Inverse probability weighting to estimate impacts of hypothetical occupational limits on radon exposure to reduce lung cancer

Radon is a known cause of lung cancer. Protective standards for radon exposure are derived largely from studies of working populations that are prone to healthy worker survivor bias. This bias can lead to underprotection of workers and is a key barrier to understanding health effects of many exposures. We apply inverse probability weighting to study a set of hypothetical exposure limits among 4137 male, White and American Indian radon-exposed uranium miners in the Colorado Plateau followed from 1950 to 2005. We estimate cumulative risk of lung cancer through age 90 under hypothetical occupational limits. We estimate that earlier implementation of the current US Mining Safety and Health Administration annual standard of 4 working level months (implemented here as a monthly exposure limit) could have reduced lung cancer mortality from 16 of 100 workers to 6 of 100 workers (95% confidence interval, 3/100, 8/100), in contrast with previous estimates of 10 of 100 workers. Our estimate is similar to that among contemporaneous occupational cohorts. Inverse probability weighting is a simple and computationally efficient way address healthy worker survivor bias to contrast health effects of exposure limits and estimate the number of excess health outcomes under exposure limits at work.