Reliability of the Icrp’s Dose Coefficients Case Studies of Potential Errors in Current Biokinetic Models and Dose Coefficients

A pilot study was conducted for the U.S. Nuclear Regulatory Agency (NRC) to examine the feasibility of characterizing and improving the reliability of current dose coefficients for a comprehensive set of radionuclides of concern to the NRC. The dose coefficients considered for inhalation and ingestion were those currently recommended by the International Commission on Radiological Protection (ICRP) for members of the public and workers. The pilot study consisted largely of a set of case studies, i.e., detailed assessments of selected radionuclides and exposure situations, representing different sources and quality of biokinetic and dosimetric information. An assessment of a case study included critical evaluation of the database and current models for the radionuclide and comparison of model predictions with those of alternate models considered to be at least as strongly founded. This paper summarizes results and conclusions of eight case studies regarding the reliability of ICRP models and dose coefficients as central estimators for adult males. The focus of this paper is on potential errors in biokinetic models, which were judged to be the dominant sources of error in dose per unit intake in most instances, and the reflection of those errors in derived dose coefficients. It is concluded that potential errors in current models and dose coefficients often arise in large part from discrepancies between the models and database and thus are reducible errors rather than genuine uncertainties. The combination of reducible errors and uncertainties can result in large errors in applications of current dose coefficients, even for frequently studied radionuclides.