Public Health England welcomes the opportunity to comment on this ICRP document that aims to clarify the appropriate use of effective dose in radiological protection. As you are aware Public Health England contributes significantly to the work of ICRP and as the UK authority with responsibility for providing advice on radiological protection, is always keen to review ICRP publications which may affect the advice it provides.
The comments enclosed below were collated from contributions from several members of staff. If you require to discuss any of these comments in detail please contact Mr T Cabianca (email given above).
Lines 110 – 113: “a) Equivalent dose (H) is not required as a protection quantity. It will be more appropriate for limits for the avoidance of tissue reactions for the hands and feet, lens of the eye, and skin, to be set in terms of absorbed dose (Gy) rather than equivalent dose (Sv).” That is acceptable for photons and electrons and beta particles, but RBE weighting will be required for other particles such as alphas and neutrons.
Lines 113 – 116: “b) While risk assessments for individuals based on organ/tissue doses and specific dose-risk models make best use of scientific knowledge, E may be used as an approximate indicator of possible risk, recognising that this is a pragmatic, but unintended, application of effective dose.” This is a bit clumsy: “approximate indicator of possible risk” is a problematic expression. It seems to be saying that E can be used for risk assessments for individuals.
Lines 113 and 120 (also 268 and 277 and throughout section 5 and 6): The document makes use of the phrase “possible risk” in relation to using effective dose as a risk indicator but it is not clear what is meant by “possible” in this context.
Lines 185 – 187: “The dosimetric quantities used in radiological protection are absorbed dose (D), with the special name of gray (Gy), and equivalent dose (H) and effective dose (E), both with the special name of sievert (Sv); the SI unit is J kg-1 in each case.” These are special units, not names. Replace ‘names’ with ‘units’
Lines 190 – 191: “Absorbed dose is the most appropriate quantity for use in setting limits on 190 organ/tissue doses to prevent tissue reactions (deterministic effects).” Again, this is only true where there no RBE considerations.
Lines 221-226: “Although E will generally be used at doses below 100 mSv, its use exceptionally in emergency exposure situations at acute doses in the range up to around 1 Sv is reasonable, noting that the possibility of occurrence of tissue reactions should also be considered at such doses if a significant contribution is made by non-uniform distribution of external dose or radionuclides concentrated in specific tissues/organs.” This is an odd statement. The risk estimates on which E is based are from exposures > 100 mSv, where there is seen to be a statistically significant dose-effect relationship. But, the report seems to be saying that the risk estimates based on those dose levels should only be used for lower dose levels, where there is no statistically significant dose effect relationship.
Lines 281 – 287: “Collective effective dose is a valuable tool in the optimisation of protection, particularly for occupational exposures. It is not intended for use in risk projection. Its use to predict potential/possible health effects should be treated with great caution, put into context and judged in relation to baseline lifetime morbidity risks. For public exposures, components of dose integration in time and space should be considered in estimating collective doses, particularly when considering exposures of large populations over very long periods of time.” The last part of this statement should emphasise that it really relates to large populations that have received small individual effective doses. ICRP may have trouble convincing with the first part – if effective dose is based on the best available data for population exposures, it is only natural that it will end up being used for collective dose calculations and projection of risk.
Line 293: ICRP introduced effective dose equivalent in Publication 26, which differed in name and definition from effective dose; it was based on the Q(L) weighted absorbed dose, the wR concept had not been introduced. Effective dose was not introduced until ICRP Publication 60.
Line 323: “it does not provide estimates of dose to specific individuals” should perhaps read “it does not provide estimates of risk/detriment to specific individuals”
Lines 620 – 632: Table 2.1 is copied from ICRP Publication 103. However, whilst the text of ICRP 103 notes that the testes are amongst the most radiosensitive organs, the table only includes the detriment analysis for the ovary (ovaries?). In effective dose the two are averaged. Should this read “ovaries and testes” in both publications or is the only measurable detriment to the testes due to hereditary effects? If this does not include the testes in the male phantom, then the average for the gonads may struggle to justify the higher wR. It is assumed that the breast values in both publications are the average of male and female? The specification of brain and salivary glands as having defined wT values, without inclusion of their data in the tables, is odd: the other 14 organs in the remainder have a higher mean detriment than the two wT=0.01 tissues that do have data provided (bone surfaces and skin). The division between remainder and specified organs is hence somewhat contentious.
Line 830: change “fndamental” to “fundamental”.
Line 1096: “radionuclides with low energy beta or alpha particle emissions”. This is a little unclear. The issue relates more to higher energy alpha particle emissions, which reach 50-60 microns in tissue, but do not reach 70 microns. It would be better to say “radionuclides with alpha particle or low energy beta particle emissions”.
