Methods old and new for analysing occupational cohort data

Our plain-English summary of Methods old and new for analysing occupational cohort data, by A.D. McDonald, J.C. McDonald (American Journal of Industrial Medicine, 1988).

The study

• Authors: A.D. McDonald, J.C. McDonald
• Journal: American Journal of Industrial Medicine
• Published: 1988;13(6):735-6
• Type of article: Letter
• Cited by: 1 article, as counted by NIH iCite
• Read it: PubMed · Publisher
• In J-M's library: Johns-Manville's company medical library held a copy (J-M nos. 000251, 000290).

In plain English

This is a letter to the editor, not a study, followed by a reply. Alison and Corbett McDonald wrote that a 1987 review of cohort-analysis methods by Alice Whittemore misreported their asbestos data. They said its Table II came not from the paper cited, on a chrysotile textile plant, but from their other 1983 paper, on a factory using chrysotile, amosite and crocidolite in mainly textile manufacture. That paper gave lung cancer standardized mortality ratios (SMRs) at five dust-exposure levels. As the scanned table reads, these rose from 66.9 at under 10 mpcf.y to 416.1 at 80 or more. The review showed only the middle three levels, which the letter called inaccurate, with invented totals. The letter writers concluded that deliberate misrepresentation of others' data for illustration should be stated. Whittemore replied that the middle-category death counts (5, 10 and 6) were accurate and the total was their sum. She said a trend test needs a level for each category. She used the midpoints of the middle three and omitted the extremes because they had none. Her aim was to explain a method, not to show asbestos unassociated with lung cancer.

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: Letter to the editor commenting on a published methods review, followed by the review author's reply; no new data
• Who was studied: The letter discusses lung cancer mortality data from the authors' own earlier cohort study of an American factory that used asbestos fibers of mixed type in mainly textile manufacture (McDonald et al., 1983b); cohort size and dates are not given in the letter
• Exposure: Chrysotile, amosite and crocidolite used in mainly textile manufacture; cumulative dust exposure in five categories expressed in mpcf.y (less than 10, 10 to less than 20, 20 to less than 40, 40 to less than 80, 80 or more)
• Main results:
  • The letter writers stated that the data in Whittemore's Table II were not from the paper cited (1983a, a chrysotile textile plant) but from their other 1983 paper (1983b), on a plant using mixed asbestos fibers
  • In the mixed-fiber textile plant, the original table gave lung cancer SMRs at five dust-exposure levels. As the jumbled scanned table reads, observed deaths and SMRs were: under 10 mpcf.y, 21 deaths, SMR 66.9; 10 to under 20, 5 deaths, SMR 83.6; 20 to under 40, 10 deaths, SMR 156.0; 40 to under 80, 6 deaths, SMR 160.0; 80 or more, 11 deaths, SMR 416.1
  • The letter writers stated that Whittemore's Table II showed only the middle three exposure levels, inaccurately and with invented totals
  • In reply, Whittemore stated that Table II accurately reported the observed lung cancer deaths in the middle three exposure categories as 5, 10 and 6, and that the total of 21 was obtained by adding them
  • Whittemore stated that a trend test requires a level attached to each exposure category. To illustrate the test, she represented each of the middle three categories by its midpoint and omitted the two extreme categories because no midpoint was available
  • Whittemore stated that the table was used to explain a statistical method, not to show that asbestos was unassociated with lung cancer in the McDonald data, and that hypothetical data would have served equally well
• The authors' conclusion: The letter writers concluded that when other people's data are deliberately misrepresented for illustrative purposes, this should be stated, adding that comment is free but the facts are sacred

Topics

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

Related studies

• Analysing health data from rare occupational cohorts — T.F. Bateson (2014)
• Comparison of exposure assessment methods for occupational carcinogens in a multi-centre lung cancer case-control study — S. Peters, R. Vermeulen, et al. (2011)
• Proxy respondents and the validity of occupational and other exposure data. The Selected Cancers Cooperative Study Group — C.A. Boyle, E.A. Brann (1992)
• Design and conduct of occupational epidemiology studies: IV. The analysis of case-control data — N. Pearce, H. Checkoway, et al. (1989)
• Effects of spirometry standards in two occupational cohorts — E.A. Eisen, L.C. Oliver, et al. (1985)
• Risk of lung cancer associated with quantitative beryllium exposure metrics within an occupational cohort — M.K. Schubauer-Berigan, J.A. Deddens, et al. (2011)

Browse more research in American Journal of Industrial Medicine, or return to medical research.