Uranium dose assessment: a Bayesian approach to the problem of dietary background.

Workers are routinely monitored by urinalysis for exposure to uranium at Los Alamos National Laboratory. Urine samples are analysed by alpha spectroscopy for 234 U, 235 U and 238 U. Interpretation of the data is complicated by the presence of natural uranium in the workers' drinking water and diet. Earlier methods used drinking water samples to estimate the dietary component in urine excretion. However, there proved to be insufficient correlation between drinking water concentration and excretion rate. Instead, an iterative calculation is used to identify a typical excretion rate for each individual in the absence of occupational intakes. Bayesian dose-assessment methods, first developed for plutonium bioassay at Los Alamos, have been adapted for uranium. These methods, coupled with an algorithm for identifying each individual's baseline level of uranium from environmental sources, yield improved reliability in the identification of occupational intakes.