Interpretation of biossay data

Draft document: Interpretation of biossay data
Submitted by Augusto Giussani, GSF
Commenting as an individual

GENERAL COMMENTS The current version of the guidelines contain several misprints, some of them are listed below. My main concern with the current document, however, is with its structure: I expect a guideline document to be a clear and hopefully slim handbook which easily explains how to perform a given task (in this case, the interpretation of bioassay data). This is only partially accomplished in Chapters 7 and 8, whereas the previous parts, expecially Chapter 4 and 5, look like a summary of generic information, and in many parts not always well structured. My personal suggestion would be to use following structure, 1) short introduction describing the aim of the guidance document and the criteria used for the definition of the different levels of task, i.e. that the effort should be proportional to the anticipated dose. 2) main body of the document, containing: (i) the general philosophy of radiation protection (Section 2 of the current document); (ii) a short overview of monitoring programme and techniques (summary of the current sections 3 to 6); (iv) a step-by-step description of the different stages, with detailed explanations of the terms and notations the first time they appear; (iv) a series of practical examples to show the application of the guidelines. 3) a number of annexes, where the most expert and interested users can find greater details on the general topics. With regard to point 1 and to the philosophy of the proportional approach, I find the distinction in Levels given in Section 8.1.4 definitively much clearer and much more effective on a practical point of view than the one reported in Section 7.2 (the latter has a somehow confusing and complex formulation, that may give the possibility of several personal interpretations). If however the intention is to avoid a rigid scheme based only on the values of anticipated dose, following formulation could be used: Level 0: no need of dose evaluation when the effective annual dose would be most likely below 0.1 mSv, even if there should be similar intakes in each monitoring interval of the year. This condition is fulfilled when the measured quantity M is lower than a critical value Mc. Tables of the critical values Mc are given for each radionuclide and for each monitoring scheme (Stage 1 of the flow chart). Level 1: dose evaluation using the default parameters of ICRP when the measured quantity in a routine monitoring exceeds the critical value Mc and the evidence of a new intake has been confirmed (Stage 2 and 3 of the flow chart). Level 2: sophisticated dose evaluation using additional information from the workplace; usually this is done after accidental intakes and for routine measurements when the evaluation in Level 1 indicated a committed dose greater than 1 mSv. (Stage 4 of the flow chart) Level 3: Expert evaluation, usually applied when the evaluation in level 2 indicated a committed dose greater than 6 mSv. This requires the availability of comprehensive data and specific parameters relating to the subject. Another concern with regard to the draft document is there are contradictory information on how to interpret data from routine monitoring. So section 5.3, Page 42 states: "The intake is often taken to occur at the mid-point of the sampling. The approach was used in ICRP Publication 78 and for consistency is also adopted in the OIR series of publications." The conclusions of ANNEX F says "The Mid-point method is adequate (…) is the default method suggested in the main text of this document. However (…) if possible the constant chronic intake should be used." Section 6.3, point (vi) says: "For each monitoring interval, the predicted value of the measured quantity is the predicted value at the end of the monitoring interval assuming a constant chronic intake….". The formula (7.1) for the critical monitoring quantity Mc is expressed in form of m(T/2), but in the following discussion on the relation between Mc and z (dose per unit content) it seems that m(chronic) should be used. Moreover, in the Example OIR-27 presented at the end of the document, the values of "m" given in the table for routine monitoring (Table 27.10) appear to have been calculated under the assumption of chronic intake, and those for "z" (Table 27.11) under the "mid-point" assumption. This is not only contradictory, but also confusing for the user. If the final philosophy of the document is to have "mid-point" as the default approach and "constant chronic" as a more refined alternative, my suggestion is to clearly separate the two approaches: the "mid-point" approach should be extensively described in the main text, together with all the related quantities, whereas the alternative approach may be introduced in an annex. As a consequence, the OIR Publications should have for each element a dedicated section for routine monitoring, with a subsection introducing tables of the quantities Mc, m and z for the "mid-point" approach, and a subsection introducing tables of the quantities Mc, m and z for the "constant chronic" approach. SPECIFIC COMMENTS Data presentation. I strongly support the idea to provide tables with "dose per unit content" (z) in addition to the prediction values m ("content per unit intake"), and I would suggest to present this concept as the one to be preferred (it is actually the most straightforward). Looking at the example OIR-27, I suggest that the tables should clearly state if the quantity reported is "m" or "z", using the same notations as given in the main text and indicating the corresponding units. SECTION 2.5 and 2.6: the sequential presentation of the information given in these two sections is not optimal, my suggestion is to merge the two sections as follows: 1. The first paragraph of 2.5 ("The main use (…) purposes worldwide.") should be followed by the first paragraphs of 2.6 ("In occupational exposure (…) material taken into the body."). Then the other paragraphs of Section 2.5, starting with "The calculation of reference dose coefficients...". The current formula of E(50) explicitly includes only inhalation and ingestion, whereas intake through the skin or wounds are considered