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


View Comment

Submitted by Ian Fairlie, Consultant on Radiation in the Environment
   Commenting as an individual
Document Dosimetric quantities


This note discusses the ICRP Foundation Document on dosimetric quantities and makes four main points:

(a) the Foundation Document fails to discuss RBE values and wR values of low-range beta emitters including tritium.
(b) the Foundation Document continues to recommend the current wR of 1 for tritium. The ICRP’s stated reasons for this policy do not withstand scrutiny
(c) one result is the current wR of 1 for tritium continues to be applied in inappropriate circumstances despite ICRP admonitions to the contrary.
(d) the ICRP preferred RBE reference radiation is presently hard gamma radiation alone: the Foundation Document’s proposal to change this to BOTH hard gammas and X-rays is unsound and should be rejected.

(a) Failure to discuss low-range beta emitters including tritium.

1. The ICRP Foundation Document on dosimetric quantities proposes to continue the ICRP policy that all electrons and photons should have a wR of 1: this includes low-range beta emitters and Auger emitters. This proposal is misguided and should be revised as much evidence suggests higher RBE values for low-range emitters. Indeed, ICRP Foundation Document and ICRP Publication 92 cite evidence of higher RBE values for lower energy radiations and acknowledge substantial RBE differences between various lower LET radiations. Nevertheless ICRP 92 recommends a single wR value of 1 for all photons / electrons and the ICRP Foundation Document proposes to continue this policy. This is an important matter as recognition of tritium’s higher RBEs would double tritium’s dose coefficients.

2. The Foundation Document is merely the latest ICRP report which refrains from examining research findings on tritium’s RBE, although it clearly should do so. This is thought to be partly due to the long-running controversy on tritium RBE and Q values. A good description of this controversy is contained in CERRIE Paper 9-1 which may be found at document 97 at the CERRIE website It appears that the controversy continues to this day.

3. There are few RBE studies of low-range beta emitters with the exception of tritium. However there are very many RBE studies on tritium, indeed it is among the most studied of all nuclides. This evidence indicates that RBE values for tritium range from 1.3 to 3.4 rather than the ICRP recommended wR value of 1. See CERRIE Paper 9 -1 and figure 1 below. This evidence comes from over 40 animal and cell studies (Straume, 1991; Straume and Carsten, 1993).

4. It is also supported by microdosimetric considerations. In the mid-1980s, a joint ICRU and ICRP report - ICRU Report 40 - stated that RBE values of photons /electrons were inversely energy dependent. That is, as photon/electron energies decreased, their RBEs increased. For photons/electrons with energies below about 10 keV, ICRU 40 stated that differences in RBE could amount to 2 at low doses and it therefore recommended a Q value of 2 for low-energy beta particles including tritium (ICRU, 1986). These recommendations were not implemented by the ICRP.

5. Other authors have remarked on the ICRP’s diffidence re the increased effectiveness of low-range emitters. Hofer (1998), for example, has stated that ICRP models and their data base for dosimetry calculations were “far from perfect” in the case of low-energy emitters.

(b) the Foundation Document continues to recommend the current wR of 1 for tritium. The ICRP’s defences for this policy do not withstand scrutiny

6. The strong scientific evidence of higher RBE values for tritium than unity and the ICRP’s continued wR recommendation of 1 has placed the ICRP in a scientific quandary. It has tried to wriggle out of this by giving a variety of reasons for its policy; however none of these withstand scientific scrutiny.

7. First, ICRP has defended its recommendation of a wR of 1 for all photons and electrons by stating that there is little epidemiological evidence on which to base any differences (preface Editorial, ICRP 92). It also stated that uncertainties in risks were large:

“… given the UNSCEAR 2000 judgment of several-fold uncertainty on the judgment of the nominal risk coefficient for cancer, including that for the dose and dose-rate effectiveness factor (DDREF), we do not see the need to ascribe different values of wR to different low-linear-energy-transfer (LET) radiations. A wR of 1 may therefore be retained for all low-LET radiations.” (preface Editorial)

8. However these points apply with greater force to neutrons and protons, but this did not stop ICRP recommendations being made for different wR values in their cases. In fact, the report recommended differences in wR values for neutrons (which vary by about a factor of 4 depending on energy), and for protons (whose wR is reduced from 5 to 2, ie a factor of 2.5). However the ICRP’s findings (see paragraph 53 of ICRP 92) of RBE differences of 2 to 3 between conventional x-rays and hard gamma rays were not accommodated. Similarly, RBE differences between beta particles and gamma rays also were not discussed.

