- Email: [email protected]
Determining the Rate of Change in Exposure to Ionizing Radiation From CT Scans: A Database Analysis From One Hospital
Determining the Rate of Change in Exposure to Ionizing Radiation From CT Scans: A Database Analysis From One Hospital Michael F. Rayo, PhDa, Emily S. Patterson, PhDa, Beth W. Liston, MD, PhDb, Susan White, PhDa, Nina Kowalczyk, PhDa
Purpose: Cancer risks associated with radiation from CT procedures have recently received increased attention. An important question is whether the combined impact of CT volume and dose reduction strategies has reduced radiation exposure to adult patients undergoing CT examinations. The aim of this study was to determine differences in radiation exposure from 2008 to 2012 to patients receiving CT scans of the abdomen, head, sinus, and lumbar spine at a midwestern academic medical center that implemented dose reduction strategies. Methods: Data were collected from two internal data sets from 2008 to 2012 for general medicine and intensive care unit patients. These data were used to calculate annual CT volume, rate, average effective dose, radiation exposure, and estimated cancer risk. Results: A 37% reduction in abdominal CT volume was found from 2008 to 2012. However, no volume reductions were found for CT examinations of the head or lumbar spine, and the decrease in sinus imaging was minimal. Dose reduction strategies resulted in 30% to 52% decreases in radiation exposure for the targeted body areas. The combined reduction in volume and dose per procedure reduced estimated induced cancers by 63%. Conclusions: Exposure to ionizing radiation from these examinations was reduced at one institution because of reduced volumes of procedures and the reduction of each procedure’s effective dose through new protocols and technologies. Although both the volume reduction and dose reduction strategies contributed to the reduced exposure, it seems that investments in implementing the protocols and new technology had the greatest effect on future cancer risk. Key Words: Radiation, CT, volume, dose reduction strategies, cancer J Am Coll Radiol 2014;-:---. © 2014 Published by Elsevier Inc. on behalf of American College of Radiology
INTRODUCTION
The use of CT imaging in the United States steadily increased from 1998 through 2008. Studies of the Medicare population demonstrated a 10.1% increase in CT volume from 1998 to 2005 [1,2]. Although the increase slowed to 5.1% from 2005 to 2008, a decrease in CT volume was not noted until 2009. From 2009 to 2010, studies of the Medicare population demonstrated a 1.7% decrease in CT scans. The decrease for hospital inpatients on Medicare during this time period was even
a
School of Health and Rehabilitation Sciences, College of Medicine, The Ohio State University, Columbus, Ohio.
b
College of Medicine, The Ohio State University, Columbus, Ohio.
Corresponding author and reprints: Michael F. Rayo, PhD, The Ohio State University, College of Medicine, School of Health and Rehabilitation Sciences, 37 W Kenworth Road, Columbus, OH 43214; e-mail: rayo.3@osu. edu. ª 2014 Published by Elsevier Inc. on behalf of American College of Radiology 1546-1440/14/$36.00 http://dx.doi.org/10.1016/j.jacr.2013.12.005
more pronounced, with CT volumes reduced by 4.5% [1,2]. In an effort to reduce patient radiation dose, health care providers have invested in technological interventions such as computer-based programs to reduce the ordering of inappropriate CT imaging [3,4]. In addition, CT equipment innovations such as tube current modulation [5,6], dual-energy imaging [6], and adaptive statistical iterative reconstruction techniques [6-8] and protocols such as weight-based variable kilovoltage selection and focused collimation produce comparable or better image quality with a reduced effective radiation dose [5,6]. The increased effort to decrease effective radiation dose is commonly attributed to 3 articles published between 2007 and 2009 [9-11]. Brenner and Hall [9] estimated that as many as 2% of all US cancers could be attributed to CT radiation exposure. Berrington de González et al [10] predicted that 29,000 future US cancers could be caused by the annual CT imaging 1
2 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
