Construction Of Long-term Survival Nomogram Prediction Model For Patients With Chronic Heart Failure Based On Lasso Regression And Cox Proportional Risk Regression

Heart failure is the end stage of various cardiovascular diseases,its prevalence rate in China continues to increase,the prognosis is poor.Through long-term follow-up of patients with chronic heart failure and collection of various clinical data,this study established a nomogram prediction model of long-term survival in patients with chronic heart failure,accurately evaluated the long-term survival of patients with heart failure,and helped patients with heart failure realize individualized evaluation.ObjectivesTo explore the use of hematological indicators,electrocardiogram,echocardiogram,social demographic data and other indicators to establish a long-term survival nomogram prediction model for patients with chronic heart failure.MethodsBring into the clinical data of 235 patients with chronic heart failure from 2006 to 2009 in Shandong Provincial Qianfoshan Hospital.Long-term follow-up was conducted,with telephone follow-up every 3 months,record the occurrence of the outcome event.Clinical data were collected including blood indicators:creatinine,uric acid,blood urea nitrogen,hemoglobin,red blood cell distribution width,triglyceride,total cholesterol,low-density lipoprotein cholesterol,albumin,serum calcium,serum potassium,serum phosphorus;Parameters of echocardiography:left ventricular end diastolic dimension,left atrium dimension,left ventricular ejection fraction,interventricular septum,pulmonary artery systolic pressure,mitral regurgitation,tricuspid regurgitation,pleural effusion,pericardial effusion;ECG index:QRS duration,QTc interval;General characteristics:sex,age,mean heart rate,mean systolic blood pressure,mean diastolic blood pressure,smoking≥20 years,drinking≥20 years;Previous medical history:whether or not diabetes mellitus,hypertension and chronic gastritis,in total of 34 indicators.Based on SPSS 26.0 and R 4.1.3,the data set was divided into training set and validation set in a ratio of 7:3,and all continuous variables were divided into dichotomous variables and rank variables.The independent prognostic factors affecting long-term survival of patients with chronic heart failure were screened by combining Lasso regression with Cox proportional hazard regression analysis.A nomogram prediction model was established,and the differentiation and accuracy of the model were evaluated by concordance index,calibration curve and time-dependent receiver operating characteristic curve.ResultsThe one-year,five-year and ten-year survival rates were 77.0%,43.0%and 21.3%,respectively.A total of 7 independent predictors of long-term survival of patient with chronic heart failure were selected by Lasso regression with Cox proportional hazard regression,which were age,left ventricular end diastolic dimension,mitral regurgitation,tricuspid regurgitation,red blood cell distribution width,uric acid and blood urea nitrogen,and the prediction model was constructed and verified internally.The C-index of training set and validation set were 0.782 and 0.791,respectively.The AUCs of one-year,five-year and ten-year survival rates of training set and validation set were 0.813,0.829,0.843,0.848,0.876 and 0.885,respectively.The calibration curves showed good consistency.Conclusions1.Patients with chronic heart failure have poor prognosis and low 10-year survival rate.2.Age,left ventricular end diastolic dimension,mitral regurgitation,tricuspid regurgitation,red blood cell distribution width,blood urea nitrogen and uric acid were all independent predictors of long-term survival in patients with chronic heart failure.3.As a quantifiable clinical scoring tool,the nomogram prediction model established in this study can predict the long-term survival rate of patients with chronic heart failure relatively accurately,and has advantages of convenience and visualization,providing a reliable scheme for clinicians to judge the condition of patients with chronic heart failure and provide individualized treatment plans.