9. The ICRP has also defended its decision to continue with a wR of 1 for tritium by referring to RBE values of 1.2 in two animal studies (Gragtmans et al, 1984 and Johnson et al, 1995) on cancer induction in mice and rats. Although animal data on carcinogenicity should be considered, the ICRP reliance on them alone, indicates a rather selective view of the large body of RBE evidence available on tritium as set out in CERRIE Paper 9-1.

10. More important, these two animal studies used 200 kVp X-rays as their reference radiation. However hard gamma radiation would have been the more relevant reference radiation to have been used, for two reasons. First, the Compton/photoelectron ratio produced by gamma rays is more similar to that produced by beta particles. Second, the ICRP 92 recommended that hard gammas are preferable as the reference radiation for a number of reasons (see paragraph 15 below). As stated by ICRP 92, X-rays are approximately twice as effective as hard gamma rays, which means that the correct RBE would be about double the observed value of 1.2, thus fitting in with most other RBE values observed with hard gammas as the reference radiation.

(c) ICRP’s failure to discuss tritium’s RBE values means the wR value of 1 is applied inappropriately

11. ICRP 92 also stated that wR and RBE values may be different as they are used for different purposes:
“…there is no conflict between the fact that all photon radiations are …given the same weight, while risks .. of soft x-rays, conventional x-rays, and hard gamma rays are assessed differently. wR is designed for the practice of radiological protection, not for specific risk assessment. … it is not meant to be applied in the derivation of quantitative risk estimates under specific conditions” (paragraphs 82 and 83).

It further explained that
“…wR is a quantity intended for use in radiological protection and was not developed for use in epidemiological studies or other specific investigations of human exposure. For these other studies, absorbed dose in the organs of interest and specific data relating to the RBE of the radiation type in question are the most relevant quantities to use.” (preface Editorial)

These views are maintained in the ICRP Foundation Document on dosimetric quantities.

12. These statements are disingenuous, because in practice RP practitioners simply do not distinguish between RBE and wR. When RP practitioners and regulators make quantitative risk estimates under specific conditions, they almost invariably use ICRP wR values. The reason is that ICRP guidance on recommended RBE values in this area does not exist (except for Auger emitters). Given this absence, wR values and effective doses will continue to be used routinely in retrospective dose assessments, epidemiological studies or medical exposures, regardless of the ICRP’s advice to the contrary.

13. If the ICRP wishes wR values not to be used in the above circumstances, then it needs to recommend RBE values for use in such situations, ie for specific nuclides, ideally in their various chemical forms (ie HTO and OBT) and for specific organs or tissues. Without such guidance, a danger would exist of different regulatory bodies using differing RBE values for the same nuclides in specific dose assessments etc – which would be a recipe for confusion.

14. One result is that tritium’s wR value of 1 continues to be applied inappropriately in epidemiological studies and other specific investigations of human exposure. This should be stopped.

(d) proposed change in preferred reference radiation

15. Relative biological effectiveness (RBE) is defined as the ratio of the absorbed dose of a reference radiation to that of a test radiation to produce the same level of biological effect observed in experiments. Two reference radiations are in common use: hard gamma rays (usually from Co-60) and X-rays (usually 200-250 kVp).

16. Paragraph 53 of ICRP 92 (2004) noted RBE differences of 2 to 3 between conventional x-rays and hard gamma rays with the former being more effective. As a result, the two reference radiations have been the source of much confusion in radiation biology in general. In particular, they have bedeviled the determination of RBEs for low-range beta emitters because 2 and 3 are their RBE values when compared with these radiations.