volume, and Smith-Bindman et al [11] suggested that 1 in 270 women undergoing CT coronary angiography may develop cancer from that procedure. Although the assumptions underlying these estimates were disputed by members of the imaging community [12-16], these articles caught the attention of the national news media [17-19], prominent health institutions such as Harvard Medical Center [20], and the FDA. Another potential cause of reduced ionizing radiation exposure was the Deficit Reduction Act of 2005 and other similar federal legislation that began reducing Medicare payments for advanced imaging beginning in 2007 [21]. Because this legislation affects only Medicare patients, it is unclear whether the observed reductions are predictive of a decrease in CT volume in the overall US patient population. In addition, it is unclear whether recent protocol and technology changes implemented for both Medicare and non-Medicare patients have affected individual and population cancer risk. As such, the purpose of this study was to determine the impact of changes in CT procedure volumes and dose reduction interventions on possible future cancer risk for the inpatient population at an academic medical center. METHODS
CT procedure volumes were obtained from an electronic database capable of providing deidentified patient information available for research purposes. Data were collected for calendar years 2008 to 2012. This study focused primarily on the inpatient population, so data were collected only from patients located in the medical intensive care unit and general medicine units. Procedure volumes for CT of the abdomen and pelvis, head, sinus, and lumbar spine were collected. Current Procedural Terminology, version 4, codes corresponding to the abdominal and pelvic codes were 72192, 72193, 72194, 74150, 74160, 74170, 74176, 74177, and 74178. Head CT codes were 70450, 70460, and 70470. Sinus CT codes were 70486, 70487, and 70488. Lumbar spinal CT codes were 72131 and 72132. Annual CT procedure volumes for each body area and for all body areas combined were
reported. The CT imaging rate was then determined by dividing CT imaging volume by the number of patient admissions for the medical intensive care unit and general medicine units. Because the data analyzed included the entire population under study for the selected years, comparative statistics were not used to determine differences. A subsequent analysis of inpatient abdominal ultrasound procedure volumes from 2008 to 2012 was conducted using the same methods and sources to help explain the findings on abdominal CT procedure volume. To calculate average effective dose in order to determine the effectiveness of the dose reduction strategies, the average dose-length products (DLP) for 2010 and 2012 were multiplied by tissue weighting factors supplied by the American Association of Physics in Medicine [22]. The average DLP and associated standard deviation for each body area was calculated for 2010 and 2012; those years were specifically selected to represent preintervention and postintervention data because all dose reduction strategies were implemented in 2011. To calculate the DLP, patient records from the institution’s PACS were randomly selected for each desired year and procedure. Data from 100 patients in each year were recorded for abdominal and pelvic CT. Eighty-seven patients in each year were sampled for head CT. For lumbar spinal CT, a smaller sample size of 10 patients was used to be proportional to the smaller annual procedure volume relative to the first 2 body areas. Insufficient data were available to perform this analysis for sinus CT. In addition, predicted induced cancers and related mortalities from these 3 body areas were calculated on the basis of estimates published in the BEIR VII report [23]. This study was approved by the institutional review board. RESULTS
Table 1 and Figure 1 show the inpatient CT procedure volumes from 2008 to 2012 for the 4 targeted body areas. Total imaging volume for CT imaging of these 4 body areas increased by 21% from 2008 to 2010 and then decreased by 30% from 2010 to 2012, for a net
Table 1. Annual inpatient CT imaging volumes and rates for the abdomen and pelvis, head, sinus, and lumbar spine from 2008 to 2012 Annual Volume (Rate per 100 Admitted Patients) Body Area 2008 2009 2010 2011 2012 Abdomen/pelvis Head Sinus Lumbar Total volume Total admissions
2,481 (39) 1,606 (25) 84 (1) 69 (1) 4,240 6,333
2,824 (42) 2,116 (31) 79 (1) 72 (1) 5,091 6,749
3,099 (42) 1,846 (25) 91 (1) 91 (1) 5,127 7,314
2,362 (30) 2,005 (25) 107 (1) 68 (1) 4,542 7,949
1,565 (17) 1,866 (20) 83 (1) 75 (1) 3,589 9,104
Rayo et al/Rate of Change in Exposure to Ionizing Radiation 3
Fig 1. Annual inpatient CT imaging volumes and rates for the abdomen and pelvis, head, sinus, and lumbar spine from 2008 to 2012.
