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

Our plain-English summary of 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, by A.I. Naimi, S.R. Cole, M.G. Hudgens, et al (Epidemiology, 2014).

The study

• Authors: A.I. Naimi, S.R. Cole, M.G. Hudgens, D.B. Richardson
• Journal: Epidemiology
• Published: 2014;25(2):246-54
• Cited by: 31 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

Background: Previous estimates of the effect of occupational asbestos on lung cancer mortality have been obtained by using methods that are subject to the healthy-worker survivor bias. G-estimation of a structural nested model provides consistent exposure effect estimates under this bias.

Methods: We estimated the effect of cumulative asbestos exposure on lung cancer mortality in a cohort comprising 2564 textile factory workers who were followed from January 1940 to December 2001.

Results: At entry, median age was 23 years, with 42% of the cohort being women and 20% nonwhite. During the follow-up period, 15% of person-years were classified as occurring while employed and 13% as occupationally exposed to asbestos. For a 100 fiber-year/ml increase in cumulative asbestos, a Weibull model adjusting for sex, race, birth year, baseline exposure, and age at study entry yielded a survival time ratio of 0.88 (95% confidence interval = 0.83 to 0.93). Further adjustment for work status yielded no practical change. The corresponding survival time ratio obtained using g-estimation of a structural nested model was 0.57 (0.33 to 0.96).

Conclusions: Accounting for the healthy-worker survivor bias resulted in a 35% stronger effect estimate. However, this estimate was considerably less precise. When healthy-worker survivor bias is suspected, methods that account for it should be used.

In plain English

This study reanalyzed records from the South Carolina Chrysotile Asbestos cohort: 2564 asbestos textile factory workers followed for lung cancer deaths over 100,765 person-years between 1940 and 2001. Exposure was the cumulative number of chrysotile fibers longer than 5 µm per ml of air, measured in units of 100 fiber-years/ml. The authors compared standard statistical models with a structural nested model, a method meant to account for the healthy-worker survivor bias. During follow-up, 142 workers (5.5%) died of lung cancer. With the structural nested model, each 100 fiber-years/ml increase in cumulative exposure was associated with a 43% shorter time to lung cancer death (survival time ratio 0.57, 95% confidence interval 0.33 to 0.96). The standard models gave 12% and 11% shorter times (0.88, 0.83 to 0.93; and 0.89, 0.82 to 0.96 after adjusting for work status). The structural nested estimate was 35% stronger but much less precise. The authors concluded that the difference suggests bias in previous analyses of these data. They noted they had no information on smoking, socioeconomic position or intermittent time off work.

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: Occupational cohort study (analysis of existing records) comparing standard Weibull accelerated failure-time models with g-estimation of a structural nested accelerated failure-time model
• Who was studied: 2564 asbestos textile factory workers in South Carolina who started work between 1940 and 1965, followed for lung cancer mortality over 100,765 person-years from 1940 to 2001; at entry, median age 23, 42% women, 20% nonwhite
• Exposure: Cumulative occupational exposure to chrysotile asbestos fibers longer than 5 µm/ml of air in an asbestos textile factory, assigned with a job-exposure matrix, expressed per 100 fiber-years/ml
• Main results:
  • 142 of the 2564 workers (5.5%) were classified as having died of lung cancer
  • Standard model not adjusted for work status: survival time ratio 0.88 (95% CI 0.83 to 0.93) per 100 fiber-years/ml increase in 10-year lagged cumulative asbestos
  • Standard model further adjusted for work status: survival time ratio 0.89 (0.82 to 0.96)
  • Structural nested model: survival time ratio 0.57 (bootstrap 95% CI 0.33 to 0.96) per 100 fiber-years/ml increase
  • The authors expressed these as a 43% shorter time to lung cancer death with the structural nested model, versus 12% and 11% shorter with the standard models
  • The structural nested model estimate was 35% stronger than both standard Weibull model estimates
• The authors' conclusion: The authors concluded that the stronger estimate from the structural nested model suggests bias in previous analyses of these data, and that methods accounting for healthy-worker survivor bias should be used when that bias is suspected
• Limitations the authors state:
  • No information on individual smoking status, socioeconomic position, or intermittent time off work
  • Confidence intervals from the structural nested model were much wider (less precise) than those from the standard models
  • The counterfactual consistency assumption needed for causal interpretation cannot be verified

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: Asbestos Exposure, Lung Cancer, confidence interval.

Related studies

• 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)
• Hierarchical latency models for dose-time-response associations — D.B. Richardson, R.F. MacLehose, et al. (2011)
• Lung cancer mortality in North Carolina and South Carolina chrysotile asbestos textile workers — L. Elliott, D. Loomis, et al. (2012)
• Analysis of occupational asbestos exposure and lung cancer mortality using the g formula — S.R. Cole, D.B. Richardson, et al. (2013)
• Accounting for outcome misclassification in estimates of the effect of occupational asbestos exposure on lung cancer death — J.K. Edwards, S.R. Cole, et al. (2014)
• Multistage modelling of lung cancer mortality in asbestos textile workers — N. Pearce (1988)

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