Ionizing radiation is a known and well-quantified human cancer risk factor; however, estimates of radiation-related cancer risk are uncertain. Sources of uncertainty include imprecision of measurements used to reconstruct radiation doses, lack of knowledge about true values of dosimetric parameters, assumptions in dosimetry models used to calculate radiation doses, as well as the inherent statistical variations in fitting dose-response models. It is important that uncertainties be incorporated properly into risk calculations and be communicated clearly.
The Radiation Epidemiology Branch (REB) is now conducting several dosimetric studies and assessments of radiation risk in which the evaluation of uncertainties plays a central role and that use the most advanced analytical techniques. The most contemporary design for dose reconstructions account for shared and unshared error structures and is most suitably implemented by Monte Carlo modeling, typically with two dimensions of sampling of uncertain parameters. Uncertainties in both physical dosimetry and personal behavior data are being integrated in a two-dimensional Monte Carlo simulation to stochastically generate multiple sets of individual dose estimates.
Studies presently underway in REB using these techniques include:Steven Simon.