| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Commentary |
National Radiological Protection Board, Chilton, Didcot, Oxon OX11 0RQ, UK
The concept of "reference doses" for common X-ray examinations was introduced in the UK in 1990 in a joint document by the Royal College of Radiologists (RCR) and the National Radiological Protection Board (NRPB) entitled Patient Dose Reduction in Diagnostic Radiology [1]. As a simple indication of "abnormally high doses", the document recommended using third quartile values of the distributions of mean doses observed for a particular examination on a representative sample of patients at each hospital participating in a national patient dose survey in the mid 1980s. In 1992 a National Protocol for Patient Dose Measurements in Diagnostic Radiology [2] provided practical guidance on how radiology departments could compare local performance with national practice using these reference doses. In view of the expected variability in doses between individual patients, mainly owing to differences in physique and pathology, it was recommended that local performance for a particular type of X-ray examination also be assessed in terms of the mean dose in a representative sample of standard sized patients. If such mean doses were found to exceed the corresponding reference dose, an investigation should take place to establish the cause and to take corrective action, unless the unusually high doses could be clinically justified. Thus, the purpose of the reference dose was to provide a trigger to the first step in the optimization of patient doses and identify those practices in most urgent need of investigation and corrective action. However, the fact that radiology departments had established a recognized practical method for measuring and comparing patient doses meant that they could easily progress beyond this first step and go on to develop their own quality assurance and clinical audit programmes that included regular patient dose assessments. It was emphasized that to be just below the reference dose was not necessarily an indication of optimum performance and, to further optimize local practice, it may be helpful to set local dose audit standards lower than the national reference doses and based on these local dose distributions.
As an investigation for situations where patient doses are unusually high and in most urgent need of reduction, a similar concept to reference doses was subsequently adopted by the International Commission on Radiological Protection (ICRP) in ICRP Publications 60 and 73 [3, 4]. The latter report introduced the term "diagnostic reference level" (DRL) and recommended that values should be selected by professional medical bodies, reviewed at intervals representing a compromise between the necessary stability and the long-term changes in observed dose distributions, and be specific to a country or region. Interestingly, the ICRP also recommended that DRLs need to be established only for common types of diagnostic examination and broad types of equipment. They are not intended to be used in a precise manner and a multiplicity of levels would reduce their usefulness.
In 1996 the NRPB reviewed all the data received from radiology departments throughout the country that had been following the national protocol [5]. Analysis of some 50 000 patient dose measurements from 375 UK hospitals showed an average 30% reduction in mean doses for common types of X-ray examination since the survey in the mid 1980s, and that less than 10% of hospitals were exceeding the original national reference doses. Despite the appreciable increase in the amount of patient dose data available, it was still concentrated on the same nine types of relatively common X-ray examination that featured in the earlier national survey. The need for a possible revision of the reference doses was discussed, but it was recognized that consultation with relevant professional and regulatory bodies would be necessary before any formal recommendations could be made.
This became more urgent in the following year with the adoption of the Medical Exposure Directive (MED) [6] by the Council of the European Union. The directive placed a requirement on all member states to promote the establishment and use of DRLs and to make available guidance for this purpose, having regardto European DRLs where available. In the UK most of the requirements of the MED, including those concerned with DRLs, were implemented last year by the Ionising Radiation (Medical Exposure) Regulations 2000 (IRMER) [7]. Specifically, IRMER requires that every hospital performing diagnostic radiology shall establish DRLs and shall have written procedures for their use, which should include local investigation and appropriate corrective action whenever they are consistently exceeded.
The Department of Health (DoH) convened a Working Party in January 2000, comprised of representatives from all the professional bodies involved in diagnostic medical exposures, to consider broad policies for the establishment and use of DRLs. The main conclusions, in relation to X-ray examinations, from this first meeting were:
Having established these broad principles, it was recognized that more detailed practical guidance on how to establish, review and check compliance with DRLs at the individual hospital or practice level in the context of the overall regulatory patient protection framework was urgently required. The Institute of Physics and Engineering in Medicine (IPEM) consequently took the lead in setting up a second DRL Working Party, mainly comprising medical physicists but still including representatives from NRPB, RCR, the College of Radiographers and the British Institute of Radiology, to provide authoritative guidance on these matters. In view of the urgency for providing some assistance to those trying to implement these new regulations, the IPEM Working Party issued a preliminary guidance statement at the end of 2000, outlining and expanding upon the practical implications of the above policies decided by the DoH Working Party [8]. The statement was essentially intended to prevent precipitant action by hospital physicists or radiographers who might, for example, have interpreted the regulations as requiring that DRLs be established for every type of procedure and be adhered to for every individual patient exposure in their radiology department. It pointed out that, in the first instance, local adoption of the new national DRLs for those types of examination regularly performed in the radiology department would be sufficient. Although there is a policy for annual review of locally adopted DRLs, such reviews can be very brief and will involve no changes if there have been no developments in national DRLs and no major local dose surveys during the year. Moreover, the recommendations in the national protocol [2] for comparing local performance with national reference levels, by monitoring the doses to a representative sample of patients at least once every 3 years, would be suitable for checking whether the DRLs were being consistently exceeded.
