- Email: [email protected]
UK population dose from medical X-ray examinations
European Journal of Radiology 50 (2004) 285–291
UK population dose from medical X-ray examinations D. Hart∗ , B.F. Wall National Radiological Protection Board, Chilton, Didcot, Oxon OX11 0RQ, UK Received 14 May 2003; received in revised form 10 June 2003; accepted 11 June 2003
Abstract Objective: To assess the annual per caput and collective effective dose to the United Kingdom population from medical and dental X-ray examinations. Method: The results of a detailed survey of the frequency of X-ray examinations during the financial year 1997/1998 were combined with contemporary data on the effective doses typically received by patients. The resulting per caput and collective dose for 1997/1998 was updated to 2001/2002 by using annual statistics on the total numbers of computed tomography (CT), interventional and conventional examinations collected by the English Department of Health. Results: The annual per caput effective dose for the UK in 2001/2002 was estimated at 0.38 mSv. Over the last 10 years CT has more than doubled its contribution and is now responsible for 47% of the collective dose from medical X-rays. The contribution from conventional radiographic and fluoroscopic examinations has nearly halved to about 34%. Interventional and angiographic procedures together contribute the remaining 19%. Conclusions: The annual per caput effective dose of 0.38 mSv is low in comparison with other countries having similarly developed systems of health-care. This is due to both a lower frequency of X-ray examinations per head of population and generally lower doses in the UK than in other developed countries. © 2003 Published by Elsevier Ireland Ltd. Keywords: Collective dose; Effective dose; Computed tomography; X-ray examinations
1. Introduction Estimates of the population dose (the collective dose, or the average per caput effective dose) provide useful information on the relative contribution of different sources of ionising radiation to population exposure. They can be used to compare the contribution from diagnostic radiology with those from natural or other artificial sources of radiation and to see how the contributions differ between different countries or regions. More specifically, they allow comparison of the contributions from different types of X-ray examination or from different medical imaging modalities. Such information provides guidance on where best to concentrate efforts on dose reduction, so as to optimise the protection of patients in the most cost-effective manner. An estimate by the National Radiological Protection Board (NRPB) in 1999 [1] put the contribution from patients undergoing X-ray examinations at nearly 90% of the total population dose from all artificial sources in the UK, with diagnostic nuclear medicine procedures contributing ∗ Corresponding author. Tel.: +44-123-582-2647; fax: +44-123-583-3891. E-mail address: [email protected] (D. Hart).
0720-048X/$ – see front matter © 2003 Published by Elsevier Ireland Ltd. doi:10.1016/S0720-048X(03)00178-5
a further 8%. (Radiotherapy exposures were deliberately excluded from this analysis). In contrast, all occupational and public exposures arising inadvertently from medical and other uses of ionising radiation, including the UK nuclear power programme, amounted to less than 3% of the total. While dominating the artificial sources of radiation exposure, medical X-rays accounted for only one-sixth of the population dose from natural background radiation. NRPB previously made an assessment of the collective dose from medical and dental X-rays in 1991. The UK collective effective dose equivalent from all diagnostic radiology was at that time roughly estimated at 20 000 man Sv, corresponding to an average per caput dose of 0.35 mSv [2]. NRPB recently made a detailed and more accurate estimate of the collective effective dose to the UK population from medical X-rays for the financial year 1997/1998. A figure of about 19 300 man Sv was derived, corresponding to an average per caput dose of 0.33 mSv [3]. That figure was subject to an estimated uncertainty of ±9%, and confirmed that the previous crude estimate [2] was reasonable. This article updates the detailed population dose estimate to the financial year 2001/2002, using annual statistics on radiology workload and some recent information on patient doses from specific examinations.
286
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
2. Method used to estimate UK population dose for 1997/1998
3. Method used to update the 1997/1998 estimate to 2001/2002
To estimate the annual UK population dose from all medical and dental X-ray examinations, information is required on the annual frequency and the mean effective dose for each type of examination. The former was provided by an NRPB survey of the frequency of 150 types of X-ray examination in the UK in 1997/1998 [4]. Nearly 60% of the hospitals in two of the eight health service regions in England provided detailed examination frequency data and the results were scaled up to the whole of the UK using radiology workload statistics from the Department of Health. Estimates of the mean effective dose for each examination were obtained from a number of sources, the predominant one being the National Patient Dose Database maintained by NRPB [5,6]. The database contained information collected in the period from 1988 to 2000 covering about 60 types of radiograph and 100 types of X-ray examination. Doses are recorded in the National Patient Dose Database mainly as entrance surface dose (ESD) values for individual radiographs and dose–area product (DAP) values for complete examinations. The typical dose for a specific radiograph or examination was taken to be the mean of the doses recorded in the National Patient Dose Database over the whole of the 1990s. Data for the whole decade were used in order to get a sufficiently representative sample size, particularly for the less common examinations. The mean dose for each examination was derived by firstly calculating the mean dose for the sample of patients measured in each radiology room and then taking the mean of these room mean values. In this way equal weight was given to each radiology room in the National Patient Dose Database. A downward trend over time in doses for common radiographic and fluoroscopic examinations, as shown in periodic analyses of the National Patient Dose Database [5,6], means that doses for conventional radiology are probably slightly overestimated for 1997/1998. NRPB report R262 [7] contains generalised conversion coefficients, in Tables 16–18, for estimating effective dose from ESD and DAP measurements, assuming that the X-ray spectra (tube voltage and total filtration) used are close to the average. Typical effective doses were derived from the mean ESD or DAP values using these conversion coefficients. For examinations consisting purely of radiographs, the effective doses from each radiograph were added to provide an effective dose for the complete examination. An effective dose estimate, derived from data in the National Patient Dose Database and conversion coefficients of the type given in NRPB-R262, was obtained for 90 examinations out of the 150. Effective doses for another 25 examinations were found in published literature. This included common computed tomography (CT) examinations, which had been covered by an NRPB survey in 1989 [8]. Effective doses for the remaining 35 examinations were estimated by comparison with similar examinations, for which doses were already available.
