British Journal of Radiology 74 (2001),926-931 © 2001 The British Institute of Radiology
Provision of MRI can significantly reduce CT collective dose
J C Clarke, FRCR, FFRRCSI1,
K Cranley, PhD, FIPEM2,
B E Kelly, MD, FRCS, FRCR1,
K Bell, FRCR, FFRRCSI3 and
P H S Smith, BA, DPhil4
Departments of 1Radiology and 3Neuroradiology, Royal Group of Hospitals, Belfast BT12 6BA, 2Northern Ireland Regional Medical Physics Agency, Forster Green Hospital, Belfast BT8 6HD and 4Northern Ireland Regional Medical Physics Agency, Headquarters, Belfast BT2 8BS, UK
 |
Abstract
|
|---|
The purpose of this study was to investigate the proportion of patients currently being investigated by CT that could be investigated by MRI with a potential reduction in exposure to ionizing radiation. The health detriment arising from the radiation dose associated with CT has been quantified in terms of the number of likely cases of serious health effects. The results show that a significant saving in the collective radiation dose is possible, with an associated detriment of between 0.23 and 0.33 cases of cancer or severe hereditary effects averted in one imaging department every year. In selecting the balance of provision of MRI and CT facilities, the health detriment associated with the radiation dose from CT should be considered.
|
Introduction
|
|---|
Increasing use of CT has had a major impact not only on patient care but also on the radiation exposure of the population. In 1989 [1] there were 200 scanners in clinical use in the UK, which accounted for 2% of all radiological examinations but contributed 17% of the collective radiation dose. Today there are more than 370 scanners [2] performing 4% of examinations and contributing 40% of the total radiation dose to the population from medical X-rays. Hall [3] and Mettler et al [4] recently reported that in one study in the USA, the CT contribution to the collective effective dose approaches 70%. In Northern Ireland the number of scanners in use has increased from 5 in 1991 to 13 in the year 2001.
Despite the fact that non-ionizing imaging techniques, including MRI and ultrasound, have greatly advanced, the popularity of CT is steadily increasing. There are many reasons for this [2], even though it is well known that CT makes a major contribution to the radiation exposure of the population. Use of CT can lead to doses per examination that are among the highest delivered within the radiology department.
There are many approaches to reducing patient exposure in CT, including careful selection of technique and consideration of image quality requirements. Helical scanning has the potential to reduce doses if pitch values greater than one are used without increasing the mA–duration product. Careful use of data reconstruction can remove the need for further exposure.
Equipment design has an important effect on patient doses, for example use of carbon fibre tables, tube-to-detector distance and filter combinations. Solid-state detectors with higher detection efficiencies than the xenon gas type provide the potential for simultaneous image quality improvement and patient dose reduction.
However, the greatest potential for reducing patient doses is careful selection of patients for the examination and, if possible, use of a different imaging modality, preferably one that does not employ ionizing radiation.
Following a recent survey of CT doses in Northern Ireland [5, 6], among the recommendations was that the monetary cost of the radiation dose saved by using MRI as an alternative should be calculated and taken into account when policy decisions are made regarding the location of CT and MRI equipment in the Health and Personal Social Services within Northern Ireland.
This paper does not attempt a full cost–benefit analysis, since there is some controversy regarding an appropriate monetary valuation of the detriment associated with collective dose for use in health care economics. Instead, it stops short at estimating the health detriment, and includes modifying factors for paediatric and geriatric patients for whom the detriment per unit dose is higher by a factor of 2 and lower by a factor of 5, respectively, rather than the average for the whole population.
|
Methods
|
|---|
The clinical indications for 1025 CT patients examined in Antrim Hospital over a 3-month period in 1998 were recorded by radiographic staff. Antrim Hospital is a 370-bed district general hospital and was one of the centres included in the Northern Ireland CT Dose Survey. The hospital's CT scanner provides a service to several smaller hospitals within the same Trust, which serves a population of approximately 350 000. The nearest MRI scanner is situated some 20 miles away, but very limited numbers of patients are currently referred for such examinations as a consequence of the waiting lists.
