Establishment And Verification Of A Nomogram To Predict The Risk Of Missed Diagnosis Of Colorectal Polyps

Objective:To analyze the risk factors for missed diagnosis of colorectal polyps,and establish and verify a nomogram model for predicting the risk of missed diagnosis of colorectal polyps after a single colonoscopy.Methods:The clinical data of patients who underwent two colonoscopies within 3 months in the Affiliated Hospital of North Sichuan Medical College from February 2017 to August 2019 were retrospectively retrieved.In the training set,single factor and multivariate Logistics regression was used to analyze the independent risk factors of colorectal polyps missed diagnosis,and Lasso regression was used to screen the risk factors as a double verification of Logistic regression;the alignment to predict the risk of colorectal polyps missed diagnosis was established through R software In the figure,the Bootstrap method is used to verify the nomogram internally,and the verification set is used for external verification.The C-index,receiver operating characteristic(ROC)curve and Calibration curve are used to evaluate the predictive performance of the nomogram model.Evaluate the clinical efficacy of the nomogram through the sensitivity,specificity and other indicators under the optimal cut-off value,and the application of clinical decision curve.Results:Age,physician’s experience,bowel preparation,observation meth-ods,withdrawal time,number of right colon polyps,number of 6-9mm polyps and ≥10mm polyps are independent risk factors for missed diagnosis of colorectal polyps,which are consistent with the risk factors screened by lasso regression.The C-index of the nomogram is 0.763(95%CI:0.721～0.805)and 0.716(95%CI:0.647～0.785)in the training set and validation set,respectively.The calibration curve fitted well.The optimal critical value of nomogram calculated by Youden index was 188.2.The sensitivity,specificity,positive predictive value and negative predictive value of the critical value are 74.4%,68.0%,41.6%,and 89.7%in the training set,respectively,and in the validation set.62.5%,72.3%,45.9%and 83.7%.Clinical decision curve analysis suggests that within the threshold probability range of 0.15 to 0.60,the use of this model can benefit some patients.Conclusion:The nomogram based on the independent risk factors of missed diagnosis of colorectal polyps has good discrimination,calibration and clinical efficacy analysis.It has a certain reference value for medical workers to analyze the risk of missed diagnosis of colorectal polyps individually,identify the high-risk population of missed diagnosis,and formulate appropriate and timely follow-up strategies.