Abstract

The 1950 I.L.O. classification of radiological abnormality in simple pneumoconiosis describes four discrete categories, but the actual amount of abnormality in an individual does not change by discrete steps. Radiological progression is the increase of a continuous variable, and methods of measuring it must deal with three problems: (1) the coarse classification into discrete categories; (2) possible variation in the rate of an individual's progression from point to point on the continuum of abnormality; (3) the possible non-uniformity of the continuum in relation to some relevant change in the lungs.

The methods proposed for measuring progression were developed from an analysis of two surveys, two and a half years apart, of 689 coal-face workers in eight pits. Some results from a third survey five and a half years after the second are also used. The merits of different methods of film reading are considered. Smooth curves were fitted to the distributions of abnormality found at each survey of each pit, and the progression of the first smooth curve into the second was estimated by the new method of “percentile pairing”. The amount of progression within category 1 was found to be about double that in category 2 in all the pits. A new scale of abnormality was derived that makes this change more nearly uniform all along the continuum, and a single index was thus obtained for comparing one pit with another. Progression was also measured on the dust retention scale of Rivers, Wise, King, and Nagelschmidt (1960) and again was greater throughout category 1 than throughout category 2, so that in terms of dust retention progression is not uniform at different points in the continuum of radiological abnormality.

Various implications of the new methods and other approaches to the analysis of progression data are critically discussed.