Register for Updates | Search | Contacts | Site Map | Member Login

ICRP: Free the Annals!

View Comment

Submitted by Gerald Kirchner, Federal Office for Radiation Protection (BfS), Germany
   Commenting on behalf of the organisation
Document Foundation docs Optimisation; Dose to Individual

Federal Office for
Radiation Protection

Willy-Brandt-Straße 5
38226 Salzgitter

 + 49 1888 333 4100

Berlin, 28 June 2005

International Commission on Radiological Protection (ICRP): Assessing Dose of the Representative Individual for the Purpose of Radiation Protection of the Public
Comments on the Draft for Consultation (web version); version 18.1 from 5 April 2005

1. Introduction
The International Commission on Radiological Protection (ICRP) is currently drafting new general recommendations on radiological protection which will be supported by a number of foundation documents. One of these foundation documents Assessing the dose of the representative individual for the purpose of radiation protection of the public should assist in defining the individual to be used for determining exposures in the ICRP system of protection. A draft version of this document has been made available by ICRP for input and comments by the international radiation protection community.
The German Federal Office for Radiation Protection (Bundesamt für Strahlenschutz, BfS) has always been interested in the development of new concepts in radiation protection and therefore followed with great interest the steps taken by ICRP to develop new recommendations. At several occasions, representatives of the BfS have been involved in discussions initiated by ICRP seeking feedback on their draft papers.
We are grateful that ICRP has given us the opportunity to comment on the draft version of “Assessing Dose of the Representative Individual for the Purpose of Radiation Protection of the Public”. We offer the following comments that will focus on items which, in our opinion, would profit from further discussions and clarification.
2. Implications on basic principles of radiation protection
2.1 General applicability of constraints
Setting limits and constraints both for workers and for members of the public has always been a key issue in radiation protection. One of the basic principles of radiation protection stipulates that limits and constraints shall apply to each individual in a society.
Paragraph (S16), however, seems to indicate that only the majority (i.e. 95 %) of the population shall be protected to the degree given by the constraint. Up to 5% of the population may receive doses above the dose constraint.
This is underlined by the statement in paragraph (19), that the Commission “...believes there should be a small probability that some individual may exceed the dose constraint.”
This fundamental change in protection philosophy has to be based on convincing arguments, which are not given in the presented (rather technical) draft. Changes in the protection philosophy should be discussed in the ICRP general recommendation on radiation protection rather than in any of the foundation documents and would profit from a broad stakeholder involvement and discussion.
In addition, the use of a 95th percentile to decide on compliance with a constraint is questionable. As no rationale is given for this value, it seems to be taken from statistical hypothesis testing. It should be remembered, however, that (i) its use is nothing else but a convention between statisticians, (ii) that “statistical significance” is different from “relevance”.
2.2 Constraints and radiation risks
The Commission proposes to reduce the number of age categories for the prospective calculation of dose for practices from six to three, while prospective calculations for emergency planning and all kind of retrospective dose assessments should still be based on six age categories (plus embryo and foetus).
One of the basic principles of our radiation protection system is that the additional health risk of an individual receiving a dose equal to the constraint or limit shall be almost identical for each member of the society, especially for each age group. It was this principle which motivated the ICRP nearly 20 years ago to develop age dependent dose coefficients for 6 age groups in order to implement the scientific evidence of age-dependent radiation risks (the alternative would have been to introduce age-dependent values of limits or constraints).
The proposed reduction of number of age groups considered for prospective dose assessments and the introduction of a new set of averaged dose coefficients for three age groups represents an implicit change of philosophy that each individual should be protected at the same level of risk. Although we are aware of practical limitations, we do not see any rationale to modify the principle. On the contrary, the move (at least in many western societies) from utilitarian approaches, including risk perception, to a focus on the individual strongly favours the philosophy that regulations should reflect identical risks to all individuals.
If, however, the Commission has strong arguments to change its traditional principle of constraints representing identical risks to all individuals, this should be stated and outlined in the new General Recommendations. Such a change should not be introduced in this mainly technical foundation document.
For practical implementation in national legislation not much is gained with this reduction of number of age groups as national procedures will still have to foresee dose assessments based on six age categories (plus embryo and foetus) for prospective calculations for emergency planning and all kind of retrospective dose assessments. In addition, countries of the European Union recently finalised the implementation of the EC directive 96/29/Euratom which stipulates the use of six age groups for dose assessment.

