P142] Interpretation of radiation risks from low dose diagnostic scans

Abstract

Purpose Scientists disagree on how to interpret risks from low dose radiation exposures. In particular, publications which rely on the Linear No Threshold (LNT) model for estimating of the number of future cancers that may be caused by medical x-rays have led to considerable debate. Some scientific bodies have described such estimates as “highly speculative”. There has also been criticism of a commonly used dose metric, Effective Dose (ED), which it is argued, should be used for “radiation protection purposes” only. Methods This paper offers a partial defence of estimates of cancer risk in a population after exposure to low dose x-rays. We argue that (i) ED is too readily dismissed as a tool for estimating risks in a population, and (ii) that the context for the unmeasurable risk from low dose x-rays is not always given sufficient attention. Results On the first point we argue that criticisms cataloguing the averaging assumptions underpinning the estimation of ED can only be used to calculate an error bar; the conclusion that ED cannot be used in risk estimates because of this error bar does not follow. Secondly, we criticize the use of the term “highly speculative” for risk from exposures below 50 mSv. It is easy to suppose that if science cannot detect a risk, then that risk must be negligible: however, this ignores the fact that because cancer is a common disease, x-rays would have to cause large numbers of cancers before the risk could be detected. In this context, unmeasurable risk cannot be equated to inconsequential risk. Conclusions Confusion about radiation risks has the potential to disrupt the optimum use of medical imaging technology. This paper argues against views which give insufficient weight to low dose radiation risk [1] , [2] .