|Comments from Switzerland on the draft report from the ICRP “Lung cancer risk from radon and progeny” (ICRP ref 4843-4564-6599, July 27, 2010)
1. The report is of high significance and arrives at an opportune moment following the publications on the analyses of pooled studies on domestic radon exposure. The consistency of the results of the three methodologies, namely the cohort studies of miners, the case-control studies on residential exposure and the application of the dosimetric model for exposure of the lungs and the use of risk factors derived from epidemiological studies on the survivors of Hiroshima and Nagasaki, now provides a solid basis for estimating the risks from radon.
2. The organisation of the report is excellent. In particular, we welcome the fact that the discussion of the reference levels which appeared in a preliminary version has been omitted, thereby enhancing clarity and focus. It may however, have been opportune to leave out the appendix B on radon dosimetry, which is not the specific topic of the publication.
3. The connection between radon and smoking addressed in paragraph 3.3 tends to predict that the excess relative risk (ERR) is constant, as is shown in the European pooled analysis (paragraph 25 of chapter 2.3). The consequences of this observation are not sufficiently clearly stated in the conclusions of the report. One should not avoid transparency on the pretext that the result could negatively influence the awareness of the population towards the risk of radon. It should be explicitly stated that, based on our present understanding, smokers are exposed by about 25 times more to the risk of radon than are non-smokers.
4. The revised assessment of the detriment from radon (LEAR) (chapter 4) represents an important part of the report. The elements required for this calculation are indicated in paragraph 40. The percentage of smokers in the population needs to be added to the suggested list. In fact, if one assumes the concept of a constant ERR, then this element is fundamental. The models used by different organisations and authors to convert ERR values into LEAR values are cited. Moreover, the Task Group itself performed different calculations using these models (see paragraph 47, chapter 4.2). It would be very useful if the chosen models and their accepted parameters (death rate from lung cancer, fraction of smokers in the population) were expressed in a more explicit manner. This could be presented for example in an appendix. In fact, this proposal is justified because this calculation is at the very heart of the problem. This approach would provide greater transparency in regard to the distribution of risk across the population, perhaps even questioning the concept of converting risk into effective dose (concept based on the fact that the individual risk is not too variable, thereby allowing an average risk to be used, which is not the case for radon exposure).
5. In the report (§ 56 of chapter 5) it is indicated that the ICRP is considering providing in the near future the dose coefficients based on the HRTM model and practical exposure scenarios. In the interim period it is recommended to use the old factors, in spite of the fact that it is known that they underestimate the risk by a factor of 2. We believe that in the interim period, on a temporary basis, a more reasonable proposal would be to multiply the dose coefficients by two.