| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Commentary |
Department of Medical Physics, Royal Infirmary of Edinburgh, 51 Little France Crescent, Edinburgh EH16 4SA, UK
We live in a risk averse society. The public is prepared to believe scare stories regarding the risks associated with many manufactured devices and materials whether they relate to genetically modified crops, mobile phone masts, or the measles, mumps and rubella vaccine. Attitude towards risk is tempered by the degree of imposition and by the perceived benefit — "my mobile phone is safe, your transmitter is dangerous".
In this climate we should ask whether it is to the benefit of the radiological community, or indeed to patients, to promote an awareness of the potential harm from X-rays when the risks are, as discussed below, negligible in most circumstances. A typical scenario is for the benefits of MRI to be promoted on the basis of the absence of risk from ionizing radiation. This is frequently stated when trying to fund a new scanner. It is surely more important to promote the benefits of MRI: high quality 3-dimensional images providing clinical information that cannot be obtained from any other diagnostic tool. Radiology does itself a disservice when it talks up the risk from imaging modalities that have an essential role in clinical diagnosis.
Radiation risk is promoted in several ways. For example the first part of a radiologist's specialist training in the UK is concerned with the requirement of the Ionising Radiation (Medical Exposure) Regulations 2000, IR(ME)R, to have had adequate training, specified in the legislation as being principally concerned with the physics of ionizing radiations and radiation protection. The implication is clear: X-ray and radionuclide imaging modalities are associated with significant hazards whereas ultrasound and MRI are not. This implication is reinforced by the possible requirement to notify the IR(ME)R Inspectorate of certain incidents involving an overexposure of a patient, for example the injection of the incorrect radiopharmaceutical, when there is no such requirement for non-ionizing radiations.
Medical imaging is a field of endeavour in which artificial energy sources are used to penetrate the body. It has been said that if X-ray examinations with present day techniques and equipment had been the first and only significant exposure of humans to ionizing radiation, then we would have no doubt that they were safe. But this was not the case. The early uses of diagnostic X-ray equipment inflicted massive doses of radiation on both operator and patient, and other uses (and abuses) of ionizing radiations have firmly established that ionizing radiations cause cancer. This fact, recognised since early in the last century, has promoted the simplistic view: non-ionizing radiation – good; ionizing radiation – bad. The view is reinforced by the layers of regulation concerning the use of these radiations with no comparable legislation governing the use of ultrasound or MRI. The introduction of ultrasound and MRI was not without concerns about their safety. These concerns have not been fully resolved but the extensive use of these modalities and the lack of any evidence of harm at the exposure levels used for imaging is taken as sufficient demonstration that any risk that there may be is negligible, particularly in the context of the proven clinical benefits.
Diagnostic radiology contributes approximately 14% of the average annual dose received by the UK population [1]. Other artificial sources in the UK contribute less than 0.4% with just 0.012% arising from the disposal of radioactive waste. The patient protection legislation in Europe is attributed to the public concern over the impact of artificial sources in the environment. The argument runs that if the public demand stringent action to minimize the dose from one particular source, then there should be concerted action to ensure that the population dose from the much greater source of radiation (i.e. medical exposure) is subject to equally rigorous standards.
In discussions of radiation hazards, one factor often neglected is that risk is proportional to dose. This is the received wisdom, promoted by the International Commission on Radiological Protection, the so called linear no-threshold theory (LNT). LNT has many challengers but it is not the purpose of this paper to debate that issue. LNT is the basis of our legislation and is required to be the basis on which we seek to minimize risk. In dealing with risk we accept that there is a 1 in 20 000 chance of developing fatal cancer following irradiation to an effective dose of 1 mSv. Legislation does not permit us to act as though the risk is any less, but at the same time we should not suggest that the risk might be any greater.
The range of doses in diagnostic radiology spans almost 5 orders of magnitude and the resultant risks vary to the same extent. A CT scan of the abdomen and pelvis gives a dose of about 10 mSv with an associated fatal cancer risk of 1 in 2000 (approximately equal to the annual death rate from all causes for people in their mid-20s). An X-ray of the hand gives a dose no greater than 0.2 µSv with a fatal cancer risk of 1 in 100 million (approximately one-tenth of the annual risk of being killed by lightning).
In recent years the contribution of CT scanning to the collective dose has been a major concern. Following publication of the UK survey on CT doses, the National Radiological Protection Board (NRPB) [2] reported that, whereas CT represented only 2% of all imaging involving X-rays, its contribution to the collective dose was approximately 23%. More recently it has been noted that the frequency of examinations has increased to about 4% and the contribution to collective dose has probably increased to 40% [3]. Such evidence, that a few examinations contribute disproportionately to the collective dose, is used to inform decisions on investment in alternative imaging modalities or in other investigation techniques. However, we can look at this information in another way. The typical dose for a chest X-ray is 17 µSv, for an X-ray of the extremities the dose per radiograph lies between 0.2 µSv and 3 µSv [4], and for dental radiology it is 4 µSv and 7 µSv, respectively, for two bitewing films using E-speed film and for a panoramic radiograph [5]. These very low dose examinations represent about 70% of all medical and dental exposures, but they contribute (on average) individual doses no greater than 3 µSv per year, that is less than 1% of the dose arising from all medical exposures. The average annual dose from air travel is 20 µSv, nearly 7 times greater [1].
IR(ME)R is concerned with all medical exposures and there is no de minimis dose below which the regulations do not apply. IR(ME)R requires the justification of individual medical exposures on the basis of the balance between risk and benefit. In applying the regulations it is important to recognise that when risk is extremely low then the benefit need only be proportionate to that risk and as has been discussed above, for the majority of X-ray examinations that risk is negligible (defined in my dictionary as "such as may be ignored because very little or very unimportant").
An NRPB leaflet containing information for patients [6] poses the question "X-rays: how safe are they?" and concludes that even for high dose examinations the dose "represents a very small addition to the underlying cancer risk from all sources". For those examinations with effective doses of 20 µSv or less (70% of the total) the risk is described as negligible. In every day language we might conclude that the majority of X-ray examinations are indeed safe.
In general the balance between risk and benefit is not judged by a simple mathematical calculation but let us try an example. It might be acceptable to advocate chest radiography (1 in 1 000 000 risk) in particular clinical circumstances even if 99 out of 100 films were expected to have no abnormal finding. The ratio of winners to losers would still be 10 000 to 1, a very substantial benefit to risk ratio. This may not be practicable as a general approach to justification, but a mathematical approach is taken in breast screening to provide the reassurance that the potential harm, i.e. numbers of cancers induced, is very much less than the numbers detected that can be successfully treated as a result.
The purpose of this communication is not to criticise the legislation surrounding radiation protection. IR(ME)R can be viewed as a guide to best practice in clinical radiology, demanding as it does, for example, high standards of training, written procedures and quality assurance. X-raying the wrong patient is bad practice even if the risk is zero. The purpose is threefold:
Received for publication November 24, 2003. Revision received February 11, 2004. Accepted for publication March 30, 2004.
 |
References |
|---|
 Top References |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Mascalchi, G. Belli, M. Zappa, G. Picozzi, M. Falchini, R. D. Nave, G. Allescia, A. Masi, A. L. Pegna, N. Villari, et al. Risk-benefit analysis of X-ray exposure associated with lung cancer screening in the Italung-CT trial. Am. J. Roentgenol., August 1, 2006; 187(2): 421 - 429. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D Rawlings Options for radiation protection of the patient Br. J. Radiol., October 1, 2005; 78(934): 877 - 879. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
