Hessel P A
Am J Ind Med. 1985;8(1):43-8. doi: 10.1002/ajim.4700080106.
The tendency for readers of chest films of the pneumoconioses to place films in the middle of the International Labour Office (ILO) categories of profusion (eg, 1/1, 2/2) was described by Morgan et al [1974], who saw this tendency as a possible source of bias. The present study reexamines the middling tendency in relation to bias and variability by comparing duplicate readings of chest films for pneumoconiosis. Readings of trained lay persons were compared to professional readings of the same films and readings of professional readers were compared with each other. Reasons for the middling tendency, which include the use of standard middle films and the instructions given to users of the classification, are suggested and possible solutions are proposed.
摩根等人[1974年]描述了尘肺病胸部X光片的阅片者倾向于将片子归类于国际劳工组织(ILO)密集度分类的中间类别(如1/1、2/2),他们认为这种倾向可能是偏差的一个来源。本研究通过比较尘肺病胸部X光片的重复阅片结果,重新审视了与偏差和变异性相关的中间倾向。将经过培训的非专业人员的阅片结果与同一批片子的专业阅片结果进行了比较,并且专业阅片者之间的阅片结果也相互进行了比较。文中提出了产生中间倾向的原因,包括使用标准的中间片子以及对分类使用者的说明,并提出了可能的解决办法。