• Jun Dong, Yuhang Shen, Zhihuai Wang, Jiankang Zhang, Xihu Qin, Chunfu Zhu, Yuan Gao, and Qiang Yu.
• Department of Hepato-biliary-pancreatic Surgery, The Affiliated Changzhou No. 2 People's Hospital of Nanjing Medical University, Changzhou, China.
• Medicine (Baltimore). 2024 Apr 26; 103 (17): e37911e37911.

AbstractHypertriglyceridemia is a common cause of acute pancreatitis (AP). Fatty liver, a manifestation of metabolic syndrome, is related to the severity of AP. The present study aimed to construct an accurate predictive model for severe AP (SAP) by combining the fatty liver infiltration on a computerized tomography (CT) scan with a series of blood biomarkers in patients with hypertriglyceridemia-associated AP (HTG-AP). A total of 213 patients diagnosed with HTG-AP were included in the present retrospective study. Clinical information and imageological findings were retrospectively analyzed. The model was constructed from independent risk factors using univariate analysis, the least absolute shrinkage and selection operator method. Subsequently, the data from the training group of 111 patients with HTG-AP was analyzed using logistic regression analysis. The efficacy of the model was verified using an external validation group of 102 patients through the receiver operating characteristic curve (ROC). Independent predictors, including serum calcium, C-reactive protein, lactate dehydrogenase and liver-to-spleen CT attenuation ratio (L/S ratio), were incorporated into the nomogram model for SAP in HTG-AP. The model achieved a sensitivity of 91.3% and a specificity of 88.6% in the training group. Compared with the Ranson model, the established nomogram model exhibited a better discriminative ability in the training group [area under the curve (AUC): 0.957] and external validation group (AUC: 0.930), as well as better calibration and clinical benefits. The present study demonstrates that the constructed nomogram based on CT findings and blood biomarkers is useful for the accurate prediction of SAP in HTG-AP.Copyright © 2024 the Author(s). Published by Wolters Kluwer Health, Inc.

32 keywords...

hide…