3. Probabilistic dose assessments
3.1 Uncertainty vs. variability
The draft is not always clear about the use of the expressions “uncertainty in data” and “(natural) variability or distribution of data”.
In model assessments and interpretation of assessment results, ‘uncertainty’ (e.g. the measurement uncertainty in environmental data) and ‘variability’ (e.g. variability in individual habit data) are different concepts. We feel that the document not always distinguishes clearly between the two concepts (e.g. in section 2.4 where ‘uncertainty’ is often used instead of ‘variability’). A clear distinction between, and a consistent use of the two concepts throughout the draft text would enhance understandability and readability of the document.
3.2 Deriving dose limits and dose constraints
At the end of paragraphs (S7), (45) the Task Group states: “The Commission recognises uncertainties in the models linking detriment to dose. These uncertainties are considered in establishing selected values of quantities such as limits and constraints.”
This statement is difficult to interpret, even more in the context of probabilistic dose assessments. It would be helpful to outline which uncertainties (and/or variability in data and/or probability distributions) have been taken into account in deriving the various limits and constraints set for protection of the individual and how this was done (e.g. by safety factors, expert judgement, probabilistic assessments?).
In our opinion, there is a close relationship between the definition and/or derivation of a dose constraint or a dose limit and the corresponding compliance criteria. Therefore, the question arises whether these newly introduced compliance criteria have an effect on the definition of the dose constraint, and/or on how numerical values for the dose constraint are derived?
3.3 Interpretation of results
Paragraphs (57), (S15), Appendix B, (52)
As discussed by the Task Group, probabilistic assessments generally will include a mixture of probability distributions and point estimates. The following cases may arise depending on the availability of appropriate probability distributions for the data included:
(a) probability distributions of concentrations of radionuclides in the environment, point estimates of habit data, point estimates of dosimetric quantities;
(b) probability distributions of concentrations of radionuclides in the environment, probability distributions of habit data, point estimates of dosimetric quantities;
(c) probability distributions of concentrations of radionuclides in the environment, probability distributions of habit data, probability distributions of dosimetric quantities.
Obviously only the full probabilistic procedure (c) simulates the distribution of doses within a population and allows to make decisions such as those discussed in (S16) and (88) .
The partly probabilistic procedures could be characterised as “dose estimation based on the probability distribution of environmental concentrations of radionuclides” for case (a) and “dose estimation based on the probability distribution of human intakes of radionuclides” for case (b).
At present, dose estimates using the full probabilistic procedure as described in (c) can not be performed, since dosimetric quantities exist only as point estimates. This should be clearly stated by the Task Group. As a consequence, the use of phrases such as “probabilistic dose estimates”, which are commonly used in the document, should be avoided.
In addition, we would welcome information on the Commission’s intention to support the development of probabilistic assessments of dosimetric quantities.

4. Identification of the representative individual
4.1 Representative individual vs. critical group
With this draft publication, the Commission reconsiders the concept of a critical group for assessing the dose for the purpose of radiological protection of the public by introducing the new concept of a representative individual.
In paragraph (53), the representative individual is defined as “...the hypothetical individual receiving a dose that is representative of the most exposed individuals in the population”. The last two sentences of paragraph (S14) state “Ultimately, a group is identified that is expected to receive the highest exposure. The average characteristics of this group are used to estimate dose to the representative individual.”
Both definitions of the representative individual appear similar to the definition of a critical group. It would help the reader to understand the new concept of the representative individual if the Commission could describe more explicitly the difference between the “representative individual” and the “critical group”.

4.2 Guidance on the procedure to identify the representative individual
For the assessment of dose to the public, the Commission proposes that characteristics of the representative individual, pathways of exposure, data of environmental contamination and time frames can be determined either in a deterministic approach or a probabilistic approach or using a mixture of these two. The Commission then provides some guidance on how to choose these parameters, however, there is little guidance on the procedure to identify the representative individual, especially for the prospective probabilistic assessment of dose.
In this context, the second sentence in paragraphs (S16) and (88) is, at least, misleading. The phrasing “...the Commission recommends that the representative individual be identified such...” could be interpreted as a definition of, or a guide to identify the representative individual rather than a compliance criteria for probabilistic dose assessments. Even if the intention was to offer a criterion to prove compliance with the dose constraint, the sentence is very difficult to understand and would profit from rephrasing.
It would be very helpful if the Commission could consider offering procedures to identify the representative individual, especially for the prospective probabilistic dose assessment.

5. Optimisation and limitation
Another basic principle of radiological protection is optimisation of protection. In the draft document on the representative individual, the Commission specifies in paragraphs (S20), (83) and (91) that “...compliance with quantitative dose constraints and optimisation of protection is necessary for radiological protection”.
In case of a probabilistic dose assessment, proving compliance with dose constraints and/or dose limits requires a somehow complicated procedure which is not easy to understand. The demonstration that the protection achieved is in compliance with the dose constraint and optimised requires further steps which are not given in this draft.

Federal Office for Radiation Protection
Contact person: Dr. Gerald Kirchner