The merits and limits of pooling data from nuclear power worker studies

Nuclear power plant workers are exposed to various sources of occupational radiation and are a suitable population to investigate the effects of low and protracted exposure. Thus, since the 1970s, analyses have been done of data from a single power plant from one country, several plants from one country, and several plants from several countries. However, comparisons between results are hampered by different designs and different inclusion criteria. Risk estimates very between studies and larger studies or pooled analyses are needed to increase precision. International pooling started in 1995, with a study that included data from Canada, the UK, and the USA.1IARC Study Group on Cancer Risk among Nuclear Industry WorkersDirect estimates of cancer mortality due to low doses of ionising radiation: an international study.Lancet. 1994; 344: 1039-1043Summary PubMed Scopus (88) Google Scholar Data were later added from a further 12 countries.2Cardis E Vrijheid M Blettner M et al.The 15-country collaborative study of cancer risk among radiation workers in the nuclear industry: estimates of radiation-related cancer risk.Rad Res. 2007; 167: 396416Crossref Scopus (560) Google Scholar In The Lancet Haematology, Klervi Leuraud and colleagues present results of extended follow-up for the INWORKS study,3Leuraud K Richardson DB Cardis E et al.Ionising radiation and risk of death from leukaemia and lymphoma in radiation-monitored workers (INWORKS): an international cohort study.Lancet Haematol. 2015; (published online June 22.)http://dx.doi.org/10.1016/S2352-3026(15)00094-0Google Scholar including cohorts from France,4Metz-Flamant C Laurent O Samson E et al.Mortality associated with chronic external radiation exposure in the French combined cohort of nuclear workers.Occup Environ Med. 2013; 70: 630-638Crossref PubMed Scopus (57) Google Scholar the UK,5Muirhead CR O'Hagan JA Haylock RG et al.Mortality and cancer incidence following occupational radiation exposure: third analysis of the National Registry for Radiation Workers.Br J Cancer. 2009; 13; 100: 206-212Crossref PubMed Scopus (347) Google Scholar and the USA,6Schubauer-Berigan MK Daniels RD Bertke SJ Tseng CY Richardson DB Cancer mortality through 2005 among a pooled cohort of U.S. nuclear workers exposed to external ionizing radiation.Rad Res. 2015; (published online May 26.)https://doi.org/10.1667/RR13988.1Crossref Scopus (83) Google Scholar and by contrast with the previous pooled analysis,2Cardis E Vrijheid M Blettner M et al.The 15-country collaborative study of cancer risk among radiation workers in the nuclear industry: estimates of radiation-related cancer risk.Rad Res. 2007; 167: 396416Crossref Scopus (560) Google Scholar included individuals exposed (or likely to have been exposed) to internally deposited radionuclides or to neutrons were included, which tripled the person-years available and increased the number of deaths despite a reduced number of workers included. The results show an increased risk for leukaemia, with an excess relative risk of 2·32–3·68 per Gy depending on the latency period, whether socioeconomic status and internal deposition were adjusted for, whether the analysis was restricted to different dose ranges, and whether all or only two countries were included in the analysis. Most of these excess risks were statistically significant, but they were based on 90% CIs. In total about 70 CIs were presented in the paper. With this high number of statistical tests the chance of finding statistically significant results is high, so the danger of false positives is not negligible and should be taken into consideration when interpreting the results. The study provides supportive evidence on the radiation risks of leukaemia after exposure to low doses, using a large dataset and adequate statistical analysis. But limitations of this type of cohort study of nuclear workers have been discussed by Leuraud and colleagues and other investigators before,8Boice J The importance of radiation worker studies.J Radiol Pro. 2014; 34: E7-12Crossref PubMed Scopus (10) Google Scholar and they somewhat hamper the study's conclusions. Heterogeneity between countries is present but not well understood, and its assessment can be a major challenge for a pooled analysis. The contribution of errors in the outcome variable (death certificates from different countries covering more than 50 years are included) is not known. Confounding by socioeconomic status and other lifestyle factors cannot be assessed completely. Additional risk factors, such as exposure to benzene and medical exposure to ionising radiation are not taken into account. Internal exposure to radionuclides, uranium, and plutonium are neither qualitatively nor quantitatively evaluated. Background radiation exposure might be larger than occupational exposure and was not incorporated into the analysis. In my view, to properly understand the mechanisms and effects of low-dose radiation, we need new data collected by comparable methods for all participants: excellent dosimetry for internal and external exposure, including organ doses, data for exposure to ionising radiation from other sources such as medical and background exposure, data on other known occupational risk factors, lifestyle factors, biological material, genetic markers, and medical history, including information on screening and medical care. We need longitudinal (prospective) data (as we have from atomic-bomb survivors in Japan) to help us understand the dose–response relationship, interaction and confounding, and outcome pathways.9Preston RJ Integrating basic radiobiological science and epidemiological studies: why and how.Health Phys. 2015; 108: 125-130Crossref PubMed Scopus (31) Google Scholar We also need more sophisticated statistical analyses: not only for dose measurement errors,10Thierry-Chef I Richardson DB Daniels RD et al.Dose estimation for a study of nuclear workers in France, the United Kingdom and the United States of America: methods for the International Nuclear Workers Study (INWORKS).Rad Res. 2015; (published online May 26.)https://doi.org/10.1667/RR14006.1Crossref PubMed Scopus (46) Google Scholar but also to deal with confounding, with errors in the outcome variable, with heterogeneity, multiple testing, and to model the dose–response relationship. Workers and the public can be protected from radiation by controlling the dose to prevent adverse health effects, mainly cancer; however, a better understanding of the effect of low-dose radiation is warranted. Large studies like INWORKS support current risk estimates, but new, creative prospective studies that include biological material and collaboration between radiation biologists and radiation epidemiologist are needed to clarify how low-dose radiation affects human beings. I declare no competing interests. Ionising radiation and risk of death from leukaemia and lymphoma in radiation-monitored workers (INWORKS): an international cohort studyThis study provides strong evidence of positive associations between protracted low-dose radiation exposure and leukaemia. Full-Text PDF Open Access

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.