The annual frequency of examinations was updated by using radiology workload statistics gathered every year by the English Department of Health, the National Health Service Breast Screening Programme, and the Dental Practice Board. For some examinations, the effective dose is known to have changed, and the population dose was altered proportionately. For those examinations for which no recent dose information was available, the effective dose was assumed to be unchanged from the 1997/1998 estimate. The English Department of Health collects statistics on the number of medical imaging procedures conducted each year by all National Health Service (NHS) hospitals in England. These are known as KH12 returns and have been collected in the current format since 1995/1996 [9]. Non-interventional and interventional procedures are treated as separate categories (the definition of interventional procedures is discussed in an explanatory note at the end of this article). Each of the two categories is broken down into the following modalities: CT, magnetic resonance imaging (MRI), obstetric ultrasound, non-obstetric ultrasound, nuclear medicine, fluoroscopic examinations, and radiographs not involving fluoroscopy. In order to ensure completeness of the statistics, the Department of Health uses the data from previous years to estimate the contribution from the few hospitals that failed to supply actual data in any specific year. These statistics thus give the total number of CT and other X-ray examinations performed each year, and the number of these that are interventional. Any trends in these numbers, therefore, provide a means to update the 1997/1998 annual frequency estimate for medical X-ray examinations to the year 2001/2002. Scotland, Wales and Northern Ireland (with a combined population of 16% of the whole UK) were assumed to have experienced similar trends to England. Mammographic screening, and imaging procedures performed outside NHS hospitals such as dental and chiropractic radiography, are not covered by the KH12 returns. We have attempted to obtain the latest frequencies for these imaging procedures from other sources in order to update their contribution to the population dose. Allowance has also been made for changes in doses for some examinations. The latest review of the National Patient Dose Database (1996–2000) [6] found reductions in doses for common radiographic and fluoroscopic examinations. Therefore, the contribution to population dose from such examinations has been reduced proportionately. It was assumed that doses from CT, interventional and angiographic examinations have remained the same.
4. Results The KH12 returns show that CT, MRI and non-obstetric ultrasound are all steadily increasing in frequency, while
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
287
Table 1 Numbers of interventional and non-interventional procedures in English NHS hospitals
Radiographs CT Fluoroscopy US Non-obstetric US Obstetric MRI Radioisotopes
1997/1998
1998/1999
1999/2000
2000/2001
2001/2002
19 474 590 1 172 656 1 179 979 2 884 883 1 905 649 473 074 722 096
19 876 933 1 254 474 1 244 632 3 068 276 1 950 158 522 138 699 654
19 967 296 1 359 852 1 256 965 3 289 919 1 965 411 585 797 727 255
19 913 022 1 488 752 1 253 847 3 430 753 1 951 829 632 594 539 141
19 806 876 1 625 304 1 222 296 3 579 412 1 992 567 705 706 537 653
Table 2 Numbers of interventional procedures in English NHS hospitals
Fluoroscopy Radiographs CT US Non-obstetric US Obstetric Radioisotopes MRI
1997/1998
1998/1999
1999/2000
2000/2001
2001/2002
217 564 49 723 13 049 70 805 19 072 4830 1422
231 204 51 225 15 374 71 843 16 881 2500 1510
241 802 53 994 17 941 81 904 15 316 3397 1215
253 532 57 357 18 339 84 408 32 756 3865 902
276 197 138 928 31 460 113 286 43 694 16 834 5735
obstetric ultrasound, radiographs, fluoroscopy and nuclear medicine (radioisotopes) have very little growth over recent years (see Table 1). Radiographic examinations are easily the most common of all these procedures. Their frequency rose slightly in the first half of the period 1997/1998 to 2001/2002, but fell in the second half. The number of CT examinations has increased by 39% in the 4 years from 1997/1998 to 2001/2002. The growth in interventional radiology involving fluoroscopy or radiographs was 55% over this period (see Table 2). The number of conventional diagnostic radiology examinations (radiographs and fluoroscopy, not including interventional procedures) has increased by only 1% for the period 1997/1998 to 2001/2002. On the assumption that typical patient doses per procedure have not changed, the 39% growth in frequency of CT examinations boosts the annual collective dose to the UK population for CT from 7662 to 10 650 man Sv for 2001/2002 (see Table 3), and the 55% growth in interventional radiology boosts it from 1239 to 1920 man Sv. As stated above, conventional radiology has increased in frequency by 1% from 1997/1998 to 2001/2002, so this small increase has been applied to all conventional radiology examinations (except mammography and dental radiography) that are listed in Table 3. However, the latest review of the National Patient Dose Database [6] showed there had been a reduction in doses for some common radiographic and fluoroscopic examinations, so this reduction has also been applied in Table 3. The four largest contributors to the UK population dose from conventional radiology are barium enema, abdomen, lumbar spine, and pelvis examinations. Together they contribute about 60% of the total population dose from conventional radiology. In Table 3 their contributions have been reduced by between 7 and 15% to match the
Table 3 Contributions to UK collective dose (man Sv)
CT Interventional Conventional radiology Barium enema Abdomen Lumbar spine Pelvis Mammography Dental Others Angiography Total a
1997/1998a
2001/2002
7662 1239 8473 2588 852 825 644 466 77 3021 1923 19 298
10 650 1920 7720 2379 749 692 559 513 82 2746 2423 22 713
Table 3 of [3].
dose reductions found for these examinations when comparing data for 1996–2000 [6] with the data for 1990–1999 that were used in the 1997/1998 population dose estimate [3]. We have assumed there has been a dose reduction for all other conventional radiology examinations of 10% for the 4-year period 1997/1998 to 2001/2002, in line with the above reductions, and with the latest review of the National Patient Dose Database [6] which found a general reduction of 16% over a 5-year period. The fifth largest contributor to UK population dose from conventional radiology is mammography. The latest data from the National Health Service Breast Screening Programme are for 2000/2001 [11] and show that the number of women being screened has increased to 1.6 million, from 1.4 million in 1997/1998. The breakdown of these figures into different types of examination is shown in Table 4,
288
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
Table 4 Update of UK collective dose from mammography Frequencies
First round of screening Subsequent screening rounds Recall for assessment Symptomatic Total
Collective dose (man Sv)
1997/1998
2000/2001
1997/1998
2000/2001
374 000 960 000 66 000 326 000
377 000 1 169 000 83 000 326 000
138.4 192.0 15.2 120.7 466
139.5 234.0 19.1 120.7 513
where we have assumed that the frequency of symptomatic referrals to hospitals has not changed. We can assume that doses per mammographic procedure have not changed significantly, because the imaging techniques and the number of views taken per breast in specific rounds of the screening programme did not change between 1997 and 2001. The application of the percentage changes in frequency to the collective dose in columns 4 and 5 of Table 4 show that mammography has increased its contribution by 10% to 513 man Sv. This has also been entered into Table 3. Data from the Dental Practice Board [12] show that the number of X-ray examinations carried out by dentists for the NHS in England and Wales has increased by 6% over the period 1997/1998 to 2001/2002. We have assumed the same increase for the private sector, and that the increases apply across the whole of the UK. Assuming no change to the dose per examination, the contribution to the collective dose, therefore, increases from 77 to 82 man Sv as shown in Table 3. It has previously been difficult to get accurate statistics on the annual number of chiropractic X-ray examinations, and this has continued to be the case. Fortunately, the estimated contribution to the 1997/1998 collective dose from chiropractic was only 34 man Sv. Even a major change in the frequency of chiropractic X-ray examinations would thus have little effect on the total UK population dose. Therefore, chiropractic has been assumed to maintain the same absolute contribution to population dose as before. Overall, the annual collective dose from conventional radiology is estimated to have decreased by about 9% from 8473 to 7720 man Sv as shown in Table 3. The KH12 returns give no specific information on angiography. We have, therefore, estimated the annual growth in the frequency of angiography in the UK from 1997 to 2002 to be the same as the annual growth in the years prior to 1997. Table 5 shows the number of angiographic examinations performed in Great Britain [4,13], and France [14] for
the period 1977–2000. Great Britain showed growth rates of 7% per annum in 1977–1983, and 6% p.a. in 1983–1997. France had a growth rate for coronary angiographies of 7% p.a. during 1991–2000, which is very similar to growth rates for Great Britain while covering years which overlap with our update period. It, therefore, seems reasonable to assume that the 6% p.a. increase in the number of angiographies in the UK continued during 1997/1998 to 2001/2002. We have also assumed that the dose per examination has remained constant. An increase of 6% p.a. amounts to a 26% increase over a 4-year period. Consequently, the collective dose from diagnostic angiography of 1923 man Sv estimated for 1997/1998 [3] has been increased to 2423 man Sv for 2001/2002, as shown in Table 3. Taking all the above mentioned increases and decreases into account, the total collective dose in the UK from all X-ray imaging procedures is estimated to have increased from 19 300 man Sv in 1997/1998 to 22 700 man Sv for 2001/2002 as shown in Table 3. The population of the UK had risen by only 1% to 59.8 million by the year 2000, and so the annual average per caput dose from medical X-ray examinations has increased slightly to 0.38 mSv.
