Problems With The Measurement Of Radiological Change

Our plain-English summary of Problems With The Measurement Of Radiological Change, by J.B. Werner (1980).

The study

• Author: J.B. Werner
• Published: Biological Effects Of Mineral Fibres (J.C. Wagner, Editor), Lyon: Iarc Scientific Publications, Volume 30, Pp. 571-577 (1980)
• Type of article: Conference paper
• Our copy: from Johns-Manville's own medical library, J-M no. 000877. The article is still under copyright, so we describe it here rather than post it.

In plain English

This paper described problems with reading chest X-rays in the Health and Safety Executive's asbestos survey. The survey began in 1971 to study illness and death among asbestos workers in factories under the 1969 Asbestos Regulations. By the end of 1978 it held 36,592 X-rays from 26,471 workers. An earlier pilot study of 101 workers had given useful results, with 587 X-rays from 1951 to 1969 read side by side by three readers. In the main survey, three readers classified each film separately, without seeing a worker's earlier films. In 6369 pairs of first and second X-rays, 1339 (21.0%) second X-rays went to lower categories and 510 (8.0%) to higher ones. The share of X-rays read as abnormal fell from 28.3% in 1972 to 3.0% in 1978. The author took this to show a large change in reading standards over time. Readers also disagreed widely: within one panel, one reader judged 0.9% of 223 films abnormal, another 23.8%. The author concluded that future studies must overcome both problems and planned to re-read all or a sample of the films.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Methodological report: an analysis of how chest X-rays were read and classified in a large occupational health survey, with an earlier pilot study
• Who was studied: Asbestos workers in factories subject to the 1969 Asbestos Regulations in the UK, surveyed from 1971. By the end of 1978 there were 26,471 workers with 36,592 examinations and X-rays, from about 800 establishments. The analyses used 32,695 X-rays read between the start of the survey and early 1978. The pilot study covered 101 workers with 587 X-rays from 1951 to 1969.
• Exposure: Work with asbestos in factories subject to the 1969 Asbestos Regulations. The article does not name fiber types or give doses. Chest X-rays were classified using the International Labour Office (ILO) scale.
• Main results:
  • Among the 6369 pairs of first and second X-rays in Table 1, 1339 (21.0%) of the second X-rays went to lower categories, against 510 (8.0%) that went to higher ones.
  • Of 61 men whose first X-rays were classified in the high categories 2 and 3, 18 (29.5%) had second X-rays classified in category 0.
  • The percentage of X-rays read as abnormal (category 1/0 or above) fell from 28.3% in 1972 to 3.0% in 1978. This pattern was not much affected by film quality.
  • The author interpreted Tables 1 and 2 as showing a very great change over time in the standards applied when reading X-rays.
  • Readers' interpretations of the same films differed considerably, sometimes placing the same X-ray at opposite ends of the scale. Only for films whose lowest reading was 0/0 did a majority show complete agreement among all three readers.
  • Within one panel, one reader judged only 0.9% of the 223 films read by that panel to be abnormal, while reader 19 found 23.8% abnormal. Readers from the Pneumoconiosis Medical Panels in general found a lower percentage of abnormal X-rays than readers in other groups.
• The authors' conclusion: The author concluded that two problems undermined the main survey: variability between readers, and changing reading standards over the seven-year reading period. Both must be overcome in any future study designed to measure radiological change. Films should be gathered and then read together over as short a period as possible. For the HSE survey, the plan was to select all or a sample of the films for re-reading in a single intensive exercise by a group of readers.
• Limitations the authors state:
  • Because the X-ray reading method changed between the pilot study and the full survey, the pilot study could not be claimed as a genuine test of the method.
  • Readers had a set of standard ILO films to compare against, but it is not known how often they used them.

Topics

Medical terms in this study, explained in our Lexicon: Chest X-Ray.

Browse return to medical research.