RT Journal Article SR Electronic T1 A Comparison of Some Alternative Procedures in the Classification of Chest Radiographs for Coal-workers' Pneumoconiosis JF British Journal of Industrial Medicine JO Br J Ind Med FD BMJ Publishing Group Ltd SP 293 OP 299 DO 10.1136/oem.20.4.293 VO 20 IS 4 A1 Rae, S. A1 Ashford, J. R. A1 Morgan, D. C. A1 Pasqual, R. S. H. A1 Pearson, N. G. YR 1963 UL http://oem.bmj.com/content/20/4/293.abstract AB This paper is concerned with the problem of reading individual chest radiographs for coal-workers' pneumoconiosis in terms of the I.L.O. (1953) classifications, using a group of four readers. It represents the first of a series of investigations by the National Coal Board's Pneumoconiosis Field Research into various aspects of the general subject of the classification of chest radiographs. Later papers in this series will cover related problems, including the effect of radiographic “technique” on reading standards, the classification of a series of radiographs of the same individual, and the use of the readings in the correlation of radiographic abnormality with environmental exposure. The effect of the circumstances under which the readings must be made in the Pneumoconiosis Field Research, including the geographical separation of the four readers into two groups of two, is described, and it is shown that any acceptable procedure must involve two distinct stages. On the “first stage” every radiograph must be read by some or all of the doctors individually, and on the “second stage” a proportion of films (those on which the individual readings are “ambiguous”) must be classified by the four doctors reading together. The general conduct of the joint reading sessions is then considered, and a description is given of a trial which was held to compare some alternative procedures. On the basis of these results an optimum reading procedure is derived. This is as follows:- (i) Every radiograph to be read independently by one or other of the doctors from each group, in such a way that each reads half of the total number of films. The doctors from the group producing the radiograph to read alternately consecutive batches of about 50 as they are taken during the survey; the doctors from the other group to read the first and second halves of the survey, respectively, the films being arranged in serial order. (ii) If these two independent readings are consistent, the common reading to be taken as definitive. All other films to be classified by joint consultation between the four doctors reading together, their assessment being arrived at by general discussion with each of the readers in turn giving the first opinion of the category of the film.