Abstract

Lindars, D. C. (1971).Brit. J. industr. Med.,28, 131-142. An elaboration of small opacity types (p, m, and n) in simple pneumoconiosis. According to the I.L.O. classification (International Labour Office, 1959), radiographs showing pneumoconiosis may be classified as p, m or n according to the greatest diameter of predominant (small) opacities. Recent work has revealed pathological and physiological differences associated with these appearances.

In response to a request for some refinement of the classification for correlation with pathological data a `pmn elaboration' has been devised, analogous to the N.C.B. elaboration of the I.L.O. classification (Liddell and Lindars, 1969). It has the following form: o/-, o/o, o/p; p/o, p/p, p/m; m/p, m/m, m/n; n/m, n/n, n/A. Instructions to readers are similar to those for use with the N.C.B. elaboration. The significance of o/- and o/o is identical in the two elaborations; n/A indicates predominance of opacity size close to that of large opacities in the I.L.O. classification.

Two hundred and forty-seven radiographs have been read, on two occasions by four film readers, using both elaborations. Analysis of the results showed that the readers had a slightly greater observer error, in terms of variance, when using the pmn elaboration than with the N.C.B. elaboration, but reading bias was less. Calculation of information transmitted showed a gain in information in the pmn elaboration over conventional p, m, n typing, comparable to the gain in information achieved by the N.C.B. elaboration.

Marginal zones were not used as frequently as in the N.C.B. elaboration. The radiograph series contained too few normal or near-normal radiographs for the lower end of the scale to be adequately studied. Improved results can be expected with increased experience and more careful framing of reading instructions for prior briefing.

It is recommended that the pmn elaboration should be used whenever typing is required for correlation between radiographic appearance and pathological or physiological data.