regression analysis to examine the relationship between massive bleeding and variables including patient demographics, disease severity, laboratory indexes and local pancreatic complications. A novel nomogram was constructed based on these factors, and was validated both internally and externally assessing its discrimination, calibration, and clinical applicability.

Results: The study involved 351 patients in the training cohort, 104 patients in the internal validation cohort, and 123 patients in the external validation cohort. Logistic regression analysis identified several independent risk factors for massive bleeding, including computed tomography severity index score above 8 points, Acute Physiology and Chronic Health Evaluation II score greater than 16 points, abdominal compartment syndrome, pancreatic fistula, and sepsis. The nomogram constructed from these factors yielded an area under the receiver operating characteristic curve (AUC) of 0.896 and a coefficient of determination (R²) of 0.093. The Hosmer-Lemeshow test indicated good model fitness (P = 0.654). Furthermore, the nomogram demonstrated reliable performance in both validation cohorts.

Conclusions: The nomogram showed strong predictive capability for massive bleeding and could be a valuable tool for clinicians in identifying SAP patients at high risk for this complication at an early stage.