ASSESSING CHANGE IN THE PNEUMOCONIOSES USING SERIAL RADIOGRAPHS

Publisher Website
Google Scholar
Abstract
The assessment of progression in coalworkers' and other pneumoconioses by comparing serial radiographs is a standard procedure. Despite general acceptance of the principles involved, several sources of bias may be inherent in this practice. This study, designed to identify and quantify the various biases, shows that the assumed chronologic sequence of the paired radiographs, viz. the position, as well as the disease stage existing for a given pair member, can significantly influence the categorization process. It demonstrates that there is marked reluctance to diagnose regression of pneumoconiosis. Although one can not hope to eliminate totally the biases in a side by side reading method, partial remedies are available. Recognition that such biases exist is a start. It is desirable to conduct trials with various readers to ascertain those who are prone to “pull” side by side interpretations towards the middle (i.e. the “concertina” effect) resulting in a lower progression score and those who are oppositely inclined. Side by side readings which differ drastically from those determined by independent interpretation should be viewed with suspicion. Variability between readers will perhaps always be a problem. However, rigid initial training, as well as periodic checks to insure that reading habits do not change, would do much to improve the validity and reliability of the categorization of pneumoconiosis.

This publication has 0 references indexed in Scilit: