A risk model for parenteral nutrition-associated liver disease in patients with severe acute pancreatitis

Abstract

Background and study aims

The aim of this study is to explore the risk factors for parenteral nutrition-associated liver disease (PNALD) in patients with severe acute pancreatitis by establishing a verification risk model.

Patients and methods

A total of 176 patients with severe acute pancreatitis from January 2019 to August 2021, were assigned into the observation group (n = 88) and control group (n = 88) based on the diagnostic results of PNALD, randomly. Their clinical data were recorded. Alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBIL) and alkaline phosphatase (ALP), etc., were detected. The logistic model and desicion tree model were used to analyze the risk factors.

Results

Patients in the observation group had higher levels of ALT, AST, TBIL, and lower level of ALP than those of control group (P < 0.05). Multivariate logistic regression analysis revealed that alcohol intake history, ALT ≥69.65 U/L, AST ≥71.27 U/L, TBIL ≥26.27 μmol/L and ALP ≤45.11 U/L were risk factors for PNALD. The levels of ALT and AST in observation group were two times as high as those in the control group, which conformed to the Danan’s criteria and accorded with the results of univariate analysis.

Conclusion

The regression model showed high consistency with the decision tree model in the prediction of risk factors. Alcohol intake history, ALT ≥69.65 U/L, AST ≥71.27 U/L, TBIL ≥26.27 μmol/L and ALP ≤45.11 U/L are risk factors for PNALD.