Radiology 2012; 264: 325-328
James H. Thrall, MD Computed tomography (CT) is by far the most important source of medical radiation exposure in the United States (1), with more than 70 million examinations performed each year. The large number of patients receiving radiation exposure from CT has raised concerns about patient safety, most importantly, the risk of radiation-induced cancers and cancer deaths later in life. In trying to estimate cancer risks for patients undergoing CT, two issues come to light. First, the effective doses of radiation associated with most CT examinations are very low, on the order of 1.0–12.0 mSv (1). Second, statistical uncertainties in available data, such as data from atomic bomb survivors in Japan and radiation workers in the United Kingdom, make direct estimates of cancer risk highly problematic for effective doses of less than 100 mSv. For this reason, investigators seeking to calculate risks for cancer induction following CT have typically used an extrapolation model called the “linear no-threshold” model. This model is endorsed by the prestigious Committee to Assess Health Risks from Exposure to Low Levels of Ionizing Radiation of the National Academies of the United States (Biologic Effects of Ionizing Radiation [BEIR] VII phase 2) (2–5).