节点文献
血红蛋白A2在珠蛋白生成障碍性贫血筛查中的截断值及应用
Analysis of the cut-off value and applied value of hemoglobin A2 in screening for thalassemia
【摘要】 目的分析该实验室血红蛋白A2(HbA2)筛查珠蛋白生成障碍性贫血(又称地贫)的最佳截断值及其应用价值。方法回顾性分析2016年1-12月来该院进行地贫筛查的育龄夫妇的血液标本,通过毛细管电泳法检测HbA2水平,以基因检测结果为诊断金标准,绘制HbA2用于α和β地贫诊断的受试者工作特征(ROC)曲线,计算曲线下面积、最佳截断值,以及相应的灵敏度、特异度。结果 ROC曲线分析显示,HbA2诊断α地贫的曲线下面积为0.766,最佳截断值为≤2.45%,其对应的灵敏度为75.4%,特异度为72.0%。HbA2诊断β地贫的曲线下面积为0.984,最佳截断值为≥3.55%,其对应的灵敏度为96.20%,特异度为99.20%。结论利用ROC曲线建立了该院HbA2筛查α和β地贫的最佳截断值,曲线下面积表明HbA2诊断β地贫的准确度较高,而诊断α地贫的准确性处于中等水平。
【Abstract】 Objective To analyse the best cut-off value and applied value of hemoglobin A2(HbA2)in screening for thalassemia in this laboratory.Methods The blood of childbearing age couple who participated in the thalassemia screening from January to December 2016 in Boai Hospital of Zhongshan were retrospectively analyzed.The HbA2 level was tested by Capillary Hemoglobin Electrophoresis.Then we used the genetic results as the gold standard for diagnosis and drew the receiver operating characteristic(ROC)curve of HbA2 to diagnoseαandβ-thalassemia respectively.Finally,we computed the areas under the curve,the best cut-off value and corresponding negative predictive value,positive predictive value.Results The ROC curve showed that the area under the curve of HbA2 in diagnosingα-thalassemia was 0.766.The best cut-off value was≤2.45%,the sensitivity,specificity were 75.4% and 72.0% respectively.The ROC curve showed that the area under the curve of HbA2 in diagnosingβ-thalassemia was 0.984.The best cut-off value was ≥3.55%,the sensitivity and specificity were 96.20% and 99.20% respectively.Conclusion The best cut-off value of HbA2 in diagnosingα-thalassemia andβ-thalassemia was established by using ROC curve.The area under the curve showed that the accuracy of HbA2 in diagnosingβ-thalassemia was good,while the accuracy of HbA2 in diagnosingα-thalassemia was at a moderate level.
【Key words】 thalassemia; hemoglobin A2; cut-off value; receiver operating characteristic curve;
- 【文献出处】 检验医学与临床 ,Laboratory Medicine and Clinic , 编辑部邮箱 ,2019年10期
- 【分类号】R556;R446.1
- 【被引频次】4
- 【下载频次】128