Risk Prediction Model Of Clinical Symptom Onset Of Coronary Heart Disease Patients And Degree Of Coronary Stenosis

Objective:The incidence of coronary heart disease in the population is increasing,and the symptoms of some coronary heart disease patients have repeatedly attack.To analyze the factors that may affect coronary heart disease and to construct a risk prediction model for the symptom onset and stenosis of patients with coronary heart disease by multivariate Logistic regression analysis,so as to provide reference for the treatment of coronary heart disease.Methods:The clinical data of 283 patients with coronary heart disease treated in the Department of Cardiology of the Second Hospital of Shandong University from June 2020 to September 2020 and 192 normal people in the examination center were analyzed retrospectively.Logistic regression method was used to analyze the factors that may affect the clinical symptom onset and stenosis of coronary heart disease,and the Receiver operating characteristic curve(ROC)was drawn.The linear regression was adopted for diagnose the multiple collinearity among the prognostic factors.According to the results of multi-factor Logistic regression,the prediction model was established,and the optimal threshold was determined by ROC.Results:Univariate logistic analysis showed that sex,age,body mass index(BMI),lymphocyte count,monocyte count,red blood cell count,hemoglobin concentration,hematocrit,average hemoglobin content,red cell distribution width(RDW)coefficient of variation,RDW standard deviation,platelet count,mean platelet volume,platelet volume distribution(PDW),Platelet-large cell ratio(P-LCR),alkaline phosphatase,direct bilirubin(DBIL),total protein,albumin,albumin/globulin,glucose,creatinine,total cholesterol(TC),high density lipoprotein-cholesterol(HDL-C),Estimated glomerular filtration rate(eGFR),neutrophil count/lymphocyte count,derived neutrophil count/lymphocyte count,monocyte count/HDL-C,TC/HDL-C,non-HDL-C/HDL-C and LDL-C load was statistically significant between the coronary heart disease clinical symptom onset group and the control group.White blood cell count,neutrophil count,DBIL,creatinine,CystatinC(CysC),carbon dioxide binding capacity,HDL-C,apolipoprotein Al(ApoA1),homocysteine,B-type natriuretic peptide,fibrinogen(Fbg),monocyte count/HDL-C,apolipoprotein B,ApoB/ApoA1,globulin/prealbumin,Fbg/albumin,Fbg/prealbumin and E peak flow velocity was statistically significant between the low and high Gensini scores group of coronary heart disease.Through linear regression,collinearity diagnosis and exclusion index,and further through multi-factor logistic regression analysis,for the prediction of coronary heart disease clinical symptom onset,seven indexes such as sex,age,BMI,PDW,total protein,albumin and glucose were selected for model fitting.The fitting regression model is as follows:P=1/1+e-y,Y=7.666-1.033X1+0.068X2+0.218 X3-0.289 X4-0.081 X5-0.175X6+0.211X7.The area under curve is 0.918,and the prediction accuracy is 87.7%.For the prediction of the degree of stenosis of coronary heart disease,globulin,ApoA1,interventricular septal thickness,left ventricular ejection fraction and IBIL were selected for model fitting.The fitted regression model was as follows:P=1/1+e-y,Y=3.492+0.286X1-4.095X2+0.536 X3-17.888X4+0.352 X5,the area under curve was 0.773,and the prediction accuracy was 80.9%.Conclusion:According to the results of this model,it has great reference value for sex,age,BMI,PDW,total protein,albumin and glucose for predicting the coronary heart disease clinical symptom onset,and globulin,ApoA1,interventricular septal thickness,left ventricular ejection fraction and IBIL for predicting the degree of stenosis of coronary heart disease.