The diagnosis model of stuck-toxin syndrome of gastric cancer was studied by logistic regression and ROC curve based on syndrome differentiation
Objective To explore the diagnostic model of gastric cancer stasis-toxicity syndrome based on evidence-based diagnosis using logistic regression and ROC curve.Methods The information from January 2023 to December 2023 of 97 gastric cancer patients was collected from the four diagnoses of traditional Chinese medicine,and the patients were classified into stasis-toxicity syndrome and non-stasis-toxicity syndrome.The multifactorial correlation between stasis-toxicity syndrome of gastric cancer and TNM staging,coagulation function,and serum inflammatory factor was investigated,and a diagnostic model of stasis-toxicity syndrome of gastric cancer was established by using logistics regression,and the ROC curves were used for the evaluation and analysis.Results The results showed that the TNM staging of patients with gastric cancer stasis-toxicity evidence was later than that of patients without stasis-toxicity evidence,and D-dimer(D-D),plasma fibrinogen(FIB),serum inflammatory factors interleukin-1(IL-1),C-reactive protein(CRP)and interleukin-6(IL-6)were higher than those of patients with non-stasis-toxicity evidence in patients with stasis-toxicity evidence,and the tumor necrosis factor(TNF)of patients with stasis-toxicity evidence was lower than that of patients without stasis-toxicity evidence.patients with stasis-toxicity syndrome.The predictive function model of stasis-toxicity syndrome was established based on the objective quantitative indexes of coagulation function and serum inflammatory factor:P=1/[1+exp(25.599-1.917×D-D-2.809×IL-1-1.984×CRP-1.508×TNM staging)].The area under the ROC curve of the predictive function model was 0.980,the sensitivity was 96.2%,the specificity was 77.8%,and the overall diagnostic concordance rate was 92.8%,and the predictive function model had a high degree of agreement.Conclusion The diagnostic model of gastric cancer stasis-toxicity evidence established by using logistic regression and ROC curve has a high coincidence rate,which is of certain reference value for improving the accuracy of identification and treatment of gastric cancer stasis-toxicity evidence.
万方数据
维普
