Effective dose: how should it be applied to medical exposures?

The effective dose (E) was created to provide a dose quantity that was related to the probability of health detriment due to stochastic effects from exposure to low doses of ionizing radiation. E is derived from the weighted sum of doses to tissues that are known to be sensitive to radiation and so can only be derived by calculation. The tissue weighting factors are derived from the extrapolation of epidemiological evidence. E was intended for use in radiation protection, but has found wide application in evaluation of doses for medical exposures involving only parts of the body. More reliance is often placed on E values and risk estimates based on E than the evidence on which it is based can justify. In this paper, the uncertainties in the estimated values of E for a reference patient and the associated risk coefficients are reviewed in order to provide an indication of how much reliance can be placed on E as an indicator of risk for patients. The relative uncertainty in estimated values of E for medical exposures for a reference patient is seen to be about +/-40%. The estimated risk of cancer may be a factor of three higher or lower when applied to a reference patient, and will be more variable when applied to an individual. A set of recommendations relating to the use of E and description of risk for medical exposures is proposed.