Paragomi P, Hinton A, Pothoulakis I, et al. - The authors present a modified model that eliminates the Morphine Equivalent Dose (MED) component and predicts severe acute pancreatitis (AP) accurately while maintaining significantly distinct early trajectories.

• The sample consisted of 1,393 participants (median age of 49 years; 51% males).

• There were 950 mild (68.2%), 315 (22.6%) moderately severe, and 128 (9.2%) severe AP cases in the study cohort.

• At each study time-point, mild cases had the lowest pancreatitis activity scoring system (PASS).

• A subset of patients was found who had outlier admission PASS values.

• In the outlier group, the MED was responsible for 70% of the PASS variation, and 66% of these patients were from US centers.

• Among the four modified models, the mPASS-1 (without MED from PASS) performed well in predicting severe AP, with an AUC of 0.88 and separate trajectories with distinct slopes amongst severity subgroups.