Inter-reader variability in chest radiography and HRCT for the early detection of asbestos-related lung and pleural abnormalities in a cohort of 636 asbestos-exposed subjects

Our plain-English summary of Inter-reader variability in chest radiography and HRCT for the early detection of asbestos-related lung and pleural abnormalities in a cohort of 636 asbestos-exposed subjects, by E. Ochsmann, T. Carl, P. Brand, et al (International Archives of Occupational and Environmental Health, 2010).

The study

• Authors: E. Ochsmann, T. Carl, P. Brand, H.J. Raithel, T. Kraus
• Journal: International Archives of Occupational and Environmental Health
• Published: 2010;83(1):39-46
• Cited by: 14 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Purpose: To compare inter-reader variability of chest X-ray and high resolution computed tomography (HRCT) scans of formerly asbestos-exposed employees over a 4-year period.

Methods: In this longitudinal study, 636 formerly asbestos-exposed persons were annually examined with chest radiographs and HRCT scans. Ten observer pairs classified the radiographs and HRCT scans, using the ILO classification and a custom-made CT classification. Inter-observer variability was calculated using the kappa-coefficient.

Results: Despite all expectations, HRCT inter-reader variability according to asbestos-related lung or pleura alterations at an early stage did not turn out to be better than X-ray inter-reader variability. Substantial inter-observer agreement was found for pleural calcifications (kappa(X-ray) = 0.63; kappa(CT) = 0.64). Averaging over kappa led to fair inter-observer agreement of both methods (kappa(X-ray) = 0.36; kappa(CT) = 0.34).

Conclusions: High resolution computed tomography scans are superior to X-rays in detecting lung alterations after asbestos exposure and are supposedly easier to interpret. Nevertheless, inter-observer variability did not differ between the two methods in this study. This was probably due to the only discrete asbestos-related lung or pleura alterations of this cohort and to the unfamiliar CT classification sheet, which was revised on the basis of the presented results.

In plain English

This German study compared how consistently pairs of doctors read chest X-rays and high resolution CT (HRCT) scans of 636 formerly asbestos-exposed people from 1993 to 1997. None had a manifest asbestos-related lung or pleural disease already reported to the German Employers' Liability Insurance Association. Four yearly examinations were planned, and with dropouts the study comprised 1,975 examinations. Seven experienced readers in ten pairs classified the X-rays with the ILO system and the HRCT scans with a new ILO-based CT sheet. Agreement between readers was measured with the kappa coefficient. HRCT showed clear signs of asbestos-related lung changes in 35.4% of participants, against 3.4% on X-ray, and pleural changes in 55.3%, against 37.9%. Averaged agreement was fair for both: kappa 0.36 for X-rays and 0.34 for CT. Agreement on pleural calcification was substantial (0.64 CT, 0.63 X-ray), but CT agreement on diffuse pleural thickening was only slight. The authors concluded that reader agreement did not clearly differ between the methods, although HRCT detected abnormalities more often. They attributed this probably to readers' inexperience with the new CT sheet and the very discrete changes in this cohort.

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: Longitudinal study of inter-observer (inter-reader) agreement comparing chest radiographs with HRCT scans, measured with the kappa coefficient
• Who was studied: 636 formerly asbestos-exposed subjects (584 male; median age 63 years, range 53 to 70) from the German registry for asbestos-exposed persons, examined at one of six clinical centres in Germany between 1993 and 1997. Inclusion required being younger than 70 and in the highest risk group for lung cancer or mesothelioma according to a mathematical model. Subjects could have no manifest asbestos-related disease of lung or pleura already reported to the German Employers' Liability Insurance Association at the time of inclusion. Drop-out rates ranged from 7.4 to 14.4% across centres.
• Exposure: Former occupational asbestos exposure; fiber type not stated. Median exposure was 19 years (inter-quartile range 23 years), and median latency from first exposure to first examination was 43 years (inter-quartile range 7 years). Imaging was planned as chest radiographs and HRCT about once a year, four times. Seven experienced observers, arranged in ten pairs, read the images independently.
• Main results:
  • In total, the study comprised 1,975 examinations read by ten observer pairs.
  • Chest radiographs showed clear signs of asbestos-induced lung changes in 3.4% of participants, with borderline findings in 10.0%.
  • HRCT showed clear signs of asbestosis in 35.4% of participants, with borderline results in 6.1%.
  • Pleural alterations (plaques, diffuse pleural thickenings or pleural calcifications) were seen in 37.9% of participants on chest X-ray and in 55.3% on HRCT.
  • Averaged over all findings, observer pairs and years, kappa was 0.36 for chest radiographs and 0.34 for CT scans, both fair agreement. The authors found no substantial difference between the techniques.
  • For pleural alterations, averaged kappa was 0.42 for chest radiographs (moderate agreement) and 0.37 for CT scans (fair agreement). For parenchymal alterations, it was 0.29 for radiographs and 0.30 for CT.
  • Agreement on the existence of pleural calcification was substantial: kappa 0.64 for CT and 0.63 for X-ray.
  • For thickness and extension of diffuse pleural thickening, kappa was 0.25 and 0.32 on radiographs but only 0.15 and 0.14 on CT scans (slight agreement).
  • Over the 4 years, a continuous improvement in agreement was seen only for detecting pleural calcifications on CT. Agreement on CT decreased for small irregular and linear opacities.
• The authors' conclusion: The authors found HRCT superior to chest radiographs in recognizing early lung and pleural alterations. They concluded that inter-reader agreement did not differ markedly between the two methods. They attributed this probably to lack of experience with the new CT classification and to a highly selective cohort with only very discrete asbestos-related changes. The flaws found, especially for diffuse pleural thickening, led to a revision of the CT classification sheet.
• Limitations the authors state:
  • Grouping the classification items into five main categories is likely to have improved the kappa results.
  • Readers lacked experience with the new CT classification, and repeated re-definition of the guidelines during the study caused uncertainty.
  • The annual readers' meetings may have made some observers less certain rather than improving agreement.
  • The definition of diffuse pleural thickening on CT was not clear enough.

Topics

Medical terms in this study, explained in our Lexicon: High-Resolution CT (HRCT), Chest X-Ray, ILO Classification, Pleura, Pleural Calcification, Asbestos Exposure.

Related studies

• Inter-reader agreement in HRCT detection of pleural plaques and asbestosis in participants with previous occupational exposure to asbestos — F. Laurent, C. Paris, et al. (2014)
• Clinical and HRCT screening of heavily asbestos-exposed workers — T. Vierikko, R. Järvenpää, et al. (2010)
• Chest radiographs in subjects with asbestos-related abnormalities: comparison between ILO categorizations and clinical reading — B. Hilt, A. Borgerson, et al. (1992)
• Comparison of chest digital tomosynthesis and chest radiography for detection of asbestos-related pleuropulmonary disease — G. Lee, Y.J. Jeong, et al. (2013)
• Pulmonary function, smoking habits, and high resolution computed tomography (HRCT) early abnormalities of lung and pleural fibrosis in shipyard workers exposed to asbestos — S. Neri, P. Boraschi, et al. (1996)
• Comparison of "B" readers' interpretations of chest radiographs for asbestos related changes — J.N. Gitlin, L.L. Cook, et al. (2004)

Browse more research in International Archives of Occupational and Environmental Health, or return to medical research.