摘要
目的 了解国内外医学影像诊断领域诊断性试验的应用状况,比较我国影像学研究中诊断性试验的应用与国际水平的差距。方法 作者对国内外医学影像界影响最大的《中华放射学杂志》和《Radiology》进行了人工查阅,并根据国际公认标准对所发表的诊断性试验进行了分析。结果 检索1998～1999年《中华放射学杂志》2卷24期,论著408篇。《Radiology》8卷24期,论著796篇。在所有论著类文章中,《中华放射学杂志》和《Radiology》的诊断试验分别占12％和23％。其中与金标准进行独立盲法对比的分别为19％和64％,同时计算了敏感性、特异性、准确性等较完整的统计学指标的分别为28％和57％。计算阳性和阴性预检值的文章比例分别为9％、9％和33％,26％。计算似然比的文章分别为0和3％。结论 我国影像诊断学领域诊断性试验的数量和质量与国际水平相比有较大差距,还远远不能满足临床的需要。改进研究方法,更多地开展高质量的诊断性试验研究对提高影像诊断的水平具有重要意义。
Objective To analyze the status of applying diagnostic test in imaging scientific study internationally and domestically, and to compare the application of the image diagnostic studies of our country with that of abroad. Method We hand-searched the diagnosis tests published in the 'Chinese Journal of Radiology',the most influential in China, and in 'Radiology' , the most influential abroad, from 1998 to 1999 respectively. Then we evaluated each of the diagnosis tests according to the international standards. Results We searched 408 original articles in 'Chinese Journal of Radiology' in which the diagnostic test articles were 12% , and 796 original articles in 'Radiology' with the diagnostic test articles 23% from 1998 to 1999 respectively. In these diagnosis tests, by comparing the 'Chinese journal of radiology' with the'Radiology', it was found that 19% applied blind comparison with Gold Standard, 28% calculated sensitivity, specificity and accuracy, 9% both calculated negative predictive value and positive predictive value and none calculated likelihood ratios in the former versus 64% , 57% , 33% and 26% and 3%respectively in the latter. Conclusions Compared with the international level, both the quality and the quantity of the diagnosis tests applied in the specialty of imaging scientific study in China are much lower and far from meeting the clinical requirement. Improving the methods of scientific study and carrying on more diagnosis tests with high qualities are of important significance in improving the diagnostic level of imaging.
关键词
影像诊断
流行病学研究
诊断性试验
循证医学
Diagnostic imaging
epidemiological methods
data interpretation, statistical evaluation