decrease of 15% over the 5-year period. This same trend was seen individually for abdominal and pelvic imaging, which also increased from 2008 to 2010, then decreased from 2010 to 2012 for a 37% net decrease. The CT procedure volumes for imaging of the head, sinus, and lumbar spine all fluctuated year to year from 2008 to 2012. Over the 5-year span, sinus CT imaging decreased by 1%, yet head and lumbar spinal CT imaging increased by 16% and 9%, respectively. The CT imaging rate per admitted patient is shown in Table 1. Rates for abdominal and pelvic and head imaging rose from 2008 to 2009, then remained stable or decreased each year through 2012, for net decreases of 56% and 19%, respectively. Sinus and lumbar spinal CT imaging rates remained relatively stable at 1% for the 5-year period. Inpatient abdominal ultrasound imaging volume and rate are shown in Table 2. Procedure volume increased by 71% from 2008 to 2009 (from 316 to 542 procedures) and remained relatively stable through 2012. The rate of ultrasound procedures increased by 61% from 2008 to 2009 and declined from 2009 to 2012. Table 3 shows the average effective doses for 2010 and 2012 for CT procedures of the abdomen, head, and lumbar spine. The effective radiation dose was reduced from 14.7 mSv (95% confidence interval [CI], 13.1e16.3 mSv) to 7.1 mSv (95% CI, 6.4e7.7 mSv) for abdominal CT procedures. The effective dose attributed to head CT procedures was reduced from Table 2. Annual inpatient abdominal ultrasound imaging volume and rate from 2008 to 2012 Annual Volume Year (Rate per 100 Admitted Patients) 2008 2009 2010 2011 2012
316 542 484 574 603
(5.0) (8.0) (6.6) (7.2) (6.6)
2.8 mSv (95% CI, 2.6e2.9 mSv) to 2.0 mSv (95% CI, 2.0e2.0 mSv). Lumbar spinal procedure effective dose was also reduced from 20.2 mSv (95% CI, 15.8e24.6 mSv) to 14.2 mSv (95% CI, 10.6e17.8 mSv). During this time period, the academic medical center implemented 4 different dose reduction strategies. These strategies included the creation of enterprise-wide standardized imaging protocols enforcing dose uniformity on the basis of the recent literature [5,6], the use of tube current modulation [5,6,24], the use of variable weight-based kilovoltage [5,6], and the use of adaptive statistical iterative reconstruction techniques [6,8]. Table 4 compares 2010 and 2012 CT imaging of the abdomen, head, and lumbar spine. Effective dose was calculated as the product of the average effective dose and CT volumes reported. Between those 2 years, the average effective dose per patient was reduced by >60% (from 23.2 to 8.6 mSv). The number of CT procedures for patients undergoing at least one CT scan was reduced from 2.2 to 1.4. The number of patients who received a cumulative annual dose of 50 mSv, which is the limit for occupational exposure [25], decreased from 10% to 0.2%. The total exposure of the population was reduced from 53,281 to 19,971 mSv. After applying the findings from the BEIR VII report [23], this dose reduction resulted in decreases of both estimated induced cancer cases (from 10.1 to 3.8) and resultant mortalities (from 5.1 to 1.9). DISCUSSION
Recent studies [2,26] have shown that the volume and rate of advanced imaging studies have recently been decreasing. This is true for both CT and MR imaging among Medicare recipients. This study shows a trend similar to those studies. This is notable because the scope of this study is different: its focus is not just Medicare recipients but all general medicine and intensive care unit patients at an academic medical center receiving 1 of these 4 procedures. The same rationale noted by the previous studies [2,26] can be used to explain this study’s overall findings. Increased attention to radiation concerns from the news media and medical journals may have led referring physicians to use more stringent criteria for when to order advanced imaging. Cost-conscious physicians may have opted for less expensive but sufficient imaging alternatives, or no imaging at all, to aid their diagnoses. Physicians may have paid more attention to the appropriateness criteria published by the ACR [27] and other organizations. The publicity generated from programs such as Image Wisely [28] may have reduced overuse. Commercial payers’ requirement of prior authority for advanced imaging may have reduced the number of inappropriate tests. Last, the recession may have affected utilization, as many commercial payers require some out-of-pocket
4 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
Table 3. Average effective dose for abdominal and pelvic, head, sinus, and lumbar spinal CT at an academic medical center for 2010 and 2012 Average Effective Dose (mSv) (95% CI) Body Area 2010 2012 Interventions Abdomen/pelvis Head Lumbar spine
14.7 (13.1e16.3) 2.8 (2.6e2.9) 20.2 (15.8e24.6)
7.1 (6.4e7.7) 2.0 (2.0e2.0) 14.2 (10.6e17.8)
DU, WB, IR DU, IR DU, WB, IR
Note: CI ¼ confidence interval; DU ¼ dose uniformity between scanners; IR ¼ iterative reconstruction of resultant imaging; WB ¼ weight-based current and voltage modifications.
