Construction and validation of a predictive model for in-hospital mortality in elderly patients with severe acute pancreatitis
- VernacularTitle:中老年重症急性胰腺炎患者住院期间死亡预测模型的构建与验证
- Author:
Jiaxu ZHANG
1
;
Ting YANG
;
Ran LI
Author Information
- Publication Type:Research Article
- Keywords: middle-aged and elderly patients; severe acute pancreatitis; mortality; predictive model; renal insufficiency
- From: Journal of Clinical Medicine in Practice 2024;28(17):51-55
- CountryChina
- Language:Chinese
-
Abstract:
Objective To develop and validate a predictive model for in-hospital mortality in elderly patients with severe acute pancreatitis (SAP). Methods A total of 368 elderly SAP hospitalized patients were selected as study objects, and were divided into mortality group (96 patients, 26.09%) and survival group (272 patients, 73.91%) based on their survival status during hospitalization. Multivariable Logistic regression analysis was performed to identify influencing factors associated with in-hospital mortality in SAP patients, and a predictive model was constructed based on these factors. Receiver operating characteristic (ROC) curves were plotted, and the predictive performance of the model was evaluated using the area under the curve (AUC), accuracy, sensitivity, and specificity. Results Univariate analysis revealed that the mortality group had a higher proportion of patients aged over 60 years, with renal insufficiency, coronary heart disease, and undergoing laparoscopic surgery. Additionally, the mortality group had significantly higher levels of red blood cell distribution width, fasting blood glucose, interleukin-6, procalcitonin, neutrophil-to-lymphocyte ratio (NLR), lactate, modified CT severity index (MCTSI) score, and bedside index for severity in acute pancreatitis (BISAP) score compared to the survival group (
P < 0.05). Multivariable Logistic regression analysis identified age, renal insufficiency, MCTSI score, fasting blood glucose, and NLR as independent influencing factors of in-hospital mortality in elderly SAP patients (P < 0.05). A regression equation was constructed based on these factors: C-index=-1.569+0.258×(age)+0.334×(renal insufficiency)+0.672×(MCTSI score)+0.281×(fasting blood glucose)+0.410×(NLR). The ROC curve analysis showed that the AUC of the model for predicting in-hospital mortality in elderly SAP patients was 0.877 (95%CI, 0.840 to 0.915), with an accuracy of 84.23%, sensitivity of 75.00%, and specificity of 87.50%. Conclusion Age, renal insufficiency, MCTSI score, fasting blood glucose, and NLR are independent predictors of in-hospital mortality in elderly SAP patients. The predictive model constructed based on these factors can assist in identifying high-risk patients and predicting all-cause mortality risk in SAP.