Occupational Intakes of Radionuclides Part 1

Thank you for the opportunity to comment on this draft document.  While the bulk of this work is welcomed and likely to prove beneficial we do have one major concern and also several minor suggestions.

The principal concern relates to the practical implications for the use of “reference biokinetic models” and how this is incorporated into the definition of “effective dose” and the use of the concept of “dose of record”.  It is recognised that these concepts were discussed in ICRP publication 103; however, they were not mentioned within ICRP publication 60, which is the basis of the currently used system of dosimetry.  This perceived change of emphasis or tone could have potentially significant, confusing and detrimental consequences for operational dosimetry.

In particular we are concerned about the situation when monitoring data obviously conflicts with the expectations from reference models.  Some of these concerns are illustrated below:  

• It is presumed that the forthcoming reference biokinetic models will have been influenced by observations of bioassay measurements from test cases and studies, perhaps involving non-human species; if so, then it would be incongruous to explicitly prohibit observations of bioassay measurements from the individual actually being investigated.

• It would be difficult to justify recording an estimate of ‘effective dose’ to a record ascribed to an individual which is clearly at variance with measurement data from that individual.

• The “dose of record” is essentially a record of exposure conditions; in real situations exposure conditions are often very poorly understood and the practices of characterising and recording them are highly variable, often influenced by local “custom and practice” methodology.  The OIR goes into great detail on reference models (which is welcome and of interest), but very little on the assessment of exposure conditions.  The current practice of basing dosimetry programmes on real monitoring data from real workers helps as a counterbalance to limit the impact of the lack of knowledge of real exposure conditions (experience and studies have shown that the practical use of the quantity “effective dose” – when estimated from real data - is often insensitive to uncertainty in exposure).  The concern is that this control will be lost, to the detriment of practical radiation protection programmes.

• If the “dose of record” is limited to source-based determinations of local exposure conditions it is doubtful whether this can be considered to be additive or transferable to records of “dose of record” made from different sources.

• This would be exacerbated if an individual transferred between sites (and respective dosimetry services) with residual excreta or body activity which was not compatible with their “dose of record”: this could be very confusing for the ‘receiving’ service.

• If biokinetic studies for a particular site or operation indicate that a non-reference model is appropriate, then it might be detrimental to safety and cost to persist with the reference model.

• Protection programmes may sometimes need to balance protecting against diverse sources of radiation hazards: e.g. enhanced protection from tritium exposure may be at the price of elevated external exposures.  In such circumstances it is obvious that the most appropriate models should be used.

• Although AMAD and absorption (i.e. lung type) are considered to be source-based parameters, in practice these parameters are usually determined, retrospectively, from bioassay measurements.

• The above examples highlight the possible confusion over what are source-based and what are individual-based parameters, and when and how they should be used.

• While it is clearly recognised that reference dosimetry and anthropomorphic models have to be used for reasons of necessity or practicality, it is not at all clear why the same should apply to biokinetic models, particularly in cases which generate sufficient bioassay measurements which are strongly related to the quantities of interest – this being the time dependent activity levels and concentrations in principal source organs.

• It is noted that biokinetic models, in essence, define the time-periods of internal exposures; it is not clear why “effective dose” should have pre-defined exposure times for internal exposures, but not for external  exposures.

• We currently implement a dosimetry system based on ICRP60 and ICRP23 (Reference Man), and record results as “effective dose”.  When we move towards a system using ICRP103 and ICRP89 (Reference Worker) then the reference anthropomorphic model has clearly changed; the OIRs would then imply that the currently defined quantity “effective dose” (using ICRP60/23 dosimetry) is different from the prospective defined quantity “effective dose” (using ICRP103/89 dosimetry): this doesn’t make sense.

• It is presumed that ICRP will periodically refine and update reference biokinetic models; the OIRs imply that following such updates the defined quantity “effective dose” could no longer be used because the reference models have changed.

(You will appreciate that we are suffering considerable confusion as to the reasons, purpose and value of these proposals, which will inhibit their practical implementation.)

The current practice is to refer to ICRP dosimetry-models and anthropomorphic-models as fixed references with no associated uncertainty; latest published ICRP biokinetic models are used as default references; these are only departed from if there is evidence to indicate they are invalid for a specific investigation and more appropriate alternatives are available (either from published sources or by local investigation).  The use of non-ICRP models are subject to standard scientific scrutiny and peer-review processes.  The results of this process are recorded as “effective dose”.  It is not clear how the latest proposals would improve on this arrangement.  It is also unclear why the quantity “effective dose” now seems to be unusable if non-ICRP Reference Biokinetic models are used.

Other comments:

Para 207: Wounds:  what would be the dose quantity recorded for intakes via wounds? How would this be compared to dose limits and estimates of “effective dose” from other exposures?

Para 229: Medical Intervention:  The same questions as for wounds apply for dose assessments post medical interventions.

Para 245(a) selection of sex-specific biokinetic model: it is not clear why a sex-specific model is recommended to calculate intake, which is then compared to a sex-averaged dose coefficient.

Para 260: include ‘nose blow samples’ as well as ‘nasal smears’.

Paras 263 – 266: air samples: while published studies (and experience) have reported poor correlations between intake estimates based on PAS and those based on bioassay it is noted that these studies have not appeared to have considered the relative uncertainties in any detail.  The estimate of uncertainty in bioassay-based estimates for low or chronic doses is difficult and their significance cannot be discounted.  For this reason it would be an overly strong comment to state that bioassay should be preferred to PAS at low level doses.  It would be more reasonable to state that the relative reliability and accuracy between these techniques for low doses is not well known, and this lack of knowledge should be taken into account when planning monitoring programmes.

Para 313: Detailed evaluation of doses: “At installations where workers have the potential to be exposed to doses higher than 1 mSv… information should be gathered on the physical and chemical characteristics of the inhaled or ingested radionuclide…”  It is noted that most facilities – particularly those handling actinides – will have a potential for doses > 1 mSv.  While it would be nice to have detailed information for all such facilities it is very unlikely to be practicable or cost effective in practice.  This paragraph could be re-drafted to be applicable where doses are likely to be > 1 mSv; or to give greater clarity as to the level of detail required of the materials, and to how this level of detail changes with expected levels of dose.

Para 329: correction for dietary intakes in excreta: we would urge caution in this – excretion rates from dietary contributions are highly variable, subtracting the presumed contribution from bioassay measurements should only be done when there is clear evidence to support it.