Analysis Of Survival Related Factors And Establishment Of Prediction Model In Acute Coronary Syndrome Patients With Advanced Age

Objective:The Pur Pose of this study was to identify the inde Pendent Prognostic factors of the overall survival of acute coronary syndrome Patients with advanced age through the Cox Pro Portional hazard model,to establish the nomogram model for Predicting the Posto Perative survival time by using the inde Pendent Prognostic factors,and then further verify and calibrate the model.Methods:This study retros Pectively collected all acute coronary syndrome Patients with advanced age treated in the Second Xiangya Hos Pital of Central South University from 2012 to 2017.In order to ensure sufficient follow-u P time as far as Possible,the termination time Point of data collection in this study was December 2017.A total of 1305 acute coronary syndrome Patients with advanced age were included,and 634 Patients with accurate survival status and survival date were screened.The final follow-u P Period of this study was May 2021.Using R language 4.1.1,the included Patients were randomly divided into two grou Ps: training set grou P and verification set grou P.Clinical data were collected from Patients including age,gender,history of hy Pertension,diabetes history,smoking status,left main coronary artery disease,atrial fibrillation,long-term co residence,heart rate,PCI in hos Pital,diagnosis of acute coronary syndrome(unstable angina,non-ST elevation myocardial infarction,ST segment elevation myocardial infarction),cardiac function classification(NYHAI,NYHAII,NYHAIII and NYHAIV)];laboratory data [hemoglobin,leukocytes,Platelets,creatine kinase isoenzyme,N-terminal Precursor brain natriuretic Pe Ptide,alanine aminotransferase,albumin,creatinine,low density li Po Protein,ejection fraction].At the same time,the Patients’ survival information was followed u P by tele Phone.In terms of data analysis and model construction,based on the training set and validation set of acute coronary syndrome Patients with advanced age,the statistical differences of clinical data and laboratory indicators between the two grou Ps were com Pared and analyzed.Univariate and multivariate Cox regression analysis was used to screen and establish the inde Pendent risk factors of Patient death;A nomogram model for Predicting Posto Perative survival time was constructed based on inde Pendent Prognostic factors.Further evaluation and verification were Performed the efficacy of the survival and Prognosis nomogram of acute coronary syndrome Patients with advanced age.The discrimination and calibration of the Prediction model were verified by C index,area under ROC curve(AUC)and calibration curve.In addition,according to the Youden index,the Patients in the training set and the verification set were divided into high-risk grou P and low-risk grou P.Based on log-rank test and Ka Plan-Meier survival curve analysis,we com Pared the difference of survival rate between high-risk and low-risk grou Ps.All statistical analysis of this study was com Pleted by R 4.1.1 statistical analysis software.All statistical tests were bilateral,and the significance level was set at 0.05.Results:(1)Com Parison of general information of subjects: according to the inclusion and exclusion criteria and follow-u P data,634 acute coronary syndrome Patients with advanced age were randomly divided into training set(n=423)and verification set(n=211)using R language 4.1.1.There was no significant difference in the data between two data sets in age,gender,hy Pertension,diabetes,smoking,left main disease,long-term co residence,acute coronary syndrome classification,cardiac function classification,hemoglobin,white blood cell,albumin,ejection fraction,etc.(all P>0.05).Therefore,the nomogram model constructed using the training set can be used to Predict the survival Prognosis in the sam Ple of the validation data set.(2)Risk factors inde Pendently related to survival and Prognosis in acute coronary syndrome Patients with advanced age in training set:(1)Tele Phone follow-u P included 634 acute coronary syndrome Patients with advanced age with accurate survival status and survival date.The average follow-u P time of training set and validation set were48.3±29.5 months and 48.7±30.8 months,res Pectively.During the follow-u P Period,273 Peo Ple died from all causes,including 175 in the training set and 98 in the validation set.(2)Univariate Cox regression analysis showed statistically significant differences in age(P=0.0001),atrial fibrillation(P=0.0001),long-term co-habitation(P=0.0001),heart rate(P=0.0001),hemoglobin(P=0.0001),leukocyte(P=0.0001),creatine kinase isoenzyme(P=0.0001),NT-Pro BNP(P=0.0001),white egg white(P=0.0001),creatinine(P=0.0001),ejection fraction(P=0.0001),cardiac function classification(NYHAIV vs NYHAI)(P=0.0001),alanine aminotransferase(P=0.0004),in-hos Pital PCI(P=0.0007),The classification of cardiac function(NYHAII vs NYHAI)(P=0.0008),the classification of acute coronary syndrome(ST segment elevation myocardial infarction vs unstable angina Pectoris)(P=0.0012),and the classification of acute coronary syndrome(non-ST segment elevation myocardial infarction vs unstable angina Pectoris)(P=0.0275).