Table 5 Annual number of angiography examinations (thousands)
Great Britain Francea a
1977
1983
92.3
135.6
Number of coronary angiographies.
1991
1997/1998
2000
304.0 131.6
236.1
Fig. 1. Contributors to UK population dose from medical X-rays (2001/2002).
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
289
Fig. 2. Biggest contributors to population dose.
5. Discussion CT now represents 7% of all X-ray examinations carried out in NHS hospitals (excluding mammographic screening) but contributes 47% of the total population dose to the UK from medical X-ray examinations. Angiographic procedures are responsible for 11% of the population dose, and interventional procedures for 8%, while conventional radiology is now down to a contribution of 34%. Thus, for the first time, CT has overtaken conventional radiology as a contributor to the UK population dose. The relative contributions of conventional, CT, angiographic and interventional procedures to the population dose from all medical X-ray examinations are shown on the pie-chart in Fig. 1. Fig. 2 shows the 13 examinations that are the biggest contributors to UK population dose, and indicates their percentage contributions to frequency and population dose. Together the 13 examinations account for nearly 80% of the population dose. The large contribution from CT to the UK population dose is comparable to the situation in other countries. UNSCEAR 2000 [15] shows that on average for countries in Health-care
level I, CT represented 6% of all diagnostic medical X-ray examinations, and contributed 41% of the total annual population dose in the period 1991–1996. Mettler et al. [16] calculated that CT scans at one large hospital in the USA during 1998 and 1999 contributed about 67% of the total effective dose from all medical X-ray examinations at that hospital. A comparison of the estimated UK annual per caput dose of 0.38 mSv from medical radiology is made with similar data from the 1990s for other countries in Table 6, using information reported by UNSCEAR [15] and other sources as referenced. The data are arranged in order of decreasing size of the annual per caput effective dose. It can be seen that the UK has a low per caput dose compared with other nations with well-developed systems of health-care. It is notably one-fifth of the value estimated for Germany and just over one-third of the values given for France and Switzerland, although it should be recognised that there are likely to be large uncertainties associated with many of these values. The relatively low value for the UK, especially in comparison with other European countries, appears to be due to
290
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
Table 6 International comparison of annual per caput effective dose from medical radiologya Country
Time period
Annual per caput effective dose (mSv)
Source
Germany France Switzerland Canada Russia Australia Norway Poland Bulgaria Portugal Sweden Romania Netherlands USA Ukraine Finland Spain (regional) UK Denmark Taiwan Brazil Malaysia
1990–1992
1.9 1.0 1.0 0.94 0.9 0.8 0.8 0.8 0.75 0.71 0.68 0.61 0.52 0.5 0.5 0.45 0.4 0.38 0.36 0.23 0.09 0.05
[17] [15] [18] [15] [15] [19] [20] [15] [15] [21] [15] [15] [22] [15] [23] [15] [24] This article [15] [25] [15] [26]
1998
1996 1993
1991
1998 1994 1990 2001/2002 1993 1994
a Based on Table 29 in Annex D, Volume 1, of the UNSCEAR 2000 Report [15].
Table 7 International comparison of examination frequencies and typical effective dosesa Type of examination
Chest Limbs and joints Lumbar spine Thoracic spine Cervical spine Pelvis/hips Head Abdomen Upper GI tract Lower GI tract Cholecystography Urography Mammography CT CT head CT body Angiography Interventional procedures
Number of examinations per 1000 population per year
Typical effective dose (mSv)
UK
Health-care level I
UK
Health-care level I
141 147 19 5 14 31 28 21 4.9 6.1 1.2 4.6 27 21
281 166 48 13 32 35 59 41 42 8.7 3.1 12 25 57
0.02
0.14
1.3 0.7
1.8 1.4
0.7 0.04 0.7 2.6 7.2
0.83 0.07 0.53 3.6 6.4
2.4 0.06
3.7 0.07
2 9
2.3 13.3
5.2 4.5
6.8 3.0
a Selected from Tables 12, 15 and 30 in Annex D, Volume 1, of the UNSCEAR 2000 Report [15].
both a lower frequency of X-ray examinations and generally lower doses per examination. This is evident from Table 7, which shows such statistics for the UK for the 1990s, compared with the average for countries in Health-care level I as reported by UNSCEAR [15]. The second and third columns show the annual numbers of medical X-ray procedures per thousand population for the UK and the average for Health-care level I. For 14 out of the 16 types of procedure shown, UK frequencies are below the average frequencies for Health-care level I. On average, the frequency of examinations in the UK is 40% lower than for Health-care level I. The fourth and fifth columns show typical effective doses to patients for some common types of diagnostic examinations for the UK and the average values for Health-care level I. For ten out of the 12 types of examination for which doses are shown, the UK doses are below the average doses for Health-care level I. On average for the examinations listed, UK effective doses are about 20% lower.
6. Conclusions The annual per caput dose from medical and dental X-ray procedures in the UK for 2001/2002 has been estimated by updating the results of a survey of the frequency of 150 types of examinations in 1997/1998 and combining them with recent estimates of the radiation doses from such examinations. The per caput dose from all X-ray imaging performed in NHS and private sector hospitals and clinics is estimated to be 0.38 mSv for the financial year 2001/2002. This estimate is similar to the previous rough estimate for 1991, and is low in comparison with that for other countries with similarly developed systems of health-care. This is mainly due to a lower frequency of X-ray examinations, but also to generally lower doses per examination in the UK. The relative contributions of some types of examination to the UK population dose from medical X-rays have changed considerably since 1991. CT has more than doubled its contribution and is responsible for 47% of the total. Interventional and angiographic procedures, which often involve prolonged fluoroscopy and hence result in high individual doses, have also increased in frequency and provide 19% of the total. Conventional fluoroscopic and radiographic examinations are making a smaller contribution. This is partly due to a drop in frequency of examinations such as barium meals that are being slowly replaced by endoscopy, and of biliary and urinary tract examinations where ultrasound imaging provides a viable alternative. However, the major factor responsible for reducing the per caput and collective doses for these conventional X-ray examinations is the average drop of about 50% in the dose per examination, seen in the period 1985–2000 and estimated from the latest review of the National Patient Dose Database [6]. These conventional X-ray examinations contribute about 34% of the total population dose, which represents a reduction by about a factor of two since 1991.
