Response to Marsh, G. M., lerardi, A. M., Benson, S. M., & Fin ley, B. L (2019). Occupational exposures to cosmetic talc and risk of mesothelioma: an updated pooled cohort and statistical power analysis with consideration of latency period. Inhalation toxicology, 31(6), 213-223

摘要

Sir Marsh et al. (2019) recently published an article in your journal providing an updated pooled analysis of four talc miner and miller cohorts. The power calculations reported by Marsh et al. (2019) overstate the actual power. The authors use reported "expected cancers" from the four component studies as the basis for their calculations. The Marsh et al. (2019) background incidence rate of mesothelioma is 2.64 cases per 100,000 person years. However, this rate derives from national and regional registries that include both asbestos-exposed and asbestos-unexposed individuals. To detect whether working in the industry increased the risk of mesothelioma, the expected cancer numbers should reflect unexposed incidence rates. Otherwise, the study is merely comparing mesothelioma rates in two exposed cohorts, which can only determine if the risk in the talc workers is higher than the risk in the other exposed population. The correct background rate for the power calculations is the incidence rate in an unexposed population. Several authors have estimated the unexposed rate and suggest a rate of 1-2 cases per 1,000,000 person years (McDonald 1985; Teta et al. 2008). Teta et al. (2008) utilized SEER data through 2002. We extended their analysis through 2016 producing an estimated unexposed rate of 2.1 per 1,000,000 (code available from the first author).