Lines 1147 – 1149: the use of H*(10) to estimate effective dose will be very imprecise for many reasons. This is perhaps justifiable in very low dose rates, but the acceptability in other circumstances is very questionable.
Lines 1150 – 1163: the use of the kerma approximation is in fact robust for typical workplace fields, but breaks down for high energy fields that may be encountered in cosmic ray or accelerator environments. The problems with lower energies only relate to photons: the text should reflect this. The claim that the new proposals will lead to a simplification of the system is contentious: many may consider the new system significantly more complex, especially for those who try to calculate or measure the proposed quantities.
Lines 1233 – 1239: this equation has several flaws. Firstly, S has a functional dependence on DT, but DT does not appear on the other side of the equation. For an integration, E will need to have a functional dependence, which it is integrated over, but no functional relationship is likely to exist. In practice this will be a summation, not an integration. From Emin to Emax, not E1 to E2. This would be an ideal point to define Emin below which collective effective dose should not be used.
Lines 1266 – 1280 (70): is this definition of effective dose exclusively for occupational exposures (i.e. defined explicitly in the adult male and female reference phantoms, averaged). The ICRP draft on environmental exposures evaluates effective dose for adults, 15 year olds, 10 year old, 5 year olds, 1 year olds and new born babies, which conflicts with the definition here.
Lines 1288 – 1295: the p in Hp(d) should be lower case.
Lines 1310 – 1312: Para 72 refers to the use of environmental monitoring of air as being used in “exceptional circumstances”. For radon in homes and workplaces, the main mechanism of determining exposure (in terms of radon concentration or interpreted as effective dose where this is needed) is to make a long term measurement of radon concentration in air. Since radon exposure is the most widespread source of exposure requiring control and is the largest source of exposure for many population groups (including UK) it is not correct to describe this as “exceptional”.
Lines 1343 – 1353: Para 75 suggests that effective dose is used for controlling existing exposures. For radon exposure, most exposures are managed in relation to a Derived Reference Level (in terms of radon concentration in air) rather than direct use of effective dose. The document should make reference to this, perhaps by noting that “in many cases occupational exposure to radon is controlled through the use of a Derived Reference Level, expressed as an annual average radon concentration in air, rather than through direct use of effective dose.”
Line 1405: Delete double quotation marks from “representative person”.
Lines 1408 ‑ 1410: The sentence starting with “Ideally such assessment would …” needs to be redrafted/deleted. This is a weak statement. For planned exposure situations, ie assessment for authorisation, no monitoring of people should take place to avoid causing unnecessary anxiety. Only monitoring of the environment is performed.
Line 1411: “Collective effective doses may also be estimated as an input to the optimisation process …”. The report does not provide clear information on how collective dose can be used in current optimisation process is not clear, more specific comments on this topic follow.
Line 1422: Replace “hypothetical” with “representative”. ‘Hypothetical person’ has no clear meaning in radiation protection.
Line 1425: “Measurements of people and the environment can be used, if available, for retrospective assessments…”. The report needs to state more clearly that measurements of doses in people should be limited to assessment of doses for some existing exposure situations, ie dose in contaminated areas following an emergency or past practices. Retrospective assessments for planned exposure situations (ie for compliance with authorisations) are only based on environmental monitoring.
Line 1432: Replace “from radionuclides” with “to radioactivity” (full sentence should be “Members of the public may be subject to external or internal exposure through various pathways to radioactivity dispersed in natural or inhabited environments.”).
Line 1434: Add “further” between “what” and “actions” (ie “what further actions”).
Lines 1434 ‑ 1436: “Effective doses are estimated as input to the optimisation process and for comparison with relevant reference levels” should be changed to “Effective doses are estimated for comparison with relevant reference levels as input to the optimisation process”. Comparison with reference levels is the key element of ICRP recommended optimisation process. As it was written it is implied that the comparison with reference levels is not part of the optimisation process.
Lines 1436 – 1438: “Depending on the nature of the release ...”. There is nothing wrong with this sentence but, given that other parts of this report make the argument for not using equivalent dose for dose limit, this paragraph should clarify what the implications of this recommendation would be for use of effective and equivalent dose.
Lines 1438 – 1439: “Emergency exposures are usually of short duration …” it should be explained what ‘short’ means? Hours, days? And short compared to what? Also it is not necessarily the case that exposures are of short duration.
