| Objective:Endoscopic retrograde pancreatic angiography(ERCP)has been widely used in the diagnosis and treatment of biliary pancreatic diseases,especially the treatment of biliary obstruction caused by malignant biliary diseases,ERCP stent placement can open the biliary tract to achieve palliative treatment.Due to the fact that patients themselves suffer from malignant diseases and the high difficulty of ERCP operation,the incidence of postoperative complications is relatively high.At present,there are more studies on postoperative pancreatitis and less studies on postoperative cholangitis,and the postoperative cholangitis of ERCP seriously affects the postoperative quality of life and prognosis of patients.Therefore,the purpose of this study want to analyze the risk factors that may lead to cholangitis in patients with malignant biliary obstruction after stent implantation,and to construct a clinical prediction model to predict the probability of cholangitis in patients with malignant biliary obstruction,so as to guide individualized ERCP treatment in patients with malignant biliary obstruction.Methods:Clinical data of 311 patients with malignant biliary obstruction who received ERCP stent implantation in our hospital from January 2016 to June2022 were collected.Factors that may lead to the occurrence of cholangitis were included in the study,such as gender,age,previous diseases(diabetes,hypertension),past smoking and alcohol use,and the location of the patient’s biliary obstruction,whether difficult intubation occurred during the operation,stent placement type,laboratory data and other relevant data.The above influencing factors were analyzed by univariate and multivariate Logistic regression,and statistically significant independent risk factors were obtained.Then,a nomogram prediction model was established.Individualized quantitative scores were performed in the prediction model according to whether patients had the above independent risk factors,and the probability of cholangitis was predicted according to the total score obtained.The area under ROC curve(AUC)was used to evaluate the prediction model differentiation,and the model consistency was verified by calibration curve.Results:1.Comparison of general dataAmong 311 patients with malignant biliary obstruction who received ERCP biliary stent implantation in our hospital,60 cases developed cholangitis and 251 cases did not.The incidence of cholangitis was 19.29%.The general data(age,sex)were comparable and did not differ significantly(P>0.05),and no significant difference was found in diabetes or hypertension(P>0.05);Comparison of obstruction sites showed that high obstruction was more likely to lead to cholangitis(P<0.05).2.Univariate and multivariate Logistic regression analysisUnivariate analysis showed that gamma-glutamyl transpeptidase(GGT,P=0.001),albumin(P<0.001),difficult intubation(P<0.001),obstruction site(P<0.001),smoking(P<0.001)is influential factors for cholangitis in patients with malignant biliary obstruction after ERCP stenting.Multivariate analysis of the above influencing factors showed that GGT(P=0.005,OR 1.001,95%CI 1.000-1.001),albumin(P=0.004,OR 4.645,95%CI 1.644-13.129),difficult intubation(P<0.001,OR 18.859,95%CI 4.807-73.992),obstruction site(P<0.001,OR 5.723,95%CI 2.587-12.663)and smoking(P=0.002,OR3.124,95%CI 1.525-6.398)were independent risk factors for cholangitis.3.Model establishment and verificationMultivariate Logistic regression analysis was carried out to find statistically significant risk factors and a nomogram prediction model was established.According to the proportion of each risk factor,GGT level and difficult intubation showed greater predictive ability for postoperative cholangitis after ERCP,followed by obstruction site,albumin level and smoking status.The probability of cholangitis can be predicted according to whether the patients have the above risk factors and the individualized score can be made in the model,and the individualized preventive measures can be taken according to the probability of cholangitis.The model differentiation was evaluated by the ROC curve,which was plotted according to the predicted probability of cholangitis in each patient.The area under the ROC curve(AUC)was 0.837(95%CI: 0.777-0.896,P<0.001),indicating that the model has a high degree of differentiation.The Calibration curve is adopted to evaluate the consistency.If the predicted probability is between 20% and 40%,the predicted probability will be slightly larger than the actual probability.The prediction that the probability of cholangitis in this patient is over 40% is consistent with the actual probability,indicating that the model has a good predictive efficiency.Conclusion:1.For patients with malignant biliary obstruction,ERCP biliary stent implantation is an effective treatment for biliary obstruction,but it has a high probability of postoperative cholangitis.For such patients,comprehensive evaluation should be carried out according to the patient’s condition and relevant examinations,and surgeons should strictly grasp the indications and contraindications of ERCP,so as to ensure the surgical safety and postoperative quality of life of patients.2.For patients with malignant biliary obstruction with ERCP stent implantation,high GGT,low albumin,difficult intubation,high biliary obstruction,and smoking were independent risk factors for cholangitis.Multiple factors independently or cooperatively lead to the occurrence of cholangitis in patients with malignant biliary obstruction after ERCP stenting.The occurrence of cholangitis in patients with malignant biliary obstruction after ERCP stenting can be reduced as far as possible by improving the patient’s general state before ERCP,performing effective biliary decompression and unobstructed biliary drainage during the operation,and strengthening the aseptic operation of ERCP at the same time.3.A nomogram prediction model was established according to the independent risk factors.The results of model differentiation and consistency evaluation showed that the nomogram prediction model had good predictive efficiency.The risk factors of the patient can be scored in the nomogram prediction model and the total score can be obtained to predict the probability of the patient developing cholangitis and take individualized preventive measures according to the probability of the patient developing cholangitis. |
PDF Full Text Request