Clinical Prediction Model Of Colorectal Adenomatous Polyps Based On NLR,TyG Index And FOBT

Objective The aim of this study was to analyze the risk factors of colorectal adenomatous polyps(CAP),establish a predictive columnar graph model and validate its predictive value,and provide a meaningful guideline basis for risk stratification and population screening of CAP.Methods Retrospective analysis of patients hospitalized for colonoscopy from May 2017 to May 2022 at Chaohu Hospital,Anhui Medical University as study subjects,and clinical data were extracted,which included gender,age,body mass index(BMI),history of previous smoking and alcohol consumption,white blood cell count,absolute neutrophil count,absolute lymphocyte count,total cholesterol,triglycerides,fasting glucose,low-density Lipoprotein cholesterol,HDL cholesterol,CEA,fecal occult blood test(FOBT),neutrophil count lymphocyte count ratio(NLR)and triglyceride glucose(TyG)index were calculated according to the relevant formula.Eligible patient data were selected according to inclusion and exclusion criteria.Patients were randomly divided into training and validation cohorts according to 7:3,and the differences in general information between the two cohorts were first compared,followed by screening variables using one-way and multi-way logistic regression,and finally clinical prediction models were developed using the R software to plot column line graphs(Nomogram),and subject operating characteristic(ROC)curves,C-index(C-index),and calibration curves(C-index)were used in the training and validation cohorts,respectively.C-index,calibration curve,and decision analysis curve(DCA)were used to test the discrimination,calibration,and clinical usefulness of the model in the training and validation cohorts,respectively.Results A total of 662 patients were included in this study,including 464 in the training cohort and 198 in the validation cohort,and in both cohorts were grouped according to pathological findings,with confirmed CAP in the adenoma group and no significant endoscopic abnormality as the control group.49.4% and 54% of the patients in the training and validation cohorts,respectively,had CAP,and the difference between them was not statistically significant(P=0.269).After univariate and multifactorial logistic regression analysis,three independent predictors of CAP were identified,including: NLR(OR=1.362,95% CI:1.072-1.729,P=0.011),TyG index(OR=1.684,95% CI:1.189-2.557,P=0.003),FOBT positivity(OR=1.652,95% CI:1.068-1.938,P=0.024),and prediction models were constructed based on these three variables and visualized in the form of column line plots.The area under the ROC curve was 0.665 and 0.705 for the training and validation cohorts,respectively,and the Calibration calibration plots showed good agreement between the predicted and actual risk of occurrence for the training and validation cohorts.the DCA curves also demonstrated the clinical utility of the column line plots within a certain range.Conclusion NLR,TyG index,and FOBT positivity are independent risk factors for the occurrence of CAP.The clinical prediction model constructed in this study has good discrimination,calibration,and clinical utility,which is beneficial for clinical screening of high-risk groups,more targeted colonoscopy,improving the detection rate of adenomas,and reducing the occurrence of colorectal cancer(CRC).