as a "special extension"; it would be probably better to use a more general expression such as E(50) = SUMj (SUMk (ejk(50)*Ijk)) where k are the possible intake modalities and j are all involved radionuclides. I also wonder whether such a deep discussion about effective dose and tissue reactions is necessary (especially with reference to epidemiological studies). SECTION 3: my suggestion is to shift this chapter at this end as an Annex. Average users will merely apply the coefficients and curves given in the OIR Publications, without having to work directly on the model structures. Only expert users will be required to do so, and in this case it might be more convenient to redirect them to the original publications or guidance documents where the applications of the models were described (this would also avoid to duplicate sources of reference information). Here anyway some corrections to the current text: Figure 3.1, 3.3 and 3.4: Gastro-intestinal tract should be substituted with Alimentary Tract. For the sake of clarity, it might also be useful to indicate explicitly in the illustrations that the material goes from ET2 into the compartment Oesophagus/slow (as already shown in Figure 3.6). Section 3.2.1: fractional depositions are given in Table 3.2 (and not in Table 3.1, as written in the text). Section 3.3: the sentence "Default assumption that total fractional absorption fA of an element and its radioisotopes to blood occurs in the small intestine, ie fSI = fA." is for me syntactically incorrect, as it makes confusion between the PROCESS of absorption (which is what occurs in the intestine) and the VALUE of the absorbed fraction fSI and/or fA (which are numerical values). I would change it into "Default assumption that absorption of an element and its radioisotopes to blood occurs exclusively in the small intestine, ie the total fractional absorption (fA) coincides with the fractional absorption from the small intestine (fSI)". Section 3.9: The NCRP commentary 15 (Evaluating the Reliability of Biokinetic and Dosimetric Models and Parameters Used to Assess Individual Doses for Risk Assessment Purposes, 1998) is quoted as Tries, 2000. SECTION 4: This section (and, to a lesser extent, also section 5) looks like a partial and not well structured presentation of current techniques, without a real guidance spirit (for example, I don't understand the reason why a very specialized and not easily accessible technique such as AMS for 14C is explicitly quoted). Moreover, some specific technical information (such as detection limits) might be superseded in the very near future, due to the rapid development that some techniques are experiencing, and this would depress the value of the document (it could soon sound "obsolete"). I would find it more effective if Section 4 and 5 were merged together into a simple section with nothing more than a listing of techniques and monitoring strategies available, whereas more detailed information could be given for each radionuclide in the OIR series of publications, where specific technical aspects could be better presented and discussed. SECTION 5: Figure 5.1 and 5.2 are missing. Section 5.8 should/could be shifted before 5.6 (or even before 5.5). Section 5.6: Should a relation (if any) between investigation levels and critical monitoring quantities Mc be given? SECTION 6.3, (iv) and (v): it should be clearly stated that the quantities given in the graphs are 1/z(t) and m(t) respectively. Additionally, it seems to me that the graphs described in point (iv), "predicted activity (…) for an acute intake of the radionuclide which corresponds to a committed effective dose of 1 mSv" should enable to obtain directly the dose, without the need of an intermediate step to assess the initial intake. SECTION 7 and 8: In my opinion it would be better to present first the task and the structure of the flow charts, and then explain in detail the practical aspects of each single step. ANNEX B, Example 1. the predicted measured value for unit intake is here indicated with f(t) instead of m(t) as in the main body of the document. For the sake of clarity, it would be useful to have a table showing for each measurement the corresponding m(t) and the derived estimate of I. In the formula for ln(I), in the middle of Page 6, a parenthesis is missing in the second term at numerator (the formula is correct also without parenthesis, however there is no "symmetry" between the two terms of the summation). ANNEX B, Example 2. It is not very clear what is the relation between this example and the guidelines. Here a method for interpretation of measurements of HTO is given. Two approaches are presented (2.2.1, simple hand calculation, and 2.2.2, assessment according to the guidelines (?)), and in both of them the assumption is made that urine concentration of H-3 is proportional to body burden. The only difference is that in the first case the area under the curve AUC is calculated in a very rough and approximate way, whereas in the second case a more appropriate multi-exponential fit procedure is performed. In both cases there is no connection between the procedure described here and the indication given in the guidelines, in particular in chapters 7 and 8. So I would use this example only in the OIR Publication on H-3 (since this application is exclusive for HTO), with following changes: the description of the ICRP assumption of equilibrium concentration between urine and whole body concentration and on the rationale for using the AUC for calculating the number of decays in the whole body should be shifted at the beginning (these assumptions are made for both methods of calculation). Moreover, the first method suggested is very rough, I wonder whether it is still convenient to suggest it, as there are a number of software and resources that enable to perform very easily fits with multi-exponential functions. Finally, the activity concentration in urine is indicated as e(t) in 2.2.1, this could create some confusion with the effective dose coefficient e(50). ANNEX F: The discussion about the unbiasedness of the Constant Chronic Intake could be improved, as the part comparing the realism of the two assumptions (Constant Chronic and Mean-Inverse) is not straightforwardly comprehensible.