The list of indications for CT scans was then reviewed by a consultant radiologist with an interest in body MRI and a consultant neuroradiologist who indicated which cases in their opinion could have been at least as well investigated by MRI if it were freely available. The preferred imaging method for each patient was recorded. This decision was taken in the context that CT and MRI facilities were readily available and that the radiation dose from CT scanning should be avoided where possible by using an alternative imaging modality that does not use ionizing radiation. Several examination types were considered, and radiologists indicated whether CT or MRI is preferable, or where CT and MRI may be of equal use in the diagnosis.
Dose calculation
The CT survey [5] included all examination types encountered over a different 1-year period for the Antrim Hospital CT scanner. As a result, dose assessment data were available for all the procedures mentioned in Table 1
, except thoracic spine and musculoskeletal examinations. These represent less than 2% of the total number of patients audited and were therefore ignored in the current study.
Two sets of patient totals were produced, one including the "CT or MRI" totals under "CT", and the other combining these figures under "MRI". Consequently, the dose saving results have a range of possible values, as indicated below. Data for calculating the average effective dose per examination were available from the CT dose survey [5]. The annual collective effective dose potentially saved corresponding to the MRI totals in the audit sample was computed for each examination. The proportion of MRI to CT scans was applied to the known annual workload for each procedure. This allowed an estimate of radiation dose saved if MRI had been used in that proportion of examinations of each type. Summing the annual collective effective dose potentially saved by MRI scans allowed an overall proportion of dose saving to be estimated.
Estimation of detriment in diagnostic radiology requires examination of the age profile of the patients exposed. Two situations where the detriment should be modified are for paediatric patients [7] (by applying a factor of 2 to reflect the fact that the number of years lost per unit dose is approximately double that for the average population) and for geriatric patients (by applying a factor of 0.2, which allows for the fact that this age group loses about one-fifth of the years of life lost on average by the general population). These modifying factors reflect the average age at exposure and an estimate of the number of years that are likely for health effects to become manifest. The values for the relevant exposed groups are given in Table 2.
|
Results
|
|---|
There were 496 male patients and 529 females in the audit, with an average age of 58.5 years and an age range of 2 days to 97 years. The number of scans performed for each type of examination and whether CT or MRI would be preferred is shown in Table 1
. In the Appendix, Table A1 gives abbreviated details of the clinical indications for each type of examination. In some instances (such as cardiovascular accident) there is currently some debate as to the best technique. Hence the indications have been listed under three different headings. A practical approach was used. In some cases, such as acute collapse, MRI may be at least as satisfactory but would not be possible in many cases owing to the patient's condition and CT would therefore be preferred for practical reasons.
Between 47% (i.e. 480/1025) and 73% (i.e. [264+480]/1025) of patients currently having CT scans in the centre studied could have been investigated at least as well by MRI (see Table 1). Table 3 provides the potential dose savings and age-corrected risk for the range of examinations. Using the total probability coefficient for stochastic effects for the whole population of 7.3 x 10-2 Sv-1 [8] as well as the appropriate modifying factors in Table 2 for each patient, the number of cases of serious health effects per year for each type of examination was calculated. These are shown in the last column of Table 3 and in total amount to between 0.23 cases and 0.33 cases per year.
These results refer to both the percentage collective effective dose saved due to use of MRI rather than CT and the corresponding potential risk averted. The range arises from the "CT or MRI" column in Table 1, where radiologists considered that the lower number is a definite potential replacement of CT by MRI and the higher figure corresponds to cases where if a CT examination can be avoided, then it is replaced by MRI.
|
Discussion
|
|---|
The CT dosimetry survey [5] did not distinguish between paediatric and adult patients. However, because it is desirable to know to what extent children and geriatric patients are investigated using CT, this retrospective audit included paediatric and geriatric patients. It has been assumed that the radiation doses received by children are on average equal to those received by adults. The effect of this assumption is minimized by the small proportion of children (4.4%) included in this audit. However, it is felt that the proportion of paediatric CT studies is increasing. Mettler et al [4] reported that, over the last decade, use of CT in one major hospital in the USA increased from 6.1% to 11.1%. While 11% of all CT scans were performed on children, a higher proportion than was previously thought, Brenner et al [9] estimated that, relative to adults, paediatric risks are an order of magnitude higher due to (i) larger doses at the same mAs and (ii) an increased lifetime risk. Brenner et al [9] also proposed that lower mAs settings can be used in paediatric CT studies without significant loss of diagnostic information. While the number of paediatric examinations is small, their importance in terms of collective dose and the potential saving from use of other modalities is doubled compared with non-geriatric patients (Table 2).
