Recently, several epidemiological studies were interested in identifying potential health effects of incorporated radionuclides. To achieve this work, exposure of individuals was quantified in order to be compared with health status. In order to quantify the exposure of workers, the measurement of retained or excreted activity also called bioassay, must be interpreted in terms of dose. The data available for dose reconstruction are mostly bioassay analyses (mostly urine) carried out to verify the absence (or presence) of incorporated radionuclides in the workers’ body. Exposure conditions, recorded in a Job-Exposure Matrix (JEM), are known more or less precisely depending on workplace and time of exposure. However, data gathered to document workers’ contaminations were collected for radiological protection purposes rather than for precise retrospective dose assessments. Therefore, a large panel of exposure scenarios could be used to reconstruct lifetime doses. Moreover, a large portion of bioassay data are recorded as below the detection limit (DL) of the measurement technique. That is why, the uncertainty on the lifetime doses is assumed to be important. The same uncertainty is expected on dose estimates for compensation claims since they are based on the same data.
In order to quantify this uncertainty, three cases of uranium exposure, all originating in the French nuclear industry, were recently distributed inside EURADOS Working Group 7 on Internal Dosimetry to a number of participants for the purposes of an intercomparison exercise aiming:
to compare dose assessment protocols of the different participants,
to identify sources of uncertainty, and
to discuss the assessment of uncertainty on dose.
16 participants estimated total committed effective dose, total equivalent doses to the lungs and to the kidneys for at least one of the three workers. Worker 1 presented a large number of bioassay results and several recorded incidents; for Worker 2 only one result out of 19 was higher than the DL and this result was obtained at a time when exposure was not possible according to the JEM; the 75 bioassay results of Worker 3 were all below the detection limit.
The dispersion of the dose assessments is important, higher than the factor of three usually acknowledged for uncertainty of internal doses. From the description provided by the participants, the protocols to evaluate doses were reviewed in details and sources of uncertainty along with reasonable modelling assumptions were identified. This work will be used as a basis for defining guidelines to reconstruct lifetime doses for epidemiological studies and for compensation claims. Finally, the influence on the dose of the different uncertainty sources will be estimated by carrying a sensitivity study comparing dose assessed strictly applying the guidelines with doses calculated under alternative but reasonable modelling assumptions identified in this intercomparison.