应用ROC曲线评价B型尿钠肽对心力衰竭的诊断价值

[目的]应用ROC曲线评价B型尿钠肽(BNP)对心力衰竭的诊断价值。[方法]将103例患者和44例正常人BNP结果用Exec l软件作累积频数分布表,然后以灵敏度为纵坐标,以1-特异性为横坐标,建立ROC曲线评价系统,确定最佳诊断分界值,计算ROC曲线下面积(AUC)。[结果]通过心力衰竭患者ROC曲线分析证实,BNP取cut-off值为100 pg/m l时,诊断心力衰竭的灵敏度为92.2%,特异度为75%,BNP取cutoff值为120 pg/m l时,诊断心力衰竭的灵敏度为86.4%,特异度为81.82%,ROC曲线下面积AUC的95%可信区间为0.933±0.027。[结论]BNP是一个评价心脏功能的良好指标,可以帮助临床筛选无症状心功能不全,使患者得到早期治疗。

Application of ROC curve to evaluate the diagnosic value of brain natriuretic peptide in heart failure

[Objective] To evaluate the effectiveness of ROC curve in evaluation the diagnosic value of brain natriuretic peptide(BNP) in heart failure. [Methods] BNP results from 103 patients and 44 healthy people were collected.Cumulative frequency distribution form with Excel was made.Best coordinate was established and the area under curve(AUC) calculated. [Results] According to the analysis of the ROC curve of heart failure patients,while the cutoff value of BNP was 100pg/ml,the sensitivity was 92.2%,the specificity was 75%;If the cutoff value of BNP was 120pg/ml,the sensitivity would be 86.4%,and the specificity would be 81.82%.The 95% confidence interval of the area under the ROC curve was 0.933±0.027. [Conclusion] BNP is an excellent marker that can help diagnose heart failure without symptoms in clinical practice.