While the number of geriatric patients is significant (35.3%), a greater effect on the results, the weighting by a factor of 0.2 significantly reduces the potential annual savings (see Table 2).
The CT survey [5] provided data about the Antrim Hospital CT scanning protocols and dosimetry for a typical adult patient. The sample of patients in the current audit included a number of paediatric patients for whom the typical adult protocol would have been varied significantly. However, the fact that protocol differences for these small patients were ignored makes no significant difference to the conclusions because the numbers are very small, particularly when the assumption that every patient is of average adult size is made. Nevertheless, the risk has been calculated using age-related corrections to the three age groups (paediatric, below 16 years; general, 16–69 years inclusive; and geriatric, above 70 years), since the relationship between exposure and detriment has been demonstrated to be age dependent [7].
It is not possible to draw up an exhaustive set of guidelines to cover every possible clinical scenario and not every radiologist will agree with our choice of scan in every situation. We do, however, believe that the choice of technique is a fair reflection of current practice.
The indications recorded in this paper are a reflection of actual practice and should not be regarded as guidelines for the use of CT.
|
Conclusions
|
|---|
In the Joint Working Party report of the Royal College of Radiologists and the National Radiological Protection Board [1] it was stated that the elimination of clinically unhelpful examinations had the greatest potential for dose reduction (20% of the collective effective dose at that time due to medical examinations). Increasing the availability of both MRI and ultrasound, particularly for young patients, was also recommended [1]. A decade later, the same principles apply, and it appears that the alternative imaging modality of MRI can increasingly be clinically justified in many routine procedures. The current study confirms that MRI indeed has a great potential to affect patient doses in diagnostic radiology. It is apparent that the radiation dose to the population could be substantially reduced if there was increased availability of MRI.
|
Appendix
|
|---|
Notes to Table A1
View this table:
[in this window]
[in a new window]
|
A1 Detailed break-down of clinical indications (see Table 1). Numbers beside the clinical indications represent the respective number of patients found in the audit
|
|
(1) CT is much better at detecting fresh intracranial haemorrhage and therefore remains the investigation of choice for head injury, and subdural and subarachnoid haemorrhage. However, MRI is considered to be superior to CT for the evaluation of posterior fossa lesions [10] and usually provides at least as much information about space-occupying lesions as CT [11] without exposure to ionizing radiation and would therefore be preferred in such cases if it were available.
It is now generally accepted that MRI should be used to investigate pituitary microadenomas. CT is still commonly used for larger lesions but has no real advantage over MRI and it was felt that all cases of suspected pituitary pathology in this series could have been satisfactorily investigated by MRI.
(2) The ability of CT to demonstrate bony detail means that it remains the examination of choice for the assessment of orbital trauma and chronic middle ear disease and for demonstrating the anatomy of the drainage of the sinuses prior to functional endoscopic sinus surgery [12].
(3) It is now generally accepted that MRI should be used whenever possible to investigate acoustic neuroma [13].
(4) In the neck, the multiplanar capability of MRI and its better soft tissue discrimination is advantageous in staging most ENT neoplasms.
(5) MRI should be used in most instances for spinal imaging, although CT is superior for the assessment of fracture fragments.
(6) In general, MRI was preferred to CT for assessment of musculoskeletal trauma but there are specific situations, such as pelvic or calcaneal fractures, where CT is superior.
(7) CT is preferred to MRI for the vast majority of chest examinations in view of its ability both to provide superb demonstration of pulmonary parenchymal abnormalities and to give excellent information about the mediastinum [14]. Although MRI has been advocated for acute aortic dissection [15], it is our opinion that CT should still be used in this situation because the examination is quicker and it is easier to monitor the patient. MRI should, however, be used for chronic aortic dissection and for the assessment of thoracic aneurysms in view of its ability to provide direct coronal images. CT was also preferred for the investigation of rib lesions in general, but it is acknowledged that MRI may have a role in specific cases.
