Semler G, Wittchen H U, Joschke K, Zaudig M, von Geiso T, Kaiser S, von Cranach M, Pfister H
Eur Arch Psychiatry Neurol Sci. 1987;236(4):214-22. doi: 10.1007/BF00383851.
The reliability of DSM-III diagnoses using an expanded version of the Diagnostic Interview Schedule (DIS), called the Composite International Diagnostic Interview (CIDI), was evaluated by examining 60 psychiatric inpatients on a test-retest basis. Acceptable agreement coefficients of (kappa) 0.5 or above were found for all but two disorders: dysthymic disorder and generalized anxiety disorder. The subclassification of DSM-III affective disorders also revealed some discrepancies between the test and the retest interviews. When compared with results from earlier versions of the DIS, diagnostic reliability was found to have improved for the DSM-III anxiety disorders in particular. These improvements can possibly be attributed to some changes in the wording of the respective items of this section. Several reasons for lowered test-retest reliability are discussed.
使用一种名为综合国际诊断访谈（CIDI）的诊断访谈表扩展版，对《精神疾病诊断与统计手册》第三版（DSM-III）的诊断可靠性进行了评估，该评估通过对60名精神科住院患者进行重测来完成。除了两种障碍（恶劣心境障碍和广泛性焦虑障碍）之外，所有障碍的一致性系数（kappa）均达到或高于0.5，这一结果是可以接受的。DSM-III情感障碍的子分类在测试访谈和重测访谈之间也显示出一些差异。与早期版本的诊断访谈表结果相比，发现DSM-III焦虑障碍的诊断可靠性尤其有所提高。这些改进可能归因于该部分各个条目的措辞发生了一些变化。文中讨论了重测可靠性降低的几个原因。