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
The increasing attention given in recent years in the UK to radiation protection for conventional examinations, with the development of national patient dosimetry protocols and reference doses, has played a significant part in this substantial reduction in population dose. Widespread local monitoring of patient doses and comparison with national norms have undoubtedly encouraged the adoption of dose-efficient procedures and the introduction of dose-saving features into X-ray imaging equipment. With the now much increased contributions of CT, angiography and interventional radiology to the population dose, there is a clear need to develop radiation protection and optimisation activities for these high dose procedures to the same level as has been already achieved for conventional radiology.
7. Explanatory note The English Department of Health defines interventional procedures in its KH12 form as “any invasive procedure during imaging which is performed other than primarily for the introduction of a contrast medium. Endoscopic Retrograde CholangioPancreatography (ERCP) would count as interventional by virtue of the endoscopy”. This differs from the more widely adopted definition of interventional radiology used by NRPB, which is “minimally invasive procedures which use imaging guidance for therapeutic purposes” [4,10]. Under the latter definition ERCPs are only interventional if they are therapeutic. This differing classification implies up to a 20% larger number of interventional radiology procedures for Department of Health statistics as compared with NRPB data if, in the worst case, no ERCPs in the latter data are therapeutic. This difference over ERCPs will become less noticeable during the coming years, since ERCPs are becoming predominantly a therapeutic procedure. Moreover, any numerical difference should have little effect on the upward or downward percentage change in the number of procedures, which is all that is needed to update 1997/1998 figures to 2001/2002. Acknowledgements We are grateful to Julie Lawlor of the Hospital Activity Statistics Team at the Department of Health, for supplying recent statistics to us in advance of publication. References [1] Hughes JS. Ionising radiation exposure of the UK population: 1999 review. Chilton: NRPB-R311, 1999. [2] Hughes JS, O’Riordan MC. Radiation exposure of the UK population—1993 review. Chilton: NRPB-R263, 1993. [3] Hart D, Wall BF. Radiation exposure of the UK population from medical and dental X-ray examinations. Chilton: NRPB-W4, 2002.
291
[4] Tanner RJ, Wall BF, Shrimpton PC, Hart D, Bungay DR. Frequency of medical and dental X-ray examinations in the UK—1997/1998. Chilton: NRPB-R320, 2000. [5] Hart D, Hillier MC, Wall BF, Shrimpton PC, Bungay DR. Doses to patients from medical X-ray examinations in the UK—1995 review. Chilton: NRPB-R289, 1996. [6] Hart D, Hillier MC, Wall BF. Doses to patients from medical X-ray examinations in the UK—2000 review. Chilton: NRPB-W14, 2002. [7] Hart D, Jones DG, Wall BF. Estimation of effective dose in diagnostic radiology from entrance surface dose and dose–area product measurements. Chilton: NRPB-R262, 1994. [8] Shrimpton PC, Jones DG, Hillier MC, Wall BF, Le Heron JC, Faulkner K. Survey of CT practice in the UK: part 2, dosimetric aspects. Chilton: NRPB-R249, 1991. [9] Acute services team of the performance analysis branch of the NHS Executive. Imaging and radiodiagnostics: England. Leeds: Department of Health (series 1995–1996 to 2001–2002). [10] World Health Organisation. Efficacy and radiation safety in interventional radiology. Geneva: WHO, 2000. [11] National Health Service Breast Screening Programme. Building on experience: breast screening programme annual review 2002. Sheffield: NHS Breast Screening Programme, 2002. http://www. cancerscreening.nhs.uk/breastscreen/publications/2002.review.html. [12] Dental Practice Board http://www.dpb.nhs.uk [13] Wall BF, Hillier MC, Kendall GM. An update on the frequency of medical and dental X-ray examinations in Great Britain—1983. Chilton: NRPB-R201, 1986. [14] Le Tourneau T, Blanchard D, Lablanche JM, Monassier JP, Morice MC, Cribier A, Meier P, Puel J. Evolution of interventional cardiology in France during the last decade (1991–2000) [in French]. Archives des maladies du coeur et des vaisseaux 2002;95(12):1188–94. [15] UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation). Sources and effects of ionising radiation. Volume 1: Sources, Annex D, Medical radiation exposures. New York: United Nations, 2000. [16] Mettler Jr FA, Wiest PW, Locken JA, Kelsey CA. CT scanning: patterns of use and dose. J Radiol Prot 2000;20:353–9. [17] Kaul A, Bauer B, Bernhardt J, Nosske D, Veit R. Effective doses to members of the public from the diagnostic application of ionising radiation in Germany. Eur Radiol 1997;7(7):1127–32. [18] Aroua A, Burnand B, Decka I, Vader J-P, Valley J-F. Nation-wide survey on radiation doses in diagnostic and interventional radiology in Switzerland in 1998. Health Phys 2002;83(1):46–55. [19] Webb DV, Solomon SB, Thomson JEM. Background radiation levels and medical exposures in Australia. Radiat Prot Aust 1999;16(2): 25–32. [20] Olerud HM, Saxebol G. Diagnostic radiology in Norway from 1983 to 1993—examination frequency and collective effective dose to patients. Radiat Prot Dosim 1997;74(4):247–60. [21] Ferro de Carvalho A, Vaz Carreiro J. Radiation protection of the patient in Portugal. Radiat Prot Dosim 1995;57(1–4):64–6. [22] Brugmans MJP, Buijs WCAM, Geleijns J, Lembrechts J. Population exposure to diagnostic use of ionizing radiation in the Netherlands. Health Phys 2002;82(4):500–9. [23] Kalmykov L, Pilipenko N, Korneeva V. Collective doses and radiation risks due to medical diagnostic exposures in Ukraine. Radiat Prot Dosim 1997;69(4):275–80. [24] Vano E, Velasco A, Moran P, Gonzalez L, Alvarez Pedrosa CS. Evolution of diagnostic radiology in a big hospital during a 5 year period, and the derived collective dose. Br J Radiol 1993;66(790):892–8. [25] Liu WH, Kai M, Ohta K, Chen FD, Kusama T. Diagnostic medical X-ray examination frequencies in Taiwan. Radiat Prot Dosim 1996;67(3):193–7. [26] Ng KH, Abdullah BJJ, Sivalingam S. Medical radiation exposures for diagnostic radiology in Malaysia. Health Phys 1999;77(1): 33–6.
