Development and Validation of a Predictive Model for Predicting Acute Kidney Injury in Patients with Acute Pancreatitis

Background: Acute kidney injury is a serious complication of acute pancreatitis and significantly increases the risk of mortality. Early identification of acute kidney injury is important for medical interventions and care options. The aim of this study is to develop a predictive model that could rapidly identify high-risk population of acute kidney injury in patients with acute pancreatitis.

Methods: Totally 808 patients with acute pancreatitis admitted to our center from January 2015 to May 2019 were included in the study and were divided into the training (n=566) and validation (n=242) cohorts randomly, in a 2:1 ratio. The least absolute shrinkage and selection operator (LASSO) regression was used for data dimension reduction and feature selection then, multivariable logistic regression analysis was used to develop the prediction model. The performance of this nomogram was evaluated with calibration and validated in the validation set. Decision curve analysis was applied to evaluate the clinical usefulness of this model.

Results: Five potential predictors (BMI, RANSON score, serum uric acid, triglycerides and lactate) from 53 high dimensional clinical variables were incorporated to develop the prediction model of acute kidney injury. The nomogram demonstrated valuable prediction performance with AUROC of 0.994 and 0.996 in the training and validation cohorts, respectively. Individual risk probability was visually scored. The nomogram achieved fine calibration and good clinical usefulness.

Conclusion: The proposed nomogram can help to identify high-risk population of acute kidney injury in patients with acute pancreatitis and facilitate timely individualized clinical decision making.


Yuling Li*, Jian Kang , Hui Wang , Dongliang Yang , Li Zhao , Chao Wen , Xiujie Zhang , Jing Song and Dongna Gao

Abstract | Full-Text | PDF

Share this  Facebook  Twitter  LinkedIn  Google+


agar io


wormax io