WHO and IARC comments to the draft ICRP recommendations Sept.15, 2006 The draft ICRP recommendations have been reviewed by radiation experts from the World Health Organization and the International Agency for Research on Cancer. The following technical comments reflect the scientific assessment of these experts and not a formal WHO/IARC policy position. Further comments are provided by PAHO, the WHO regional office for the Americas.. General comments: Overall, WHO and IARC staff recognize the efforts of ICRP to provide updated recommendations on radiological protection, and welcomes the opportunity to contribute to the development of these recommendations. The current draft is well developed, even though certain concepts (e.g. constraints / dose limits) continue to pose challenges in terms of ease of explanation. One way to respond to this challenge could be the inclusion of more practical examples e.g. describing the concept of constraint in a specific situation. Evidence from the Chernobyl accident (as summarized in the recent WHO report on health effects) clearly points to the specific need for radiation protection of children, in particular with respect to thyroid cancer development. We support a stronger focus on this issue in the document, preferably with a separate subchapter in chapter 5. We would like to suggest that particular care is taken to further simplify the language of the recommendations. ICRP recommendations are read and translated by persons with many different language backgrounds. The applicability of the recommendations in low-resource settings will be enhanced by ease of understanding and an utmost degree of clarity. Specific comments: (29) In the following sentence: "Both high and low doses may cause stochastic effects, which are cancer or hereditary effects, which may be observed as a statistically detectable increase in the incidences of these effects occurring long after exposure", words cancer or hereditary effects are used as the only two possible outcomes of stochastic nature, which is not exactly accurate claim in the light of emerging evidence of other non-cancer diseases being associated with radiation exposure, such as cardio-vascular diseases. Consider rephrasing as follows: "Both high and low doses may cause late health effects, e.g. cancer or hereditary effects, which may be observed as a statistically detectable increase in the incidences of these effects occurring long after exposure" (49) We are concerned that the descriptive terminology of tissue reactions still remains vague. Sometimes the term appears to mean tissue damage reaction, and sometimes it appears to be used as indicating tissue reaction, meaning a physiological response. A deterministic effect is not a physiological tissue reaction but a tissue damage reaction induced by radiation exposure. Some further clarification is recommended. (54) and following There are relatively acute (short-term latency) and chronic (long-term latency) effects on cancer induction as seen in various epidemiological studies. The rapid increase of childhood thyroid cancer around Chernobyl is an ideal model to discuss the latency of radiation-induced thyroid cancer but in this section there is no specific discussion on the molecular mechanism of Chernobyl thyroid cancers. (57) We wish to emphasize that there are situations involving radiation exposures of the public where stakeholders, public health officials and the interested public strongly request information on possible population effects. This is important in particular for the planning of health services, to put the consequences of the accident in the broader public health perspective and to communicate risks to the affected persons/publications. We do not support the ICRP argumentation that the LNT is solely a radiation-protection and not a risk-estimation or -projection concept. The US NAS BEIR VII model, for example, is a risk estimation model - see for example Executive summary p.15 - "The committee has developed and presented its best possible risk estimates for exposure to low-dose low-LET radiation in humans subjects" and "The committee concludes that current scientific evidence is consistent with the hypothesis that there is a linear, no-threshold dose-response relationship between exposure to ionizing radiation and the development of cancer in humans". In BEIR VII, this model was applied to the US population and to background radiation exposure levels to project cancer risk. The BEIR VII model was also applied to the European population to evaluate the possible burden of cancer incidence and mortality which may have resulted to date and may be expected in the future due to radiation from the Chernobyl accident (Cardis et al, 2006 - work of a working group which included three members of the BEIR VII committee). In both instances, the limitations and uncertainties of the risk projections were extensively discussed in particular in terms of uncertainties in the magnitude of the risk at very low doses, uncertainties in dose estimates, uncertainties in population structures and mortality and incidence rates over time. We believe that such carefully conducted predictions are in fact useful and important for public health planning and risk communications, as long as they are based on sound and well presented assumptions and that the magnitude and sources of uncertainties are clearly indicated. Concerning the adequacy of the LNT model at very low doses, we would also like to cite the recent work by Brenner (2006) which, like BEIR VII, concludes, based on epidemiological data, mechanistic data and biophysical arguments, that even the lowest dose has the potential to increase the risk of cancer. While there are uncertainties about the exact magnitude of this risk at very low doses, the effect could in fact be either lower or higher than that predicted by the LNT. We therefore recommend that the current statement be modified: the ICRP reservations concerning the uncertainties of such predictions should be noted clearly. The usefulness of carefully conducted and qualified predictions for the purposes listed above should however be recognised. Such evaluations of the possible health impact of radiation exposure