UK population dose from medical X-ray examinations D. Hart∗ , B.F. Wall National Radiological Protection Board, Chilton, Didcot, Oxon OX11 0RQ, UK Received 14 May 2003; received in revised form 10 June 2003; accepted 11 June 2003
Abstract Objective: To assess the annual per caput and collective effective dose to the United Kingdom population from medical and dental X-ray examinations. Method: The results of a detailed survey of the frequency of X-ray examinations during the financial year 1997/1998 were combined with contemporary data on the effective doses typically received by patients. The resulting per caput and collective dose for 1997/1998 was updated to 2001/2002 by using annual statistics on the total numbers of computed tomography (CT), interventional and conventional examinations collected by the English Department of Health. Results: The annual per caput effective dose for the UK in 2001/2002 was estimated at 0.38 mSv. Over the last 10 years CT has more than doubled its contribution and is now responsible for 47% of the collective dose from medical X-rays. The contribution from conventional radiographic and fluoroscopic examinations has nearly halved to about 34%. Interventional and angiographic procedures together contribute the remaining 19%. Conclusions: The annual per caput effective dose of 0.38 mSv is low in comparison with other countries having similarly developed systems of health-care. This is due to both a lower frequency of X-ray examinations per head of population and generally lower doses in the UK than in other developed countries. © 2003 Published by Elsevier Ireland Ltd. Keywords: Collective dose; Effective dose; Computed tomography; X-ray examinations
1. Introduction Estimates of the population dose (the collective dose, or the average per caput effective dose) provide useful information on the relative contribution of different sources of ionising radiation to population exposure. They can be used to compare the contribution from diagnostic radiology with those from natural or other artificial sources of radiation and to see how the contributions differ between different countries or regions. More specifically, they allow comparison of the contributions from different types of X-ray examination or from different medical imaging modalities. Such information provides guidance on where best to concentrate efforts on dose reduction, so as to optimise the protection of patients in the most cost-effective manner. An estimate by the National Radiological Protection Board (NRPB) in 1999 [1] put the contribution from patients undergoing X-ray examinations at nearly 90% of the total population dose from all artificial sources in the UK, with diagnostic nuclear medicine procedures contributing ∗ Corresponding author. Tel.: +44-123-582-2647; fax: +44-123-583-3891. E-mail address: [email protected] (D. Hart).
0720-048X/$ – see front matter © 2003 Published by Elsevier Ireland Ltd. doi:10.1016/S0720-048X(03)00178-5
a further 8%. (Radiotherapy exposures were deliberately excluded from this analysis). In contrast, all occupational and public exposures arising inadvertently from medical and other uses of ionising radiation, including the UK nuclear power programme, amounted to less than 3% of the total. While dominating the artificial sources of radiation exposure, medical X-rays accounted for only one-sixth of the population dose from natural background radiation. NRPB previously made an assessment of the collective dose from medical and dental X-rays in 1991. The UK collective effective dose equivalent from all diagnostic radiology was at that time roughly estimated at 20 000 man Sv, corresponding to an average per caput dose of 0.35 mSv [2]. NRPB recently made a detailed and more accurate estimate of the collective effective dose to the UK population from medical X-rays for the financial year 1997/1998. A figure of about 19 300 man Sv was derived, corresponding to an average per caput dose of 0.33 mSv [3]. That figure was subject to an estimated uncertainty of ±9%, and confirmed that the previous crude estimate [2] was reasonable. This article updates the detailed population dose estimate to the financial year 2001/2002, using annual statistics on radiology workload and some recent information on patient doses from specific examinations.
286
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
2. Method used to estimate UK population dose for 1997/1998
3. Method used to update the 1997/1998 estimate to 2001/2002
To estimate the annual UK population dose from all medical and dental X-ray examinations, information is required on the annual frequency and the mean effective dose for each type of examination. The former was provided by an NRPB survey of the frequency of 150 types of X-ray examination in the UK in 1997/1998 [4]. Nearly 60% of the hospitals in two of the eight health service regions in England provided detailed examination frequency data and the results were scaled up to the whole of the UK using radiology workload statistics from the Department of Health. Estimates of the mean effective dose for each examination were obtained from a number of sources, the predominant one being the National Patient Dose Database maintained by NRPB [5,6]. The database contained information collected in the period from 1988 to 2000 covering about 60 types of radiograph and 100 types of X-ray examination. Doses are recorded in the National Patient Dose Database mainly as entrance surface dose (ESD) values for individual radiographs and dose–area product (DAP) values for complete examinations. The typical dose for a specific radiograph or examination was taken to be the mean of the doses recorded in the National Patient Dose Database over the whole of the 1990s. Data for the whole decade were used in order to get a sufficiently representative sample size, particularly for the less common examinations. The mean dose for each examination was derived by firstly calculating the mean dose for the sample of patients measured in each radiology room and then taking the mean of these room mean values. In this way equal weight was given to each radiology room in the National Patient Dose Database. A downward trend over time in doses for common radiographic and fluoroscopic examinations, as shown in periodic analyses of the National Patient Dose Database [5,6], means that doses for conventional radiology are probably slightly overestimated for 1997/1998. NRPB report R262 [7] contains generalised conversion coefficients, in Tables 16–18, for estimating effective dose from ESD and DAP measurements, assuming that the X-ray spectra (tube voltage and total filtration) used are close to the average. Typical effective doses were derived from the mean ESD or DAP values using these conversion coefficients. For examinations consisting purely of radiographs, the effective doses from each radiograph were added to provide an effective dose for the complete examination. An effective dose estimate, derived from data in the National Patient Dose Database and conversion coefficients of the type given in NRPB-R262, was obtained for 90 examinations out of the 150. Effective doses for another 25 examinations were found in published literature. This included common computed tomography (CT) examinations, which had been covered by an NRPB survey in 1989 [8]. Effective doses for the remaining 35 examinations were estimated by comparison with similar examinations, for which doses were already available.
The annual frequency of examinations was updated by using radiology workload statistics gathered every year by the English Department of Health, the National Health Service Breast Screening Programme, and the Dental Practice Board. For some examinations, the effective dose is known to have changed, and the population dose was altered proportionately. For those examinations for which no recent dose information was available, the effective dose was assumed to be unchanged from the 1997/1998 estimate. The English Department of Health collects statistics on the number of medical imaging procedures conducted each year by all National Health Service (NHS) hospitals in England. These are known as KH12 returns and have been collected in the current format since 1995/1996 [9]. Non-interventional and interventional procedures are treated as separate categories (the definition of interventional procedures is discussed in an explanatory note at the end of this article). Each of the two categories is broken down into the following modalities: CT, magnetic resonance imaging (MRI), obstetric ultrasound, non-obstetric ultrasound, nuclear medicine, fluoroscopic examinations, and radiographs not involving fluoroscopy. In order to ensure completeness of the statistics, the Department of Health uses the data from previous years to estimate the contribution from the few hospitals that failed to supply actual data in any specific year. These statistics thus give the total number of CT and other X-ray examinations performed each year, and the number of these that are interventional. Any trends in these numbers, therefore, provide a means to update the 1997/1998 annual frequency estimate for medical X-ray examinations to the year 2001/2002. Scotland, Wales and Northern Ireland (with a combined population of 16% of the whole UK) were assumed to have experienced similar trends to England. Mammographic screening, and imaging procedures performed outside NHS hospitals such as dental and chiropractic radiography, are not covered by the KH12 returns. We have attempted to obtain the latest frequencies for these imaging procedures from other sources in order to update their contribution to the population dose. Allowance has also been made for changes in doses for some examinations. The latest review of the National Patient Dose Database (1996–2000) [6] found reductions in doses for common radiographic and fluoroscopic examinations. Therefore, the contribution to population dose from such examinations has been reduced proportionately. It was assumed that doses from CT, interventional and angiographic examinations have remained the same.