Many radiology departments (at least 75%) will naturally find that their mean dose for a particular examination is below even the new national DRL and may wish to set lower local reference doses as a guide to the further optimization of local practice. Additionally, they may wish to set reference levels for types of examination for which no national DRLs are currently available. If local patient dose data are available from a sufficiently large number of hospitals, then regional DRLs can be established in the same way as the national DRLs, i.e. as the third quartile of the distribution of the mean doses to representative groups of standard sized patients at each hospital. Regional DRLs still have the function of identifying those hospitals using abnormally high doses where patient protection measures are most urgently required. However, if patient dose data are available from only one or two hospitals, they can only be used to monitor local trends in patient dose with time, or perhaps to demonstrate differences between X-ray rooms, practitioners or operators. Such local data are a valuable tool for local dose audit and optimization initiatives but they cannot, by themselves, be used to set a reference level to indicate whether local doses are abnormally high. As before, wide variations in individual patient doses are to be expected and it is only sensible to compare the mean dose values (or perhaps the median, which is less influenced by extreme outliers) on representative groups of patients to monitor local trends with time, equipment or technique. Mean or median values, used as dose audit standards to monitor local trends, are not the same as third quartile values used to indicate abnormally high doses on a national or regional scale. It is consequently rather misleading to refer to such values as DRLs, as originally defined by the ICRP and the MED, and perhaps an alternative name such as "local audit standards" would be more suitable.
With national DRLs currently available for only seven types of radiographic examination (skull, chest, thoracic spine, lumbar spine, abdomen, pelvis and intravenous urography) and two types of fluoroscopic examination (barium meal and enema), it is important to consider the need and priorities for expanding this list to other examinations. There are considerable resource implications involved in setting DRLs and in checking whether they are being consistently exceeded, so it is sensible to concentrate any further effort by staff of busy radiology departments on those examinations that are frequently undertaken and that involve the most significant radiation doses and risks. DRLs are probably already available for a sufficient number of common radiographic examinations to ensure adequate optimization of the techniques and equipment used in any hospital for this type of relatively low dose X-ray examination. An exception is dental radiography, which now accounts for about one-third of all X-ray examinations [9] and is mostly conducted in small dental practices with only remote access to radiation protection advice. Consequently, the NRPB is developing reference doses for intra-oral and panoramic dental radiography based on nationwide data collected by its Dental X-ray Protection Services [10], which it will recommend to the IPEM and DoH DRL Working Parties for consideration as national DRLs. Otherwise, it is the higher dose examinations involving CT or extensive fluoroscopy that most merit further attention.
CT now comprises approximately 5% of all medical X-ray examinations [9], with effective doses of up to 20 mSv and radiation-induced cancer risks of up to 1 in 1000 per examination. The urgent need for reference doses for routine CT examinations has already been recognized inEuropean Guidelines on Quality Criteria for Computed Tomography [11]. The European CT reference doses are largely based on UK data collected over the past 10 years and, in the short term, probably represent the best available source of national DRLs for this important imaging modality. Meanwhile, the NRPB, through the IPEM DRL Working Party, will be urgently looking at ways for revising and extending these reference doses to incorporate the results of more recent UK CT patient dose surveys.
Hospitals conducting substantial numbers of angiographic or interventional procedures involving extensive periods of fluoroscopy should consider developing DRLs for some of the more common procedures. However, as examinations become more complicated, and particularly when the outcome is a surgical intervention and not just a diagnosis, there is likely to be much greater individual variation in the patient doses required. Unavoidable complications frequently arise with the positioning of catheters and guidewires, and interventional procedures are usually clinically open-ended, having to continue until the surgery is complete. In these situations it is important to specify the type and extent of the examination or procedure as closely as possible to try and reduce this source of variation in patient doses. Even so,uncertainties in the mean values on small (realistic) samples of patients will be high and comparisons with national DRLs or the determination of trends in local dose audits will be far from precise. However, as the ICRP stated [4], DRLs are not intended to be used in a precise manner but simply as a rough indication of situations that do not appear to be providing the best radiation protection for patients. Again, they provide just the first step in the optimization process by most effectively directing corrective action to where doses and risks are highest.
So what is the way forward? First, don't panic. It is probably true to say that by following earlier recommendations from professional bodies on patient dose monitoring, radiology departments in the UK have already been following the spirit of the MED far more closely than those in any other member state of the European Union. Indeed, in the consultative document on IRMER, the DoH recognized that most of the procedures required by the proposed regulations already existed and only needed a degree of formalization. The costs associated with establishing DRLs at the local level were consequently expected to be small. All that is required, both to comply with the new regulations and to provide the most cost effective protection for patients, is to formalize existing procedures whereby representative patient doses are regularly compared with the new national DRLs for those examinations routinely performed in the radiology department. Lower regional DRLs or local audit standards can be established if there are sufficient dose data to do so, and these should eventually be extended to a number of relatively high dose procedures involving CT or extensive fluoroscopy that are frequently conducted in the department. The value of local patient dose data will be considerably increased if, as well as being used in local dose audits, they are forwarded to the NRPB for national collation and review. In this way national DRLs will be more representative of current radiological practice throughout the UK and will become available for a wider range of important radiographic examinations.
Received for publication February 5, 2001. Revision received April 10, 2001. Accepted for publication May 1, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  F. A. Mettler Jr, W. Huda, T. T. Yoshizumi, and M. Mahesh Effective Doses in Radiology and Diagnostic Nuclear Medicine: A Catalog Radiology, July 1, 2008; 248(1): 254 - 263. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
