Predictive model for progression to severe acute pancreatitis in patients without organ failure at admission

A significant subset of patients with acute pancreatitis (AP) who do not present with organ failure (OF) at admission nonetheless progress to severe acute pancreatitis (SAP). We aimed to develop and validate a simple scoring system to predict progression to SAP specifically in AP patients without initial OF.

This study utilized a prospectively maintained AP database. Patients admitted without OF were included. Variable selection employed multivariable logistic regression, Boruta algorithm, and LASSO regression. A nomogram was developed, from which a simplified ACR score was derived.

Of 3,813 eligible AP patients without OF at admission, 458 (12.0%) progressed to SAP. Six variables were independently associated with SAP progression. The resulting nomogram demonstrated strong discrimination, with AUCs of 0.834 (training), 0.833 (internal test), and 0.885 (external validation). The simplified ACR score (range 0-9) showed comparable performance (AUC 0.827, 95% CI: 0.807-0.846) and was significantly superior to APACHE II (AUC 0.655, p < 0.001), SIRS (AUC 0.732, p < 0.001) and BISAP (AUC 0.718, p < 0.001). Stratification into low- (0-3 points), intermediate- (4-6 points), and high-risk (7-9 points) groups revealed sharply increasing SAP incidence (2.9%, 16.5%, and 52.1%, respectively; p < 0.001) and worsening clinical outcomes.

The ACR score, comprising six readily available clinical parameters, effectively stratifies the risk of SAP progression in AP patients without organ failure at admission.