17. As a result, ICRP 92 (2004) advised that hard gamma rays were the preferable reference radiation. It stated unequivocally (in paragraph 28)
“While there is no need for an exclusive convention, it is nevertheless convenient to adopt a reference radiation that is understood to apply whenever there is no explicit statement to the contrary. There are practical arguments to favour gamma rays for this purpose. It is difficult and expensive to determine the initial slope of dose responses of the induction of cancer in animals, especially with low-dose-rate x-rays rather than low-dose-rate gamma rays. For this and a number of other reasons, hard gamma rays are preferable as the reference radiation because:

• most experimental animal studies of cancer induction and life shortening (and deterministic effects) have been carried out with gamma rays, and, importantly, some with exposures at low dose rates;
• the most important body of data for estimating radiogenic cancers in humans are from the atomic bomb survivors who were exposed to hard gamma rays;
• hard gamma rays have the lowest LET (dose average LET, LD, 0.4 keV/mm or less) among photon radiations;
• the distribution of the deposition of energy from gamma rays in large fields is more uniform than with x rays."

18. Perplexingly, the 2005 ICRP Foundation Document on dosimetric quantities now proposes that BOTH hard gamma and X-rays be used as reference radiations - the opposite of what the ICRP 92 said a mere year ago. The ICRP Foundation Document is proposing to return to the previous unfortunate situation where both gammas and X-rays were confusingly used. This flies in the face of the evidence cited by ICRP 92 that X-rays are approximately twice as effective as hard gamma rays. This proposal is simply not supported by the available evidence and should be withdrawn.

19. The ICRP Foundation Document attempts to justify its unfortunate proposal by stating:
“One argument for this approach is the generation of secondary radiation in the human body. If a body is exposed to mono-energetic gamma radiation, the resulting photon radiation field inside the body comprises not only the incident radiation, but also a large fraction of scattered photons with much lower energies resulting from single and multiple Compton scattering (Harder 2004). This situation is quite different from that in investigations on small micro-organisms or single cells, where the scattered photon contribution to the total dose is negligible. It is, therefore, justified for the selection of radiation weighting factors with respect to the human body to use an average of experimental RBE data related to radiation fields which use as a reference, either high energy (> 200 keV) x-rays or 60Co-gamma radiation. As a consequence ICRP has used a broad range of RBEM values related to different photon fields (mostly 60Co ã-rays or high energy x-rays) for the evaluation of wR values for other radiations. Although this approach has resulted in some debate it has allowed most of the available experimental RBE data to be taken into account in the selection of wR values.”

20. This argument is unconvincing. It only applies to mono-energetic gamma radiation, and does not discuss low-range beta emitters, Auger emitters, alpha emitters, or even low-energy X-rays whose energy distributions are later acknowledged to be likely to be heterogeneous. But it is precisely these radiations on which more detailed discussion and advice on wR and RBE is required.

21. The argument is unsatisfactory for other reasons. Depending on the irradiation configuration, of course there will be varying spectra of primary and secondary scattered radiation in tissue. But there seems little logic in saying that therefore both gammas and X-rays can both be defined as optimum reference radiations. Of course, human bodies will have more internally generated secondary photons than mice or cells. But even in experiments, it is usual to add material on the incident side of the sample to ensure charged particle equilibrium before it; often backscatter material is added behind the sample. In some experiments, samples are placed in a large phantom (water or Perspex) to simulate being within the body. Clearly many variations exist which can make a substantial difference to biological effectiveness (~ factor 2). Therefore it is unconvincing to say both these radiations should therefore be treated as equally good reference radiations.

22. The Foundation Document’s reversal of view barely a year after ICRP 92 is disturbing. Its refusal to discuss the reasonable and logical arguments of ICRP 92 on the matter is also disturbing. Those with a slightly cynical point of view may be forgiven for concluding that the real reason for this proposal may have more to do with “fitting the science around the policy” of keeping the wR of tritium at 1, than the above rather specious reasoning. The point here is that using hard gammas as the preferred reference radiation results in the RBE for tritium shifting to >2 rather than the ICRP’s existing wR of 1. This is revealed in figure 1 below which indicates a range of 1.3 to 3.4 and a median of 2.2 for tritium’s RBE values compared with hard gamma radiation. The full list of RBE experiments (using both X-rays and gammas) may be found in CERRIE Paper 9-1 at