Lines 1439 – 1440: “differences in dose as a function of age at exposure”. It is not clear what this expression means as doses for emergency exposure situations are generally not calculated as a function of age at exposure (they may be for epidemiological studies but they are not for the purpose of emergency planning and response). It would be better to say “differences in dose for different age groups”. Also it is not clear why a statement about differences in dose for different age groups is in the same sentence as something about short duration; it doesn’t read well.
Lines 1440 – 1441: “Consideration of exposures of pregnant and breast-feeding women …”. This is also a consideration for planned exposure situations (eg for 89Sr, 32P, 33P, 45Ca) not just emergency exposure situations.
Lines 1445 – 1447: “It is important to recognise uncertainties …”. This depends on when the measurements are carried out. In the immediate aftermath, there will be high levels of uncertainty but as more information becomes available over time, the assessment of doses becomes more certain.
Lines 1450 – 1451: “Retrospective assessments can also be used to refine the prospective dose assessments to reduce uncertainties and to improve the optimisation process” is confusing; it is not clear what it means. Perhaps some one of more examples could be given to clarify.
Lines 1452 – 1535: Paragraph (84) to (89) should be moved to the beginning of the section 4.2 as an introduction.
Lines 1459 – 1460: “Internal exposures can occur by inhalation of radionuclides in the air or by ingestion of radionuclides in food or water.” Skin absorption (eg for 3H) can be another pathway.
Lines 1463 – 1464: “assessed for a real or, more frequently, hypothetical person receiving a dose that is representative …” is misleading and should be replaced with “assessed for a person or group of people receiving a dose that is representative …”. Doses for optimisation purposes are never calculated for real members of the public. The term ‘real’ is probably used to mean that the habit data of the representative person may be based on data for ‘real’ people (ie residents in the area) rather than generic values.
Lines 1464 – 1465: “(the right tail of a distribution of individual doses within a particular cohort)” is confusing as there not a unique way to identify the representative person. Also the term ’tail’ is too generic and unclear.
Lines 1465 – 1466: Delete single quotation marks from ‘representative person’ (twice).
Line 1466: Delete single quotation marks from ‘critical group’.
Lines 1466 – 1467: “The concept of the ‘representative person’ was introduced in Publication 101 (ICRP, 2006) to replace the less quantitatively defined concept of the ‘critical group’”. It would be useful to include the actual definition of representative person and also explain why critical group was less quantitatively defined.
Lines 1470 – 1473: Mention of reference person is confusing (can be confused with representative person) and not relevant here. Most dose assessments rely on tabulated dose coefficients which are already averaged for different genders. We suggest that the 2 sentences should be changed to: “In the dose assessment process, a number of different age groups can be considered. Typically doses from internal irradiation are assessed using tabulated dose coefficients for different age groups calculated by ICRP (ICRP, 2012b). The full set of six age-groups considered by ICRP are the 3 month-old infants, 1 year, 5 years, 10 years, and 15 years old children and adults.”
Lines 1476 ‑ 1484: “However, this is not always the case …. If a more detailed assessment is required, the annual intake by the mother should be assumed to occur over the nine months of pregnancy and three months of breastfeeding.” This guidance comes from the report Guidance on the application of dose coefficients for the embryo, fetus and breastfed infant in dose assessments for members of the public, Documents of the HPA, publication RCE 5 published by HPA in March 2008 and not from the ICRP as it is implied here. A reference to the HPA document should be included.
Lines 1484 – 1486: ICRP is providing conversion coefficients for age related external exposures in the draft publication currently available for comment. This text will be obsolete as soon as that document is published – coherence between the drafts is required.
Lines 1488 – 1489: “Publication 101 (ICRP, 2006) concludes that consideration of three age groups, 1 year and 10 years old children and adults, is sufficient for most dose assessments …” The rationale for calculating doses to children and infants for regulatory purposes is not clear. There may be specific situations which require assessment of doses to children and more particularly infants but in general the difference in doses between age groups (especially between adults and 10 children) is not significant when compared to the uncertainties in the dose calculations. For the purpose of compliance of doses with a dose limit or constraint as a general rule it would be sufficient to calculate doses for adults only especially when site specific information is not available and therefore general assumptions have to be made on the habits of the representative person.
Lines 1491 – 1492: “In general, uncertainties in estimating exposures are large in comparison with differences in dose coefficients for different age-groups”. This sentence seems to contradict the previous sentence on the necessity to calculate doses for three age groups for regulatory purposes (ie for comparison with dose limits, dose constraints and reference levels) unless determined by specific local conditions.
