Variability in interpretation among B-readers in the U.S. Navy Asbestos Medical Surveillance Program

Our plain-English summary of Variability in interpretation among B-readers in the U.S. Navy Asbestos Medical Surveillance Program, by J.G. Muller, J.M. Lieske, J.E. Hernandez, et al (Military Medicine, 2008).

The study

• Authors: J.G. Muller, J.M. Lieske, J.E. Hernandez, G. Kubiak, W.G. Rudolph, M.H. Jan
• Journal: Military Medicine
• Published: 2008;173(4):375-80
• Cited by: 4 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

U.S. Navy Asbestos Medical Surveillance Program B-reading reports were analyzed for variability in identifying pneumoconiosis in 278,944 chest radiographs done from 1990 to 2004 using commercial database and statistical software. Reported presence of pneumoconiosis by 33 B-readers ranged from 1.85 to 32.28%. The reported presence of parenchymal abnormalities ranged from 0.11 to 15.07%; increased profusion (to 1/0 or greater) of small parenchymal opacities ranged from 0.10 to 17.70%; presence of large opacities ranged from 0.00 to 0.67%; and presence of pleural abnormalities ranged from 1.08 to 23.93%. Ranges appear to be decreasing slightly as rates of abnormal findings are diminishing. Differences between U.S. East Coast, West Coast, and midcontinent were not statistically significant.

In plain English

This study looked at how consistently doctors certified as B-readers classified chest X-rays in the U.S. Navy Asbestos Medical Surveillance Program, which enrolls current workers with current or past asbestos exposure. The authors analyzed 278,944 B-readings made by 33 B-readers from 1990 to 2004. Each film was read by only one B-reader. The share of films each reader called abnormal overall ranged from 1.85% to 32.28% (mean 13.01%). Reported abnormalities of the lung tissue itself (parenchymal abnormalities) ranged from 0.11% to 15.07% of a reader's films. Reported pleural abnormalities ranged from 1.08% to 23.93%. Readers who reported more of one type of abnormality tended to report more of the others, except for large opacities. Average rates were similar for readers on the East Coast, the West Coast and midcontinent. The range of reported parenchymal abnormalities narrowed in 2000 to 2004. The authors concluded that variability between B-readers remains very high. They stated that no reading was verified as valid, so the study measured how much readers differed, not which readings were accurate.

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: Retrospective analysis of existing B-reading records from a medical surveillance program, comparing rates of reported abnormalities between B-readers
• Who was studied: 278,944 chest X-ray readings made by 33 certified B-readers (physicians) in the U.S. Navy Asbestos Medical Surveillance Program, 1990 to 2004. Films were sent to the next B-reader on a rotating list.
• Exposure: The X-rays were of current workers with current or past asbestos exposure. Fiber types and doses were not given. The factor compared was the individual B-reader, also grouped by the reader's region and city size.
• Main results:
  • Across the 33 B-readers, the percentage of films reported as overall positive had a mean of 13.01% and ranged from 1.85% to 32.28%.
  • The percentage of films reported with parenchymal abnormalities (small and large opacities) ranged from 0.11% to 15.07% between readers, a 134-fold range.
  • The percentage of films reported with increased profusion of small parenchymal opacities (1/0 or greater) ranged from 0.10% to 17.70% between readers.
  • The percentage of films reported with pleural abnormalities ranged from 1.08% to 23.93% between readers.
  • The percentage of films reported with large opacities ranged from 0.00% to 0.67% between readers.
  • For each reader, rates of the different abnormal findings were correlated with each other (p values of 0.002 or less), except for large opacities.
  • There was no statistically significant association between the percentage of films a reader called unreadable and the reader's rates of abnormal findings. The reader with the highest percentages of overall positives, parenchymal abnormalities and increased profusion was one of three readers who reported no films as unreadable.
  • Mean overall positive rates were 13.47% for East Coast readers, 12.00% for West Coast readers and 13.49% for midcontinent readers.
  • The range between readers of reported parenchymal abnormalities decreased in 2000 to 2004.
  • Correlation between the reader's city size and the reader's rate of increased profusion was just above the 0.05 significance level.
• The authors' conclusion: The authors concluded that variability between B-readers remains very high.
• Limitations the authors state:
  • No reading was verified as valid.
  • Each film was read by only one B-reader.
  • The authors did not check whether some readers received films of workers with especially high asbestos exposure.
  • Random distribution of films to readers was assumed but not ensured.
  • Converting text entries into numbers may have introduced a small amount of error.

Topics

The National Library of Medicine indexes this study under subjects our site covers: The Navy.

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

Related studies

• Changes in B-readings over time in the United States Navy Asbestos Medical Surveillance Program — J.G. Muller, W.G. Rudolph, et al. (2007)
• Navy Asbestos Medical Surveillance Program 1990-1999: demographic features and trends in abnormal radiographic findings — D.M. Dell, B.K. Bohnker, et al. (2006)
• Trends in pleural radiographic findings in the Navy Asbestos Medical Surveillance Program (1990-1999) — J.G. Muller, B.K. Bohnker, et al. (2005)
• Navy Asbestos Medical Surveillance Program (1984-1990): linear regression analysis for effect of asbestos exposure on pulmonary function testing — B.K. Bohnker, L.S. Betts, et al. (2004)
• Navy Asbestos Medical Surveillance Program (1991-1999): linear regression analysis for the effect of asbestos exposure on pulmonary function testing — B.K. Bohnker, L.S. Betts, et al. (2004)
• Asbestos-related cancers among 28,300 military servicemen in the Royal Norwegian Navy — L.A. Strand, J.I. Martinsen, et al. (2010)

Browse more research in Military Medicine, or return to medical research.