Abstract:

Objective To explore the factors influencing the occurrence of acute acalculous cholecystitis (AAC) in patients with severe acute pancreatitis (SAP), and construct a prediction model. Methods The clinical data of 220 patients with SAP were retrospectively collected. Patients were divided into the AAC group (64 cases) and the non-AAC group (156 cases) according to whether AAC occurred during hospitalization. The general clinical data including age, gender, smoking, drinking, body mass index (BMI), cause of SAP, fasting time, underlying diseases and concurrent sepsis were compared between the two groups. Laboratory indexes [systemic inflammatory response index (SIRI), C-reactive protein to albumin ratio (CAR), white blood cell count to mean platelet volume ratio (WMR)]were also compared between the two groups. Multivariate Logistic regression was used to analyze the factors influencing the occurrence of AAC in patients with SAP during hospitalization. Based on the influencing factors, a nomogram prediction model was constructed. The performance of the prediction model was evaluated using the receiver operating characteristic (ROC) curve, the calibration curve and clinical decision analysis (DCA). Results Compared with the non-AAC group, the AAC group showed a longer fasting time, a higher proportion of combined sepsis and higher levels of TBil, SIRI, CAR and WMR at admission (P<0.05). Multivariate Logistic regression analysis showed that long fasting time, high levels of SIRI, CAR, WMR and TBil were independent risk factors for AAC in patients with SAP (P<0.05). Based on the five variables, a regression equation was developed as follows: Logit (P)=-11.364+0.444×fasting time+0.217×SIRI+0.278×CAR+1.869×WMR+0.053×TBil, and a nomogram prediction model was constructed. ROC curves showed that the area under the curve (AUC) and 95%CI, sensitivity and specificity of the nomogram risk prediction model for predicting AAC were 0.872 (95%CI: 0.824-0.921), 76.56% and 87.18%, respectively. Bootstrap validation results showed good consistency between the calibration curve and the actual curve, and Homser-Lemeshow test showed good goodness of fit. DCA showed that the nomogram prediction model could bring clinical benefits to patients when the threshold was between 0.04 and 0.99. Conclusion The nomogram prediction model constructed based on these risk factors demonstrates good predictive performance and can provide reference for early identification of patients at high risk of AAC in clinical practice.

Key words: pancreatitis, acalculous cholecystitis, risk factors, nomograms, severe acute pancreatitis, prediction model