Lines 1520 – 1521: “This is particularly important when determining whether actions, such as evacuation or decontamination, are required in an emergency exposure situation.” It is important to recognise that when deciding if evacuation should be implemented the key unknown parameter is the source term (and the duration of the release) and therefore matter less whether the other parameters are cautious or otherwise.
Line 1521: Delete “or decontamination”. For an offsite release from nuclear licensed sites, detailed emergency plans will be used to automatically trigger evacuation. The benefits and harms of evacuation will have already been assessed as part of emergency planning.
Line 1532: Perhaps it should be “not available” (ie “information on radionuclide characteristics is not available”).
Line 1532: Add “or other suitable information” after “monitoring data” (ie “When no monitoring data, or other suitable information, are available …”).
Line 1534: “… and dose coefficients that result in higher dose estimations”. It should be recognised that ICRP Publication 72 contains guidance on the most likely chemical form for material in the environment and that these are frequently not the most cautious chemical form available (ie most defaults lung absorption rates are type M rather than type S).
Lines 1543 – 1544: “For public exposures, collective effective doses can be used as part of the optimisation process for planned, existing or emergency exposure situations.” This section does not elaborate or provides details on how collective doses can be used for optimisation purposes. In the current system optimisation is performed using comparison of individual doses with dose constraints or reference levels. No equivalent criteria exist for collective doses; therefore it is not clear what a collective effective dose means in terms of risk. As a result collective doses can only be relative quantities used to compare one option with another. In the past criteria for collective doses existed (for exemption a criterion of 1 manSv was used for exemption) but are no longer used. In the past collective doses was used as input to Cost Benefit Analysis, but currently optimisation is generally not performed through the use of CBA or collective doses but by using other techniques (BAT) which do not require CBA. ICRP should state the case for using collective doses to the general population for optimisation purpose as this calculation does not seem to be required.
Lines 1547 – 1549: this is an ideal opportunity for ICRP to define the minimum acceptable dose to be used for collective dose calculations. The use of the LNT for nSv exposures is absurd, since there is no evidence for detriment AND no possibility that the doses can be determined with any accuracy. The text higher up implies that the minimum effective dose could depend on the population size, which would need some supporting justification (it is probably not true). Table 5.2 only considers assessments of risk from doses in the 10-100 mSv as being based on LNT, with below 10 mSv the risks being ‘inferred’. Some consistency between the doses range where collective dose is useful and the applicable range for LNT is needed.
Lines 1553 – 1555: “For example, following a severe nuclear accident …. to give an indication of possible health impact to help with planning and selecting from various protection options.” This sentence seems to contradict paragraph (66), which states that ‘the computation of cancer deaths based on collective effective doses involving trivial exposures to large populations is not reasonable and should be avoided (ICRP, 2007a). The same paragraph suggests that collective doses should be used for comparative studies rather than health impact assessment. Additionally ICRP should provide more detailed information on how the calculation should be performed. For example what population is considered, what integration time? If vast majority of people receive individual doses of the order of nSv to µSv should these doses really be included in the estimation of the overall collective doses? These doses are really trivial to be included.
Lines 1556 ‑ 1558: “In retrospective assessments of planned or existing exposure situations, … to inform medical and epidemiological evaluation.” This sentence is a bit vague and would benefit if some practical examples were provided. It is not clear what this means in relation to planned exposure situations, maybe it is valid for existing exposures (eg post-accident remediation). We suggest that the term “dose reconstructions” should be used instead of 2retrospective assessment”. Retrospective assessment is generally used for assessments performed to show compliance (eg discharges), not for assessment of doses actually received by members of the public.
Lines 1558 – 1564: “It is essential that such analyses using collective effective dose … and that comparisons with natural disease incidences determine whether epidemiological analyses may provide statistically significant results for populations.” Again this sentence seems to indicate that collective doses can be used in some type of risk assessment which seems to contradict what is stated in paragraph 66.
Lines 1565 – 1578: “As discussed in Publication 101 (ICRP, 2006) and Publication 103 (ICRP, 2007a), it is recommended that … and distribution in space (local, regional, global), and time (short-, medium- and long-term). However, there are problems in implementing such recommendations for public exposures … collective dose estimates are based on food production data, and the distribution of individual doses is not known (Smith et al., 2006).” The first 2 sentences seem to contradict the other two. On one hand the first two sentences say that ICRP 103 recommends that collective doses should be split according to bands, on the other the second two say that this disaggregation cannot be done for public exposures. Therefore, what is the ICRP recommendation? Do need to disaggregate or not? Doesn’t the fact that we cannot disaggregate the collective effective dose implicitly indicate that the calculation of collective doses is not that useful?