4. Results The KH12 returns show that CT, MRI and non-obstetric ultrasound are all steadily increasing in frequency, while
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
287
Table 1 Numbers of interventional and non-interventional procedures in English NHS hospitals
Radiographs CT Fluoroscopy US Non-obstetric US Obstetric MRI Radioisotopes
1997/1998
1998/1999
1999/2000
2000/2001
2001/2002
19 474 590 1 172 656 1 179 979 2 884 883 1 905 649 473 074 722 096
19 876 933 1 254 474 1 244 632 3 068 276 1 950 158 522 138 699 654
19 967 296 1 359 852 1 256 965 3 289 919 1 965 411 585 797 727 255
19 913 022 1 488 752 1 253 847 3 430 753 1 951 829 632 594 539 141
19 806 876 1 625 304 1 222 296 3 579 412 1 992 567 705 706 537 653
Table 2 Numbers of interventional procedures in English NHS hospitals
Fluoroscopy Radiographs CT US Non-obstetric US Obstetric Radioisotopes MRI
1997/1998
1998/1999
1999/2000
2000/2001
2001/2002
217 564 49 723 13 049 70 805 19 072 4830 1422
231 204 51 225 15 374 71 843 16 881 2500 1510
241 802 53 994 17 941 81 904 15 316 3397 1215
253 532 57 357 18 339 84 408 32 756 3865 902
276 197 138 928 31 460 113 286 43 694 16 834 5735
obstetric ultrasound, radiographs, fluoroscopy and nuclear medicine (radioisotopes) have very little growth over recent years (see Table 1). Radiographic examinations are easily the most common of all these procedures. Their frequency rose slightly in the first half of the period 1997/1998 to 2001/2002, but fell in the second half. The number of CT examinations has increased by 39% in the 4 years from 1997/1998 to 2001/2002. The growth in interventional radiology involving fluoroscopy or radiographs was 55% over this period (see Table 2). The number of conventional diagnostic radiology examinations (radiographs and fluoroscopy, not including interventional procedures) has increased by only 1% for the period 1997/1998 to 2001/2002. On the assumption that typical patient doses per procedure have not changed, the 39% growth in frequency of CT examinations boosts the annual collective dose to the UK population for CT from 7662 to 10 650 man Sv for 2001/2002 (see Table 3), and the 55% growth in interventional radiology boosts it from 1239 to 1920 man Sv. As stated above, conventional radiology has increased in frequency by 1% from 1997/1998 to 2001/2002, so this small increase has been applied to all conventional radiology examinations (except mammography and dental radiography) that are listed in Table 3. However, the latest review of the National Patient Dose Database [6] showed there had been a reduction in doses for some common radiographic and fluoroscopic examinations, so this reduction has also been applied in Table 3. The four largest contributors to the UK population dose from conventional radiology are barium enema, abdomen, lumbar spine, and pelvis examinations. Together they contribute about 60% of the total population dose from conventional radiology. In Table 3 their contributions have been reduced by between 7 and 15% to match the
Table 3 Contributions to UK collective dose (man Sv)
CT Interventional Conventional radiology Barium enema Abdomen Lumbar spine Pelvis Mammography Dental Others Angiography Total a
1997/1998a
2001/2002
7662 1239 8473 2588 852 825 644 466 77 3021 1923 19 298
10 650 1920 7720 2379 749 692 559 513 82 2746 2423 22 713
Table 3 of [3].
dose reductions found for these examinations when comparing data for 1996–2000 [6] with the data for 1990–1999 that were used in the 1997/1998 population dose estimate [3]. We have assumed there has been a dose reduction for all other conventional radiology examinations of 10% for the 4-year period 1997/1998 to 2001/2002, in line with the above reductions, and with the latest review of the National Patient Dose Database [6] which found a general reduction of 16% over a 5-year period. The fifth largest contributor to UK population dose from conventional radiology is mammography. The latest data from the National Health Service Breast Screening Programme are for 2000/2001 [11] and show that the number of women being screened has increased to 1.6 million, from 1.4 million in 1997/1998. The breakdown of these figures into different types of examination is shown in Table 4,
288
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
Table 4 Update of UK collective dose from mammography Frequencies
First round of screening Subsequent screening rounds Recall for assessment Symptomatic Total
Collective dose (man Sv)
1997/1998
2000/2001
1997/1998
2000/2001
374 000 960 000 66 000 326 000
377 000 1 169 000 83 000 326 000
138.4 192.0 15.2 120.7 466
139.5 234.0 19.1 120.7 513
where we have assumed that the frequency of symptomatic referrals to hospitals has not changed. We can assume that doses per mammographic procedure have not changed significantly, because the imaging techniques and the number of views taken per breast in specific rounds of the screening programme did not change between 1997 and 2001. The application of the percentage changes in frequency to the collective dose in columns 4 and 5 of Table 4 show that mammography has increased its contribution by 10% to 513 man Sv. This has also been entered into Table 3. Data from the Dental Practice Board [12] show that the number of X-ray examinations carried out by dentists for the NHS in England and Wales has increased by 6% over the period 1997/1998 to 2001/2002. We have assumed the same increase for the private sector, and that the increases apply across the whole of the UK. Assuming no change to the dose per examination, the contribution to the collective dose, therefore, increases from 77 to 82 man Sv as shown in Table 3. It has previously been difficult to get accurate statistics on the annual number of chiropractic X-ray examinations, and this has continued to be the case. Fortunately, the estimated contribution to the 1997/1998 collective dose from chiropractic was only 34 man Sv. Even a major change in the frequency of chiropractic X-ray examinations would thus have little effect on the total UK population dose. Therefore, chiropractic has been assumed to maintain the same absolute contribution to population dose as before. Overall, the annual collective dose from conventional radiology is estimated to have decreased by about 9% from 8473 to 7720 man Sv as shown in Table 3. The KH12 returns give no specific information on angiography. We have, therefore, estimated the annual growth in the frequency of angiography in the UK from 1997 to 2002 to be the same as the annual growth in the years prior to 1997. Table 5 shows the number of angiographic examinations performed in Great Britain [4,13], and France [14] for
the period 1977–2000. Great Britain showed growth rates of 7% per annum in 1977–1983, and 6% p.a. in 1983–1997. France had a growth rate for coronary angiographies of 7% p.a. during 1991–2000, which is very similar to growth rates for Great Britain while covering years which overlap with our update period. It, therefore, seems reasonable to assume that the 6% p.a. increase in the number of angiographies in the UK continued during 1997/1998 to 2001/2002. We have also assumed that the dose per examination has remained constant. An increase of 6% p.a. amounts to a 26% increase over a 4-year period. Consequently, the collective dose from diagnostic angiography of 1923 man Sv estimated for 1997/1998 [3] has been increased to 2423 man Sv for 2001/2002, as shown in Table 3. Taking all the above mentioned increases and decreases into account, the total collective dose in the UK from all X-ray imaging procedures is estimated to have increased from 19 300 man Sv in 1997/1998 to 22 700 man Sv for 2001/2002 as shown in Table 3. The population of the UK had risen by only 1% to 59.8 million by the year 2000, and so the annual average per caput dose from medical X-ray examinations has increased slightly to 0.38 mSv.