should not be dismissed so cavalierly. Furthermore the term "formal purposes of public health" is unclear. We recommend to delete this term. Line 3 of this article – we presume what is meant is “unambiguously” (146) The sentence “In the case of low individual dose which are small fractions …, the use of collective dose for risk estimates is not a reasonable procedure as it both aggregates too much information for the decision making process and combines several sources of uncertainty.” is unclear. It is not clear what is being meant here – that collective doses should not be estimated? That they shouldn’t be used but for what purpose? What “decision process” does this refer to? (147) This paragraph appears unnecessarily defensive and prescriptive. We suggest that it is revised simply as follows: “Collective dose is mainly an instrument for optimisation, for comparing radiological technologies and protection procedures. Collective dose is not a useful or adequate tool for epidemiology” (162) Second bullet point states: " emergency situations are unexpected situations that occur during the operation of a practice, requiring urgent action. Emergency situations may arise from practices”. This definition is incomplete and needs to be refined/elaborated as it appears to exclude accidental exposures of general public, e.g. Goyania accident and many other similar events involving orphan sources found at scrap yards, etc. (164) It is not exactly clear how the three identified categories of exposure correlate to the types of exposure situations defined in (162). It would be useful to state here that for each exposure category all three types of exposure situations may apply (e.g. planned occupational and emergency occupational, planned medical and accidental medical over-exposure, etc). Similarily, it should be more clearly stated whether the population-based screening of healthy persons (i.e. mammography) falls under medical or public exposures. These persons are not commonly regarded as patients as implied in (168). (177) Concerning the protection of the fetus, it seems more consistent with other statements that, after declaration of pregnancy, working conditions should be such that a limit of 1mSv for the total time of the pregnancy is not exceeded. (230) We are not in agreement with the second half of this paragraph – “However, such calculations can be a useful tool for preliminary judgments to examine the feasibility of an epidemiological study in a specific situation, or the plausibility of attributing observed health effects to a source of exposure.” We believe that the collective dose is not a useful tool to examine the feasibility of an epidemiologic study – what is needed for this purpose is at least crude information about the distribution of exposure in the population – this includes the mean or median of course but also quite importantly the range, which is not given by the collective dose -. Similarly the “plausibility of attributing observed effects to a source of exposure” is better covered by the text in paragraph (229) that states that “each relevant exposure situation must be carefully analysed to identify the individual characteristics and exposure parameters that best describe the exposure distribution among the concerned population for the particular circumstance.” We suggest therefore that paragraph 230 is removed entirely. The first sentence is implicit in paragraph 229 and is stated in a number of places in this document. If it needs to be repeated here, it should be in the first sentence of paragraph 229. (301) a) With respect to radon, constraint and Action level are seemingly used interchangeably. If not intended, this needs clarification. b) The constraints for radon in dwellings (interpreted as Action level values) seem out of date and are, as stated, maintained for the sake of practicality and continuity only. In practice, however, most countries have reduced their Action levels much below 600 Bq/m3. The new epidemiological data cited in previous paragraphs provide evidence that most health effects are initiated well below 100 Bq/m3. WHO therefore finds it insufficient to merely accept that countries use lower Action levels, as stated in para (300). Proactively, ICRP should recommend lower constraints, more in line with recent indoor data. In addition, the issue of prevention of high radon levels in the construction of new housing stock should be named and discussed more prominently in the Recommendations. This is an important way to achieve reduction of radon concentrations at population-level, and thereby an effect on population lung cancer rates. (310) The following sentence at the end of the paragraph may be deleted (or modified otherwise): "Thus dose estimates should not be regarded as measures of health detriment beyond times of around several hundreds of years into the future. Rather they represent indicators of the protection afforded by the disposal system", as it is not clear what is the significance of health detriment beyond several 100s of years while ICRP Recommendations change every 10-15 years. (335) Consider adding a sentence about malicious events and nuclear weapon explosions as an additional type of emergencies. (338) Throughout the entire document, the terms deterministic effects and tissue reactions are used interchangeably. It seems to be more consistent to use one of them or use one followed by another in brackets. E.g.. at the end of the (338): "Nevertheless, at some level of dose, approaching that which would cause tissue reactions, some kind of intervention will become almost mandatory" may read better if "deterministic effects (tissue reactions)" was used. 9.2. Do "existing situations" apply to natural sources only? If not, it may be useful to define better how long it takes to consider a situation as an existing situation? E.g., twenty years past the Chernobyl accident makes the contaminated territories "an existing situation"? Or sixty years past the atomic bombs in Japan? (359) Please insert the following sentence before the last sentence about ILO's role: "The World Health Organization uses ICRP Recommendations as a reference for its guidelines and policies on radiation health and safety.