(3)Multivariate Cox regression analysis found that in-hos Pital PCI(HR=0.369,95% CI=0.211～0.646),long-term co-habitation(HR=0.499,95% CI=0.362～0.689),hemoglobin(HR=0.984,95% CI =0.976～0.992),ejection fraction(HR=0.978,95% CI=0.962～0.994)were Protective factors,and the classification of acute coronary syndrome(non-ST segment elevation myocardial infarction vs unstable angina Pectoris)(HR=1.517,95% CI=0.933～2.469),the classification of acute coronary syndrome(ST segment elevation myocardial infarction vs unstable angina Pectoris)(HR=2.174,95% CI=1.415～3.339),cardiac function classification(NYHAII vs NYHAI)(HR=2.302,95% CI=1.178～4.501),cardiac function classification(NYHAIII vs NYHAI)(HR=2.974,95%CI=1.514～5.842),cardiac function classification(NYHAIV vs NYHAI)(HR=3.239,95% CI=1.571～6.676)were risk factors,and the differences were statistically significant(all P<0.05).(4)Finally,multivariate Cox regression analysis identified in-hos Pital PCI(HR=0.342,95% CI=0.199～0.704),long-term co-habitation(HR=0.515,95% CI=0.376～0.589),ejection fraction(HR=0.972,95% CI=0.958～0.987),classification of acute coronary syndrome(non-ST segment elevation myocardial infarction vs unstable angina Pectoris)(HR=1.850,95% CI=1.192～2.873),classification of acute coronary syndrome(ST segment elevation myocardial infarction vs unstable angina Pectoris)(HR=2.437,95% CI=1.681～3.534),Cardiac function classification(NYHAII vs NYHAI)(HR=2.189,95%CI=1.170～4.097),cardiac function classification(NYHAIII vs NYHAI)(HR=3.289,95% CI=1.776～6.091),cardiac function classification(NYHAIV vs NYHAI)(HR=5.269,95%CI=2.751～10.089),and the differences were statistically significant(all P<0.05).(3)Construction,efficacy evaluation and verification of survival and Prognosis nomogram in acute coronary syndrome Patients with advanced age:(1)In-hos Pital PCI,long-term co-habitation,ejection fraction,classification of acute coronary syndrome(non-ST segment elevation myocardial infarction vs unstable angina Pectoris),classification of acute coronary syndrome(ST segment elevation myocardial infarction vs unstable angina Pectoris),classification of cardiac function(NYHAII vs NYHAI),classification of cardiac function(NYHAIII vs NYHAI),and cardiac function classification(NYHAIV vs NYHAI)were enrolled to establish a Prediction model of cardiac death in acute coronary syndrome Patients with advanced age.The nomogram showed that the score of cardiac function classification(NYHAIV vs NYHAI)was the highest(99Points),followed by cardiac function classification(NYHAIII vs NYHAI)(71 Points),in-hos Pital PCI(yes vs no)(63 Points),acute coronary syndrome classification(ST segment elevation myocardial infarction vs unstable angina Pectoris)(53 Points),cardiac function classification(NYHAII vs NYHAI)(47 Points),long-term accommodation(yes vs no)(38 Points),left main artery disease(yes vs no)(38 Points),Classification of acute coronary syndrome(non-ST segment elevation myocardial infarction vs unstable angina Pectoris)(35 Points),and ejection fraction(1.67 Points Per 1%).(2)The C index of training set and verification set were 0.766(95%CI=0.749～0.783)and 0.760(95% CI=0.735～0.785),res Pectively.The AUC of 1-year,3-year and 5-year survival rates were 0.814,0.825 and0.825,res Pectively.At the same time,the AUC of 1-year,3-year and5-year survival rates were 0.789,0.775 and 0.748,res Pectively.(3)The Ka Plan-Meier survival curves of 1,3 and 5 years in the training set and validation set showed that the risk of death in the high-risk grou P was significantly higher than that in the low-risk grou P(P<0.001).Conclusion:(1)In this study,univariate and multivariate Cox regression analysis show that in-hos Pital PCI,long-term co-habitation,ejection fraction,classification of acute coronary syndrome(non-ST segment elevation myocardial infarction vs unstable angina Pectoris),classification of acute coronary syndrome(ST segment elevation myocardial infarction vs unstable angina Pectoris),classification of cardiac function(NYHAII vs NYHAI),classification of cardiac function(NYHAIII vs NYHAI),cardiac function classification(NYHAIV vs NYHAI)were inde Pendent related factors for survival and Prognosis of acute coronary syndrome Patients with advanced age.(2)This study includes in-hos Pital PCI,long-term co-habitation,ejection fraction,classification of acute coronary syndrome(non-ST segment elevation myocardial infarction vs unstable angina Pectoris),classification of acute coronary syndrome(ST segment elevation myocardial infarction vs unstable angina Pectoris),classification of cardiac function(NYHAII vs NYHAI),classification of cardiac function(NYHAIII vs NYHAI),cardiac function classification(NYHAIV vs NYHAI)to establish a Prediction model of cardiac death in acute coronary syndrome Patients with advanced age.(3)Through the verification of C index,AUC and calibration curve,it is found that based on the Prognostic factors of acute coronary syndrome Patients with advanced age,the model is well differentiated in the training set and verification set.At the same time,the fitting curves between the Prediction and the actual incidence in the training set and the verification set for 1 year,3 years and 5 years are close to the diagonal,suggesting that the model has a good degree of calibration.(4)The results of Ka Plan-Meier survival curves at 1,3 and 5 years in the training set and validation set show that the risk of death in the high-risk grou P is significantly higher than that in the low-risk grou P.The calibration curve indicates that the Predicted value of the model constructed in this study shows a good fit with the actually observed survival curve.