Table 5 Annual number of angiography examinations (thousands)
Great Britain Francea a
1977
1983
92.3
135.6
Number of coronary angiographies.
1991
1997/1998
2000
304.0 131.6
236.1
Fig. 1. Contributors to UK population dose from medical X-rays (2001/2002).
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
289
Fig. 2. Biggest contributors to population dose.
5. Discussion CT now represents 7% of all X-ray examinations carried out in NHS hospitals (excluding mammographic screening) but contributes 47% of the total population dose to the UK from medical X-ray examinations. Angiographic procedures are responsible for 11% of the population dose, and interventional procedures for 8%, while conventional radiology is now down to a contribution of 34%. Thus, for the first time, CT has overtaken conventional radiology as a contributor to the UK population dose. The relative contributions of conventional, CT, angiographic and interventional procedures to the population dose from all medical X-ray examinations are shown on the pie-chart in Fig. 1. Fig. 2 shows the 13 examinations that are the biggest contributors to UK population dose, and indicates their percentage contributions to frequency and population dose. Together the 13 examinations account for nearly 80% of the population dose. The large contribution from CT to the UK population dose is comparable to the situation in other countries. UNSCEAR 2000 [15] shows that on average for countries in Health-care
level I, CT represented 6% of all diagnostic medical X-ray examinations, and contributed 41% of the total annual population dose in the period 1991–1996. Mettler et al. [16] calculated that CT scans at one large hospital in the USA during 1998 and 1999 contributed about 67% of the total effective dose from all medical X-ray examinations at that hospital. A comparison of the estimated UK annual per caput dose of 0.38 mSv from medical radiology is made with similar data from the 1990s for other countries in Table 6, using information reported by UNSCEAR [15] and other sources as referenced. The data are arranged in order of decreasing size of the annual per caput effective dose. It can be seen that the UK has a low per caput dose compared with other nations with well-developed systems of health-care. It is notably one-fifth of the value estimated for Germany and just over one-third of the values given for France and Switzerland, although it should be recognised that there are likely to be large uncertainties associated with many of these values. The relatively low value for the UK, especially in comparison with other European countries, appears to be due to
290
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
Table 6 International comparison of annual per caput effective dose from medical radiologya Country
Time period
Annual per caput effective dose (mSv)
Source
Germany France Switzerland Canada Russia Australia Norway Poland Bulgaria Portugal Sweden Romania Netherlands USA Ukraine Finland Spain (regional) UK Denmark Taiwan Brazil Malaysia
1990–1992
1.9 1.0 1.0 0.94 0.9 0.8 0.8 0.8 0.75 0.71 0.68 0.61 0.52 0.5 0.5 0.45 0.4 0.38 0.36 0.23 0.09 0.05
[17] [15] [18] [15] [15] [19] [20] [15] [15] [21] [15] [15] [22] [15] [23] [15] [24] This article [15] [25] [15] [26]
1998
1996 1993
1991
1998 1994 1990 2001/2002 1993 1994
a Based on Table 29 in Annex D, Volume 1, of the UNSCEAR 2000 Report [15].
Table 7 International comparison of examination frequencies and typical effective dosesa Type of examination
Chest Limbs and joints Lumbar spine Thoracic spine Cervical spine Pelvis/hips Head Abdomen Upper GI tract Lower GI tract Cholecystography Urography Mammography CT CT head CT body Angiography Interventional procedures
Number of examinations per 1000 population per year
Typical effective dose (mSv)
UK
Health-care level I
UK
Health-care level I
141 147 19 5 14 31 28 21 4.9 6.1 1.2 4.6 27 21
281 166 48 13 32 35 59 41 42 8.7 3.1 12 25 57
0.02
0.14
1.3 0.7
1.8 1.4
0.7 0.04 0.7 2.6 7.2
0.83 0.07 0.53 3.6 6.4
2.4 0.06
3.7 0.07
2 9
2.3 13.3
5.2 4.5
6.8 3.0
a Selected from Tables 12, 15 and 30 in Annex D, Volume 1, of the UNSCEAR 2000 Report [15].
both a lower frequency of X-ray examinations and generally lower doses per examination. This is evident from Table 7, which shows such statistics for the UK for the 1990s, compared with the average for countries in Health-care level I as reported by UNSCEAR [15]. The second and third columns show the annual numbers of medical X-ray procedures per thousand population for the UK and the average for Health-care level I. For 14 out of the 16 types of procedure shown, UK frequencies are below the average frequencies for Health-care level I. On average, the frequency of examinations in the UK is 40% lower than for Health-care level I. The fourth and fifth columns show typical effective doses to patients for some common types of diagnostic examinations for the UK and the average values for Health-care level I. For ten out of the 12 types of examination for which doses are shown, the UK doses are below the average doses for Health-care level I. On average for the examinations listed, UK effective doses are about 20% lower.
6. Conclusions The annual per caput dose from medical and dental X-ray procedures in the UK for 2001/2002 has been estimated by updating the results of a survey of the frequency of 150 types of examinations in 1997/1998 and combining them with recent estimates of the radiation doses from such examinations. The per caput dose from all X-ray imaging performed in NHS and private sector hospitals and clinics is estimated to be 0.38 mSv for the financial year 2001/2002. This estimate is similar to the previous rough estimate for 1991, and is low in comparison with that for other countries with similarly developed systems of health-care. This is mainly due to a lower frequency of X-ray examinations, but also to generally lower doses per examination in the UK. The relative contributions of some types of examination to the UK population dose from medical X-rays have changed considerably since 1991. CT has more than doubled its contribution and is responsible for 47% of the total. Interventional and angiographic procedures, which often involve prolonged fluoroscopy and hence result in high individual doses, have also increased in frequency and provide 19% of the total. Conventional fluoroscopic and radiographic examinations are making a smaller contribution. This is partly due to a drop in frequency of examinations such as barium meals that are being slowly replaced by endoscopy, and of biliary and urinary tract examinations where ultrasound imaging provides a viable alternative. However, the major factor responsible for reducing the per caput and collective doses for these conventional X-ray examinations is the average drop of about 50% in the dose per examination, seen in the period 1985–2000 and estimated from the latest review of the National Patient Dose Database [6]. These conventional X-ray examinations contribute about 34% of the total population dose, which represents a reduction by about a factor of two since 1991.
