Non-asbestos-related mesothelioma: chance, bias, or a real phenomenon?

Our plain-English summary of Non-asbestos-related mesothelioma: chance, bias, or a real phenomenon?, by P. Glasziou (The Medical Journal of Australia, 1995).

The study

• Author: P. Glasziou
• Journal: The Medical Journal of Australia
• Published: 1995;162(8):431-2
• Type of article: Comment
• Read it: PubMed · Publisher

In plain English

This was a short commentary, not a new study, on Gun's article in the same issue. Gun's article suggested something besides asbestos causes some mesothelioma, because men's risk stayed higher than women's after allowing for asbestos. The author focused on one possible bias: incomplete adjustment for occupational asbestos exposure. Using an imaginary disease, "mesoid," the author showed how men's overall relative risk of 2.1 versus women became 1 once data were split by asbestos exposure. When 10% of people were assumed to be misclassified for exposure, the split relative risks fell below 2.1 but stayed above 1. This gave the impression that asbestos did not explain all of the excess in men. The author concluded that, without data on how reliably occupational exposure had been assessed, it could not be judged whether Gun's results reflected a small classification error or another cause. In a printed reply, Gun stated that a hypothetical 10% exposure-misclassification bias would mean the study's relative risk of 3.3 for men versus women without occupational exposure came from a population whose "true" figure was 2.0.

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: commentary (personal view) responding to another article, using a hypothetical worked example; no new data collected
• Who was studied: No study population of its own. The commentary discusses a study by Gun, published in the same issue, that compared mesothelioma risk in men and women in Australia.
• Exposure: Occupational asbestos exposure, and how errors in classifying who was exposed could distort adjustment for it
• Main results:
  • In the hypothetical 'mesoid' example, men had an overall relative risk of 2.1 compared with women. Within both the asbestos-exposed and the unexposed groups the relative risk was 1, so the apparent sex difference was entirely due to asbestos acting as a confounder.
  • When the same hypothetical data were given 10% misclassification of asbestos exposure, the relative risks for men in both the exposed and unexposed groups were less than 2.1 but still greater than 1. This created the impression that asbestos did not explain all of the association.
  • The author noted that statistical adjustment for misclassification is possible only when misclassification rates are known. This is generally not the case, so some residual confounding usually remains.
  • In the published reply, Gun described the author's point as a possible non-differential bias (equal probability of misclassifying cases and non-cases). Gun stated that such a bias could overestimate the men-to-women relative risk among people not occupationally exposed to asbestos. As an example, applying a 10% bias would mean that the relative risk of 3.3 found in Gun's study had arisen from a population in which the "true" figure was 2.0. Gun gave male incidence as 8.5 and female incidence as 2.6 per 1,000,000 person-years. Gun added that differential bias, such as recall bias, could lead to an underestimate.
  • In the reply, Gun stated that the National Mesothelioma Surveillance Program registered 159 men and 80 women with no apparent occupational exposure to asbestos.
• The authors' conclusion: The author concluded that, without data on the reliability of occupational asbestos exposure assessment, it could not be determined whether Gun's findings were due to a small error in classifying exposure or whether another cause of mesothelioma should be sought.

Topics

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

Medical terms in this study, explained in our Lexicon: Mesothelioma.

Related studies

• Evaluation of Confounding and Selection Bias in Epidemiological Studies of Populations Exposed to Low-Dose, High-Energy Photon Radiation — M.K. Schubauer-Berigan, A. Berrington de Gonzalez, et al. (2020)
• Estimating the effect of cumulative occupational asbestos exposure on time to lung cancer mortality: using structural nested failure-time models to account for healthy-worker survivor bias — A.I. Naimi, S.R. Cole, et al. (2014)
• Assessing the component associations of the healthy worker survivor bias: occupational asbestos exposure and lung cancer mortality — A.I. Naimi, S.R. Cole, et al. (2013)
• Non-response bias in a surveillance program for asbestos-related lung cancer — L. Knoll, M.K. Felten, et al. (2011)
• Children's environment in Central Europe: threats and chances — S. Böse-O'Reilly, U. Heudorf, et al. (2007)
• The fake fat phenomenon in organizing pleuritis: a source of confusion with desmoplastic malignant mesotheliomas — A. Churg, P. Cagle, et al. (2011)

Browse more research in The Medical Journal of Australia, or return to medical research.