Construction Of A Predictive Model For Postoperative Prognosis And Inflammatory Nutritional Indicators In Patients With Colorectal Cancer

Objective: This paper aims to establish a predictive model of inflammatory nutritional indicators in patients after colorectal cancer patients and validate the clinical value of this prediction model in the prognosis of patients after colorectal cancer patients。Methods: A total of 462 patients with colorectal cancer diagnosed by pathology in the First Affiliated Hospital of Dali University from January 2014 to January 2022 were collected retrospectively,and a total of 462 subjects met the criteria according to the inclusion criteria and exclusion criteria.Preoperative general data and postoperative pathological data were collected.The number of all eligible cases was randomly divided into two groups according to7:3: the training set and the validation set.In the training set,univariate regression was selected for variable screening,multivariate COX analysis was used to screen variables,independent risk factors for colorectal cancer prognosis were obtained,and a nomogram was drawn to construct a risk prediction model,the consistency index(C-index)and subject working curve(ROC)were used to evaluate the model discriminant ability,the calibration curve was used to evaluate the calibration ability,and the decision analysis curve was used to evaluate the clinical benefit of the model.Outcome:1.The 462 patients included were randomly divided into 329 training sets and 133 validation sets according to the ratio of 7:3,and the differences between the variables of the training set and the validation set were compared,except for "diabetes history",the remaining variables(P>0.05),and the data of the two groups were statistically significant.2.In the training set,univariate COX analysis: age,diabetes,BMI index,TNM stage(III,IV),N stage(N1,N2),M stage,CEA,PLR,SII and other variables(P<0.05),indicating that the above variables are related to the prognosis of colorectal cancer patients.Multivariate COX analysis: TNM stage(stage IV),M stage,CEA,PLR,SII,diabetes and other variables(P<0.05),which are independent risk factors affecting the prognosis of colorectal cancer patients,and a nomogram prediction model was established on this basis.3.The 1-year,3-year and 5-year survival rates of colorectal cancer patients predicted by the nomogram model tend to be consistent with the actual results,and the prediction accuracy is good.Comparing the C-index index,AUC area and clinical decision curve(DCA curve),the results show that the prediction performance of the new predictor index is more accurate than that of traditional TNM staging,and has clinical application value.Conclusion:1.TNM stage(stage IV),M stage,CEA,PLR,SII,diabetes are independent risk factors for the prognosis of colorectal cancer patients.2.The C-index index of the nomogram model was 0.747/0.753 and 0.701/0.711 in the prediction model and TNM stage,respectively,and the prediction accuracy of the model was high,which provided a visual reference for clinical evaluation of the prognosis of colorectal cancer patients.3.TNM stage(stage IV),M stage,CEA,PLR,SII,diabetes and other variables to comprehensively predict five-year survival in colorectal cancer patients is more accurate than traditional TNM staging.