Predictors of acute pancreatitis after endoscopic transpapillary interventions

Objective. To develop a model for predicting the individual risk of developing acute post-manipulation pancreatitis before performing endoscopic transpapillary intervention. Material and methods. A cohort study of the results of endoscopic transpapillary interventions in 366 patients was conducted. For comparative analysis, patients in the general sample were divided into groups depending on gender, age, presence or absence of obesity, obstructive jaundice and chronic pancreatitis. Logistic regression analysis was used to build a prognostic model, and ROC analysis was used to stratify patients according to the degree of risk of developing acute post-manipulation pancreatitis. Results. Acute post-manipulation pancreatitis significantly more often complicated the early postoperative period in women than in men (12.5% vs 4.3%, p=0.028). Also, this complication was significantly more often diagnosed in patients under 40 years of age, compared with patients 41—60 years of age and over 60 years of age (30.4% vs 14.5% (p=0.018) and 30.4% vs 3.1% (p<0.001), respectively). At the same time, chronic pancreatitis was a significant factor in reducing the risk of this complication (p=0.017). Based on the results of logistic regression analysis, a forecasting model for acute post-manipulation pancreatitis was developed. The specificity and sensitivity of the model were 71.2% and 73.1%, respectively. Conclusion. Female gender, young age and the absence of chronic pancreatitis significantly increase the risk of developing acute post-manipulation pancreatitis. The proposed prediction model, including variables, allows us to identify patients at high risk of developing acute post-manipulation pancreatitis before endoscopic transpapillary intervention.