Limitations of minimally acceptable interpretive performance criteria for screening mammography.
Limitations of minimally acceptable interpretive performance criteria for screening mammography.
复制标题
筛查乳房 X 光检查的最低可接受解释性能标准的局限性。
DOI：
10.1148/radiol.101735
复制
发表时间：
2011
期刊：
影响因子：
19.7
通讯作者：
Miglioretti,DianaL
中科院分区：
文献类型：
--
作者：
Doyle,GregoryP;Onysko,Jay;Pogany,Lisa;Major,Diane;Caines,Judy;Shumak,Rene;Wadden,Nancy;Carney,PatriciaA;Sickles,EdwardA;Monsees,BarbaraS;Bassett,LawrenceW;Miglioretti,DianaL
给编辑的信也有同样的效果，相反，报告称，将肺动脉造影测试集中肺栓塞的患病率从 20% 增加到 60%，会增加怀疑水平、敏感性和接受者操作特征曲线下的面积。在另一项研究中，Gur 等人 (5) 证明，与实验室测试组相比，临床实践中的性能更高，观察者之间的变异性更低，即使在两种环境中读取同一组乳房 X 光照片也是如此。因此，测试情况下观察者之间变异性的测量可能无法准确反映实践中的变异性。总体而言，这些研究的结果表明，案例集的丰富程度和报告环境可能是研究中实现合理现实条件的因素。我对社论作者关于研究人员应如何据此进行研究设计和报告的观点感兴趣。
LETTERS TO THE EDITOR the same effect and, conversely, reported that increasing the prevalence of pulmonary emboli in a test set of pulmonary arteriograms from 20% to 60% increased levels of suspicion, sensitivity, and area under the receiver operating characteristic curve. In a separate study, Gur et al (5) demonstrated that performance was higher and variability between observers was lower in clinical practice than in a laboratory test set, even when the same set of mammograms was read in both environments. Therefore, measures of variability between observers in the test situation may not accurately reflect variability in practice. Overall, the findings in these studies suggest that the level of enrichment of the case set and the reporting environment may be factors in achieving reasonably realistic conditions in research studies. I would be interested in the editorial authors’ perspectives on how researchers should approach research design and reporting in light of this.