Lines 1578 – 1580: “However, collective effective doses can be estimated for specific population groups living in defined geographical areas over different time periods as discussed in Publication 103 (ICRP, 2007a).” This recommendation is too weak and generic and ICRP should provide more details (and make stronger recommendations) on both the population and the integration time. In the UK we recommend to calculate collective dose at 500 years (and sometimes 100,000 years) for the UK, European and World population for authorisation/permitting assessments. Both the integration time and the populations recommended are far too wide for the collective doses to be meaningful and can also be easily misused. Collective doses should be calculated for a much shorter population (eg up to a few thousands of people living around the installation) and shorter integration times (eg no more than 100 years).
Lines 1580 – 1582: “Per-caput doses can also be estimated … which can provide useful input to optimisation and comparative studies (Smith et al., 2006).’ It is debatable whether per‑caput doses can be useful at all in the optimisation process for the following reasons: firstly there are no dose criteria which can be used to compare per-caput doses with (dose limits, dose criteria and reference levels apply to individual doses to the representative person) and therefore it is not clear what purpose a per-caput dose would serve; secondly per-caput doses can be of the order of less than 1 µSv, that is a negligible level and therefore do not provide any indication of health impact; thirdly whilst collective effective doses are integrated over a period of hundreds of year per-caput doses can’t be and are generally given as annual values and their calculation from collective doses is not straightforward. ICRP should provide more details on how to calculate per-caput doses and explain more clearly how they can be used in the optimisation process in a practical way.
Lines 1589 – 1590: “… dose assessments can contribute appropriately to decision making for periods spanning a few generations but …”. It would be helpful if ICRP clarified what period “a few generations” includes (500 years).
Line 1596: Add “dose indices” before “quantities” (ie “in terms of dose indices, quantities that can be measured…”). We suggest that the overarching term ‘dose indices’ should be used term for diagnostic and interventional dose quantities here and as appropriate through the rest of the text. We are aware that not of all of the dose indices include this term in their titles but this is a useful term and indicates that this is a surrogate for dose measurement.
Line 1667: We suggest that a brief paragraph should be added to Section 5.1 on deterministic risk, highlighting the risk of deterministic effects to skin, especially from some interventional procedures for larger patients. This may be mostly a reference back to section 2.2, and other appropriate documents by ICRP and others. The reason why we believe that it would be useful to include this section is that while it is not appropriate to assess skin dose in terms of effective dose, deterministic skin effects are a potential consequence to the patient which occurs soon after exposure and are likely to trigger alarm, including worry about developing cancer. It might also be appropriate to reference the potential for staff to receive deterministic eye doses; and the importance of mitigating this hazard by using the appropriate PPE.
Line 1692: “considerably from values obtained for adults.” Reference(s) should be given.
Line 1772: ‘Tracking of patient dose’. This is a topic which can be considered contentious with respect to its radiation protection benefits to the patient. For what purpose will the cumulative medical dose to an individual patient be tracked? Will it be for research into epidemiology, or reputedly for the patient’s benefit? If for the patient’s benefit, how? All individual diagnostic or therapeutic radiation exposures are (or should be) justified on the basis of the current medical benefit to the patient of having that exam/treatment, and that decision should not be affected by what previous doses the patient has received. At best keeping a record of what tests a patient has undergone may prevent diagnostic exams being repeated unnecessarily but one does not need to know exam doses to make that decision. If only dose indices values are recorded it may well be very difficult to assess effective dose from that data at a later time if not enough information about the individual exposures are retained. As discussed elsewhere in Chapter 5 dose indices values may give a deceptive indication of effective dose if a non-standard exposure protocol is used. If the data is not being collected for the patient’s personal medical benefit then there are likely to be legal data protection and privacy regulations involved in the collecting and retaining the data and around permitted access to this information. This would certainly be the case within the EU At the very least there will be a need for informed consent from the patient to collect this information. Ethically, the patient must be accurately briefed on how medically beneficial or not collecting and retaining this information is likely to be to for them personally. One reason patients may think it is useful to have this data is that it could be used to sue for malpractice later, should they develop cancer. Given that the ICRP is a promulgator of radiation protection advice it would be good if a brief note of caution about the usefulness of this practice is added to this paragraph.
Lines 2046 – 2054: The use of 1 year, 10 year and adult for effective dose seems in conflict with paragraph 70 and also with the finer age grid used in the new environmental exposures draft publication.
Line 2048: “Person” should be in lower case (ie it should be ‘person’ rather than ‘Person’).
Line 2049: Delete double quotation marks from “critical group”.