Commentary. Radiation dose in CT: are we meeting the challenge?

Abstract

CT has developed dramatically: scanning is faster, images are better, applications have grown and, of course, radiation doses have come down, have they not? Well, no, actually; quite the contrary in fact. When CT was new it was appreciated that it was a relatively high dose technique but there was overriding clinical justification for using it; in the brain no technique could approach it and when body CT began it concentrated on patients with malignant disease, where radiation dose was of less concern. Now circumstances are different. The technique is widely used, extensively in benign disease as well as in young patients for whom radiation protection considerations are paramount. Awareness dawned of the magnitude of the challenge following a national dose survey in the UK. In 1989 the National Radiological Protection Board showed that despite comprising only 2% of all examinations, CT contributed around 20% of the collective dose to the population from diagnostic imaging [1]. Subsequent analyses for the UK suggest that this latter figure may have risen to 40% [2]. One department in the USA has claimed that CT now represents 67% of the collective dose that it delivers [3]. Patient doses from CT are among the highest in diagnostic radiology; an abdominal examination in an adult with an effective dose of 10 mSv has been estimated to increase the lifetime risk of fatal cancer by 1 in 2000 [4]. Radiation exposure from CT is rising, not falling. This increase is not simply owing to CT replacing other techniques. We know that large variations in CT practice exist; experience from review clinics suggests volumes of exposure, number of slices and repeat exposures can be widely different within the same clinical application, with little apparent clinical justification. Surveys of practice and dose suggest that (effective) dose for a given examination may vary by a factor of 40 between departments for a UK patient [1], or a factor of 20 in Norway [5]. Overall, the evidence indicates a strong trend of increasing population dose owing to rising use of CT and to increased dose per examination. It now seems clear that variations in practice have become more important than scanner technology in determining the dose to the patient [6]. Ease of use may be contributing to this syndrome. Early CT scanners were rigid tools and extending an examination implied a time penalty. The introduction of spiral CT reduced this dramatically and probably contributed significantly to new variations in practice, as few technique guidelines were available when the new technology was introduced. For example, contrast enhanced studies have become more widespread and multiphase enhancement has become common [7]. Although this latter technique has extended the application of CT, its use when a smaller number of phases would suffice cannot be justified. There is also anecdotal evidence that workload pressure may be adding to the problem. In the face of rising demand, radiologists may rely on CT examinations being performed with standard ‘‘catch all’’ protocols while they concentrate on other work; retrospective reporting of a comprehensive examination is efficient for the radiologist. Inexperienced radiologists in particular are likely to feel more confident the more sections they have available to read. However, this removes one of the key elements of radiation protection in CT, namely supervision by the radiologist, who should terminate the examination at the point at which it has delivered the information required for clinical management. The approach of ‘‘imaging overkill’’ may be tempting in current circumstances but cannot withstand serious enquiry. A new urgency exists in the form of multislice CT [8]. This technique offers almost no resistance to extending the examination, introduces new applications and will be widely used despite the fact that, examination for examination, absorbed dose may be up to 40% higher [9, 10]. Interventional CT and CT fluoroscopy pose a particular problem. The latter may use an exposure rate 10 times that of conventional fluoroscopy Received 27 March 2001 and in revised form 31 May 2001, accepted 18 June 2001. The British Journal of Radiology, 75 (2002), 1–4 E 2002 The British Institute of Radiology