Figure 1 [NOT VISIBLE HERE]. Distribution of RBE values for HTO/gamma comparisons

23. The question arises here as to why the ICRP appears to persist with its strange attitudes on tritium. Why, for example, does it refrain from discussing the evidence of higher RBEs for tritium? The experience of Harrison et al (2002) in this regard is illuminating if discomforting. Harrison et al reviewed the data on which current ICRP dose coefficients for HTO were based and estimated that the RBE for tritium (HTO) was uniformly distributed between 1 and 2.5 with a median of 1.75. The authors in 2002 stated that an ICRP Task Force would be considering recommendations to double tritium’s dose. These recommendations were never implemented: the wR for tritium (HTO) remains at unity as of July 2005 and tritium’s dose coefficients have not been increased. Nevertheless, one author, Lambert (2001) has used increased dose coefficients for tritium in specific dose studies.

24. Indeed the question may be asked why the ICRP refuses to discuss tritium at all. For example, the ICRP’s disinclination to discuss tritium contrasts with its welcome discussion on Auger emitters whose radiations exhibit similar properties to that of tritium. Could this be because tritium discharges loom large in military and civil nuclear (including possible future fusion) installations? If so, this would be a regrettable intrusion of political bias into what should be a strictly scientific consideration.


25. The Foundation Document’s policy not to recommend realistic wR values for low-range photons and electrons is unsatisfactory for a number of reasons as discussed above.

26. The ICRP’s disinclination to allocate higher wR factors to low-range photons and electrons is not precautionary: that is, it does not err on the safe side for radiation protection purposes. It may have adverse implications, as regards radiation protection, because low-energy beta emitters are commonly discharged from nuclear and other industrial facilities. It may also have adverse consequences for radiation protection in nuclear medicine, as regards widely-used Auger emitters, and for mammography x-rays.

27. The following recommendations are made

• wR values greater than 1 rising to ~3 should be assigned to low-energy (< 50 keV) electrons via a continuous RBE-energy function
• one result would be for tritium’s wR to be increased to >2
• RBE advice for all nuclides (in specific chemical forms) should be drawn up, published and used in specific risk estimations, epidemiology studies and retrospective dose assessments in preference to generic wR values
• in the interim, immediate advice should be issued to cease using the inappropriate RBE value of 1 for tritium in epidemiological studies or other specific investigations of human exposure.

CERRIE (2004) Report of the Committee Examining the Radiation Risks of Internal Emitters.
Gragtmans NJ et al (1984) Occurrence of Mammary Tumours in Rats after Exposure to Tritium Beta Rays and 200 kVp X-rays. Radiat Res 99, 636-650.
Harrison J, Lambert B, Khursheed A (2002) Uncertainties in the Dose Coefficients for Intakes of HTO and OBT by Members of the Public. Radiat Prot Dosim Vol 98, No 3, 299-311.
Hofer KG (1998) Dosimetry and Biological Effects of Incorporated Auger Emitters. Radiat Prot Dosim Vol 79 Nos 1- 4 pp 405-410.
ICRP (2004) Relative Biological Effectiveness (RBE), Quality Factor (Q), and Radiation Weighting Factor (wR).
ICRP Publication 92. Annals of the ICRP 33 No, 4. Pergamon Press, Oxford.
ICRU (1986) The Quality Factor in Radiation Protection. Report of a Joint ICRU and ICRP Task Group. ICRU Report 40. International Commission on Radiation Units and Measurements, Bethesda, MD, US.
Johnson JR, Myers DK, Jackson JS, Dunford DW, Gragtmans NJ, Wyatt HM, Jones AR, and Percy DH (1995) Relative biological effectiveness of tritium for induction of myeloid leukaemia. Radiat. Res. 144, 82-89.
Lambert BE (2001) Journal of Radiol Prot Vol 21. pp 333-335.
Straume T (1991) Health Risks from Exposure to Tritium. Lawrence Livermore National Laboratory Report, UCRL-LR- 105088.
Straume T and Carsten AL (1993) Tritium radiobiology and relative biological effectiveness. Health Phys. 65, 657-672.