Risk Factors Analysis And Prediction Model Construction Of Left Ventricular Dysfunction After Heart Valve Replacement

Objective:To explore the risk factors of left ventricular dysfunction（LVD）in patients with heart valve replacement（HVR）,and establish a reliable and accurate clinical prediction model to provide evidence for early identification of high-risk patients with LVD after HVR.Methods:In this study,patients who received HVR in the Department of cardiac surgery of the first hospital of Lanzhou University from January 2020 to December 2020 were selected as the research objects.According to the inclusion and exclusion criteria,258patients were finally included.The relevant clinical data were collected.According to whether LVD occurred after operation,the patients were divided into LVD group and non LVD group.The independent risk factors of LVD after HVR were screened out by univariate analysis and multivariate logistic regression analysis,and the risk prediction model was established.The nomogram was drawn with R software,the receiver operating characteristic curve（ROC）and calibration chart were used to evaluate the prediction model,Hosmer-Lemeshow goodness of fit test was used to evaluate the goodness of fit of the prediction model,and bootstrap self-sampling method was used to internally verify the prediction model.Results:1.Univariate analysis showed that there were significant differences between LVD group and Non-LVD group in gender,history of atrial fibrillation,history of pulmonary hypertension,history of coronary heart disease,classification of cardiac function,total bilirubin,albumin,left ventricular end-diastolic diameter（LVEDD）,left ventricular end-systolic diameter（LVESD）,left atrial anterior-posterior diameter,left ventricular ejection fraction（LVEF）and cardiopulmonary bypass time（CPB）（P<0.05）.2.Multivariate logistic regression analysis showed that history of pulmonary hypertension,history of coronary heart disease,albumin<41.65g/l,LVEDD≥70.5mm,LVEF<54.5%,and CPB time≥125.5min were independent risk factors for LVD after HVR（P<0.05）.3.Construction of LVD risk prediction model after HVR:the above six independent risk factors were incorporated into the regression equation to successfully build a prediction model.The area under ROC curve（AUC）=0.888（95%CI:0.830-0.945）,and the best critical point is 0.621.At this time,the sensitivity is 79.5%and the specificity is 82.6%;The calibration curve drawn is close to the standard curve,which shows that the prediction model has a good calibration degree;Hosmer-Lemeshow test results show that X~2=8.259,P=0.409>0.05,indicating that the goodness of fit of the model is good;Bootstrap self-sampling method was used for internal verification of the model,and the C-index was 0.888,indicating that the discrimination and accuracy of the prediction model of LVD after HVR established in this study were good.Conclusions:1.Preoperative history of pulmonary hypertension,history of coronary heart disease,albumin<41.65g/l,LVEDD≥70.5mm,LVEF<54.5%,and CPB time≥125.5min are independent risk factors for LVD after HVR,which should be paid close attention to in clinical work.2.The prediction model based on the above risk factors has good prediction efficiency,which helps to improve the ability of early identification of high-risk patients with LVD after HVR,so as to make timely intervention and improve the prognosis of patients,and has important clinical value.