(8) Indications for MRI in the abdomen are growing, but the ability of CT to demonstrate all the organs, including the bowel, gives it the edge in many clinical situations. CT at present remains the technique of choice for staging or follow-up of intra-abdominal malignancy [16], most pancreatic or adrenal lesions, suspected leaking aneurysms, undiagnosed abdominal masses or pyrexia of unknown origin.
There is currently controversy as to whether CT or MRI is the optimum examination for liver metastases, but it is generally acknowledged that MRI cholangiography is superior to CT [17].
(9) MRI is now the examination of choice for staging most pelvic malignancies owing to its multiplanar capability [18]. CT would, however, be preferred for trauma or suspected abscess formation.
|
Acknowledgments
|
|---|
The authors are grateful to Dr Stephen McKinstry, Consultant Neuroradiologist, for his valuable assistance and advice.
Received for publication March 1, 2000.
Accepted for publication May 14, 2001.
|
References
|
|---|
-
Royal College of Radiologists/National Radiological Protection Board. Patient dose reduction in diagnostic radiology. Documents of the NRPB 1992;Vol. 1(3).
-
Shrimpton PC, Edyvean S. CT scanner dosimetry. Br J Radiol 1998;71:C1–3.
-
Hall EJ. CT scanning; risk versus benefit. J Radiol Prot 2000;20:347–8.[Medline]
-
Mettler FA, Wiest PW, Locken JA, Kelsey CA. CT scanning: patterns of use and dose. J Radiol Prot 2000;20:353–9.[Medline]
-
DHSS (NI). Report of a CT radiation dose survey in Northern Ireland. Clinical Imaging Services Advisory Committee—Computed Tomography Working Party, June 1998.
-
Clarke J, Cranley K, Robinson J, Smith PHS, Workman A. Application of draft European Commission reference levels to a regional CT dose survey. Br J Radiol 2000;73:43–50.[Abstract]
-
Shrimpton PC, Wall BF, Croft JR, Webb GAM. Medical exposure. Documents of the NRPB 1993;Vol. 4(2).
-
ICRP. Recommendations of the International Commission on Radiological Protection. Annals of the ICRP 1990;Vol. 21 Nos 1–3.
-
Brenner DJ, Elliston CD, Hall EJ, Berdon WE. Estimated risks of radiation-induced fetal cancer from pediatric computed tomography. AJR 2001;176:289–96.
-
Teasdale GM, Hadley DM, Lawrence A, et al. Comparison of magnetic resonance imaging and computed tomography in suspected lesions in the posterior cranial fossa. BMJ 1989;299:349–55.
-
Armstrong P, Keevil SF. Magnetic resonance imaging—2: clinical uses. BMJ 1998;1;303:105–9.
-
Zinreich SJ, Kennedy DW, Rosenbaum AE, et al. Paranasal sinuses: CT imaging requirements for endoscopic surgery. Radiology 1987;163:769–75.[Abstract/Free Full Text]
-
Royal College of Radiologists. Making the best use of a department of clinical radiology: guidelines for doctors (4th edn). London: Royal College of Radiologists, 1998.
-
Dixon AK. The appropriate use of computed tomography. Br J Radiol 1998;70:S98–S105.
-
Panting JR, Norrell MS, Baker C, Nicholson AA. Feasibility, accuracy and safety of magnetic resonance imaging in acute aortic dissection. Clin Radiol 1995;50:455–8.[Medline]
-
Fukuya T, Honda H, Hayashi T, et al. Lymph node metastases: efficiency of detection with helical CT in patients with gastric cancer. Radiology 1995;197:705–11.[Abstract/Free Full Text]
-
Guibaud L, Bret P, Reinhold C, et al. Bile duct obstruction and choledocholithiasis: diagnosis with MR cholangiography. Radiology 1995;197:109–15.[Abstract/Free Full Text]
-
Mayo-Smith WW, Lee MJ. MR imaging of the female pelvis. Clin Radiol 1995;50:667–76.[Medline]
Top
Abstract
Introduction
