作者: Clement T. Loy , Les Irwig
关键词:
摘要: ContextAlthough it is common practice to read tests with clinical information, whether this improves or decreases the accuracy of test reading uncertain.ObjectiveTo determine whether diagnostic are more accurate when with clinical information without it.Data SourcesMEDLINE search (1966-December 2003) extended by reference lists and articles citing retrieved (Web Science, 1985-December 2003).Study SelectionAll comparing twice same readers, once information, but otherwise under identical conditions. Only that reported sensitivity specificity or receiver operating characteristic (ROC) curves were included.Data ExtractionData extracted one author reviewed independently the other. When data difficult interpret, differences resolved by discussion.Data SynthesisSixteen met inclusion criteria. Eleven compared areas under ROC for information, and 5 compared only specificity. Ten used actual clinical information; 6 constructed was plausible. Overall, improved although the effect smaller in using actual when compared those information. There no instances which resulted significant reduction in accuracy. In some instances, accuracy came from loss specificity.ConclusionsAt least examined, diagnostic tests seems justified. Future studies should be designed investigate best way providing These studies also give an estimate information used, display identified points, include a wider range tests.