摘要
Abstract
Objective To investigate the risk factors for nonalvular atrial fibrillation and to establish a prediction score system for it.Methods Totally 236 hospitalized patients with nonvalvular atrial fibrillation and 291 patients with the sinus rhythm rate who were selected as control group,who were admitted to cardiovascular internal medicine,were enrolled during January 2014 to January 2016.Collected the dates include Demographic,color sonography,laboratory examinations for all patients.Nonvalvular atrial fibrillation risk factors and it OR values were determined after statistically analyzed date by One-Way ANOVA,multivariate logistics regression analysis.Prediction score system for nonvalvular atrial fibrillation was further established by area under the ROC curve and Hosmer-Lemeshow goodness of fit test.Results One-Way ANOVA showed that age,stroke,heart failure,Diabetes,Hypertension,Left atrial diameter,Left ventricular ejection fraction,left ventricular fractional shortening,Left ventricular end-diastolic diameter,left ventricular end-systolic dimension,homocysteine,Triglycerides,Totl cholesterol,high-density lipoprotein,Low density lipoprotein and ceruloplasmin were risk fators for nonvalvular atrial fibrillation.Logistic analysis showed that left atrial diameter and homocysteine were independent factors associated with nonvalvular atrial fibrillation,and Triglycerides,Totl cholesterol,high-density lipoprotein and Low density lipoprotein were the protective factors to a certain extent,after correct the confounding factors such as age,gender and common complication.Prediction score system was established which the highest score was 3.5.A risk score of 1.25 points was determined by Youden′s index,as the optimal cut-off for predict nonvalvular atrial fibrillation.Patients with <1.25 points were considered by low risk,and ≥1.25 points were considered at high risk for nonvalvular atrial fibrillation.The prediction score system of nonvalvular atrial fibrillation showed adequate discrimination(area under ROC curve was 0.841) and calibration(Hosmer-Lemeshow statistic test,P=0.235).Conclusion Logistic analysis showed that left atrial diameter and homocysteine were independent factors associated with nonvalvular atrial fibrillation,and Triglycerides,Totl cholesterol,high-density lipoprotein and Low density lipoprotein were the protective factors,after correct the confounding factors such as age,gender and common complication.The clinical prediction score system may help clinicians to make pre-intervention for patients with high nonvalvular atrial fibrillation risk.关键词
房颤/危险因素/Logistic回归Key words
nonvalvular atrial fibrillation/risk factors/logistic regression/prediction score system