D. Hart, B.F. Wall / European Journal of Radiology 50 (2004) 285–291
The increasing attention given in recent years in the UK to radiation protection for conventional examinations, with the development of national patient dosimetry protocols and reference doses, has played a significant part in this substantial reduction in population dose. Widespread local monitoring of patient doses and comparison with national norms have undoubtedly encouraged the adoption of dose-efficient procedures and the introduction of dose-saving features into X-ray imaging equipment. With the now much increased contributions of CT, angiography and interventional radiology to the population dose, there is a clear need to develop radiation protection and optimisation activities for these high dose procedures to the same level as has been already achieved for conventional radiology.
7. Explanatory note The English Department of Health defines interventional procedures in its KH12 form as “any invasive procedure during imaging which is performed other than primarily for the introduction of a contrast medium. Endoscopic Retrograde CholangioPancreatography (ERCP) would count as interventional by virtue of the endoscopy”. This differs from the more widely adopted definition of interventional radiology used by NRPB, which is “minimally invasive procedures which use imaging guidance for therapeutic purposes” [4,10]. Under the latter definition ERCPs are only interventional if they are therapeutic. This differing classification implies up to a 20% larger number of interventional radiology procedures for Department of Health statistics as compared with NRPB data if, in the worst case, no ERCPs in the latter data are therapeutic. This difference over ERCPs will become less noticeable during the coming years, since ERCPs are becoming predominantly a therapeutic procedure. Moreover, any numerical difference should have little effect on the upward or downward percentage change in the number of procedures, which is all that is needed to update 1997/1998 figures to 2001/2002. Acknowledgements We are grateful to Julie Lawlor of the Hospital Activity Statistics Team at the Department of Health, for supplying recent statistics to us in advance of publication. References [1] Hughes JS. Ionising radiation exposure of the UK population: 1999 review. Chilton: NRPB-R311, 1999. [2] Hughes JS, O’Riordan MC. Radiation exposure of the UK population—1993 review. Chilton: NRPB-R263, 1993. [3] Hart D, Wall BF. Radiation exposure of the UK population from medical and dental X-ray examinations. Chilton: NRPB-W4, 2002.
291
[4] Tanner RJ, Wall BF, Shrimpton PC, Hart D, Bungay DR. Frequency of medical and dental X-ray examinations in the UK—1997/1998. Chilton: NRPB-R320, 2000. [5] Hart D, Hillier MC, Wall BF, Shrimpton PC, Bungay DR. Doses to patients from medical X-ray examinations in the UK—1995 review. Chilton: NRPB-R289, 1996. [6] Hart D, Hillier MC, Wall BF. Doses to patients from medical X-ray examinations in the UK—2000 review. Chilton: NRPB-W14, 2002. [7] Hart D, Jones DG, Wall BF. Estimation of effective dose in diagnostic radiology from entrance surface dose and dose–area product measurements. Chilton: NRPB-R262, 1994. [8] Shrimpton PC, Jones DG, Hillier MC, Wall BF, Le Heron JC, Faulkner K. Survey of CT practice in the UK: part 2, dosimetric aspects. Chilton: NRPB-R249, 1991. [9] Acute services team of the performance analysis branch of the NHS Executive. Imaging and radiodiagnostics: England. Leeds: Department of Health (series 1995–1996 to 2001–2002). [10] World Health Organisation. Efficacy and radiation safety in interventional radiology. Geneva: WHO, 2000. [11] National Health Service Breast Screening Programme. Building on experience: breast screening programme annual review 2002. Sheffield: NHS Breast Screening Programme, 2002. http://www. cancerscreening.nhs.uk/breastscreen/publications/2002.review.html. [12] Dental Practice Board http://www.dpb.nhs.uk [13] Wall BF, Hillier MC, Kendall GM. An update on the frequency of medical and dental X-ray examinations in Great Britain—1983. Chilton: NRPB-R201, 1986. [14] Le Tourneau T, Blanchard D, Lablanche JM, Monassier JP, Morice MC, Cribier A, Meier P, Puel J. Evolution of interventional cardiology in France during the last decade (1991–2000) [in French]. Archives des maladies du coeur et des vaisseaux 2002;95(12):1188–94. [15] UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation). Sources and effects of ionising radiation. Volume 1: Sources, Annex D, Medical radiation exposures. New York: United Nations, 2000. [16] Mettler Jr FA, Wiest PW, Locken JA, Kelsey CA. CT scanning: patterns of use and dose. J Radiol Prot 2000;20:353–9. [17] Kaul A, Bauer B, Bernhardt J, Nosske D, Veit R. Effective doses to members of the public from the diagnostic application of ionising radiation in Germany. Eur Radiol 1997;7(7):1127–32. [18] Aroua A, Burnand B, Decka I, Vader J-P, Valley J-F. Nation-wide survey on radiation doses in diagnostic and interventional radiology in Switzerland in 1998. Health Phys 2002;83(1):46–55. [19] Webb DV, Solomon SB, Thomson JEM. Background radiation levels and medical exposures in Australia. Radiat Prot Aust 1999;16(2): 25–32. [20] Olerud HM, Saxebol G. Diagnostic radiology in Norway from 1983 to 1993—examination frequency and collective effective dose to patients. Radiat Prot Dosim 1997;74(4):247–60. [21] Ferro de Carvalho A, Vaz Carreiro J. Radiation protection of the patient in Portugal. Radiat Prot Dosim 1995;57(1–4):64–6. [22] Brugmans MJP, Buijs WCAM, Geleijns J, Lembrechts J. Population exposure to diagnostic use of ionizing radiation in the Netherlands. Health Phys 2002;82(4):500–9. [23] Kalmykov L, Pilipenko N, Korneeva V. Collective doses and radiation risks due to medical diagnostic exposures in Ukraine. Radiat Prot Dosim 1997;69(4):275–80. [24] Vano E, Velasco A, Moran P, Gonzalez L, Alvarez Pedrosa CS. Evolution of diagnostic radiology in a big hospital during a 5 year period, and the derived collective dose. Br J Radiol 1993;66(790):892–8. [25] Liu WH, Kai M, Ohta K, Chen FD, Kusama T. Diagnostic medical X-ray examination frequencies in Taiwan. Radiat Prot Dosim 1996;67(3):193–7. [26] Ng KH, Abdullah BJJ, Sivalingam S. Medical radiation exposures for diagnostic radiology in Malaysia. Health Phys 1999;77(1): 33–6.