Asbestos and lung cancer. Attributability in the face of uncertainty

Our plain-English summary of Asbestos and lung cancer. Attributability in the face of uncertainty, by P.E. Enterline (Chest, 1980).

The study

• Author: P.E. Enterline
• Journal: Chest
• Published: 1980;78(2 Suppl):377-9
• Cited by: 19 articles, as counted by NIH iCite
• Read it: PubMed
• In J-M's library: Johns-Manville's company medical library held a copy (J-M no. 000799).

In plain English

This short paper was not a new study. The author used data from earlier studies, the author's own and Selikoff's, to show how epidemiologic data can estimate the probability that a particular person's lung cancer was caused by asbestos. In an earlier study of men who made asbestos products, lung cancer deaths ran at 64 per 100,000 per year, against 31 among general-population males. On that basis the odds that a given case was caused by asbestos were about 50-50. In a study of 1348 retired asbestos workers, the estimated probability was about 17 percent for men first exposed less than 20 years before death at dust levels under 10 mppcf (relative risk 1.2). It was 79 percent for men first exposed 30 or more years earlier at levels above 10 mppcf (relative risk 4.7). Using Selikoff's data on 17,800 insulators, asbestos multiplied lung cancer death rates by 4.9 in smokers and 4.4 in nonsmokers, giving probabilities of 79 and 77 percent. The author concluded that a single case cannot be attributed with certainty, but its probability can be estimated, and that smoking history did not appear to change it much.

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 commentary illustrating how to estimate the probability of causation in an individual case, using previously published data (the author's earlier studies and Selikoff's insulator data); no new study was reported
• Who was studied: Data drawn from earlier work: a large group of men who manufactured asbestos products; 1348 men who retired from the asbestos industry during 1941-67; and Selikoff's study of 17,800 asbestos insulators in the United States and Canada
• Exposure: Asbestos exposure in asbestos products manufacturing and in insulation work, analyzed by time since first exposure and by mean dust level (under vs over 10 mppcf); cigarette smoking also considered. Fiber types are not named.
• Main results:
  • In the author's earlier study of asbestos products workers, lung cancer death rates were 64 per 100,000 per year, versus 31 per 100,000 per year in general-population males. On that basis the odds that a given case was caused by asbestos were about 50-50.
  • Among retired asbestos workers first exposed less than 20 years before death and exposed at under 10 mppcf, the relative risk was 1.2. That translates to about 1 in 6, or 17 percent, of their lung cancer cases being due to asbestos.
  • Among retired asbestos workers first exposed 30 or more years earlier at levels above 10 mppcf, the relative risk was 4.7. That gives a probability of about 79 percent that an individual case was caused by asbestos dust.
  • In Selikoff's insulator data, asbestos exposure multiplied lung cancer death rates by 4.9 in smokers and by 4.4 in nonsmokers.
  • The estimated probability that asbestos caused lung cancer was 79 percent for asbestos workers who smoked and 77 percent for those who did not.
  • In the addendum, for an asbestos worker with lung cancer who smoked, the author calculated these probabilities: 0.888 that the cancer was due to smoking, 0.975 that either smoking or asbestos was responsible, and 0.705 that both played a role.
• The authors' conclusion: The author concluded that an individual case of disease cannot be attributed with certainty to an occupational exposure, but its probability can be estimated from epidemiologic data and adjusted for time since first exposure and exposure intensity. The author also concluded that smoking history and calendar time do not appear to change this probability much for asbestos and lung cancer.
• Limitations the authors state:
  • The author noted that interpreting causal probability for two interacting factors such as smoking and asbestos requires clarifying what is meant by causation.

Topics

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

Medical terms in this study, explained in our Lexicon: Lung Cancer.

Related studies

• Prediction in the face of uncertainty: the United States versus Reserve Mining Company — A.L. Brown (1986)
• Human cancer risk from ingested asbestos: a problem of uncertainty — W.J. Nicholson (1983)
• Should uncertainty concerning the risk of malignancy be included in diagnostic (nongynecologic) cytopathology reports? — D.N. Poller, F. Schmitt (2021)
• Manufactured uncertainty: protecting public health in the age of contested science and product defense — D. Michaels (2006)
• Sources of uncertainty in dose-response modeling of epidemiological data for cancer risk assessment — L. Stayner, A.J. Bailer, et al. (1999)
• Divided by an ocean of water but united in an ocean of uncertainty: A transatlantic review of mesothelioma surgery guidelines — D.A. Waller, I. Opitz, et al. (2021)

Browse more research in Chest, or return to medical research.