costs for advanced imaging. In addition to these explanations, a change in the default order sets in the institution’s electronic health record reduced the number of imaging procedures automatically ordered for most clinical indications. However, changes in head, sinus, and lumbar spinal CT volumes did not follow the trends noted in these national studies, and changes in CT imaging rates varied widely among the 4 body areas analyzed in this study. These findings may be better explained by the relative opportunity to decrease inappropriateness imaging, as documented by the ACR [27], for these body areas. The increase in inpatient head CT volume and smaller rate decrease relative to abdominal imaging is likely due to CT’s being the most appropriate modality for imaging head traumas, hearing loss and vertigo, and severe and sudden headache symptoms. In addition, CT was found to be the preferred modality at this institution because of the speed and accuracy of the test. The insubstantial volume changes and static imaging rates of sinus and lumbar spinal CT may be explained by the lack of opportunity to reduce inappropriate imaging for these procedures. For some time, imaging of the paranasal sinuses has been discouraged unless the patient is immunodeficient, has a neurologic deficit, or experiences chronic, recurrent sinusitis. Lumbar spinal imaging is also often discouraged, with MR as a consistently preferred modality when patient comorbidities make advanced imaging appropriate. Surprisingly, abdominal CT volume decreases were not associated with meaningful abdominal ultrasound procedure volume increases. We expected to see evidence of prescribing behaviors changing in favor of abdominal ultrasound, as it are less expensive, portable, Table 4. Comparison of 2010 and 2012 CT imaging statistics for patients undergoing at least one abdominal, head, sinus, or lumbar spinal CT scan Variable 2010 2012 Number of patients undergoing CT Average effective dose/patient (mSv) Number of CT studies/patient Percentage of patients receiving an effective dose >50 mSv Total radiation exposure (mSv)
2,321 23.2 2.2 10
2,340 8.6 1.4 0.2
53,281
19,971
and often the most appropriate imaging alternative to CT for the abdomen [27]. What we found was that, after both abdominal CT and ultrasound volumes increased from 2008 to 2009, abdominal ultrasound volumes plateaued, fluctuating year to year for a net increase of 61 procedures (11%) between 2009 and 2012, which is less than one-tenth of the decrease in abdominal CT volume during the same time period. It is likely that advanced abdominal imaging volume in general decreased for this patient population. The implemented dose-reduction strategies seem to be the main contributor to the reductions in effective dose per patient (from 23.2 to 8.6 mSv), the percentage of patients with annual effective dose >50 mSv (from 10% to 0.2%), predicted induced cancers (from 10.1 to 3.8), and predicted cancer deaths (from 5.1 to 1.9). These improvements were also primarily responsible for the almost 100-fold decrease in patients receiving a cumulative annual radiation dose of >50 mSv. This is significant as dose reductions for these patients have been predicted to have a disproportionately high effect on the overall cancer risk [29]. The finding that the dose restriction strategies had more impact than CT volume decreases in reducing the patients’ effective dose is consistent with the conclusion drawn by others that imaging technology manufacturers may play the largest role in making patients safer [30]. These new technologies provide new imaging opportunities [5-8] and facilitate the implementation of new protocols [5,6] that have been found to reduce the effective radiation dose to patients without compromising image quality. It is unclear whether the decrease in observed abdomen CT volume, the relatively static volumes of sinus and lumbar spinal CT volume, or the year-toyear fluctuations in the head CT data provide adequate foreshadowing for future CT volume trends. Another possible predictor, decreasing CT imaging rate for abdomen and head CT, seems to indicate increasingly conservative prescribing behavior for referring physicians and radiologists. Whatever the CT volume and rate may be in the future, however, the dose reduction strategies used at this institution have greatly reduced the predicted cancer risk to this patient population.
Rayo et al/Rate of Change in Exposure to Ionizing Radiation 5
These findings suggest that although efforts should continue in trying to reduce inappropriate diagnostic imaging for both cost and patient safety reasons, implementing protocol and technology improvements may be more effective in reducing patients’ effective radiation dose and estimated future cancer risk. This study was limited because it was constrained to the inpatient patient population of one midwestern tertiary care academic medical center. The DLPs reported were the result of an average taken by randomly sampling procedures, not of the actual DLP recorded for each procedure in the data set. Effective doses reported were the product of these DLPs and an estimated tissue weighting, not actual cell damage to the patient. We feel that our method of estimation is justified, however, given the small confidence intervals that resulted and the large reductions in dosage after the interventions. Another limitation is that because of the data available for this study, the relative effect size of each dose reduction strategy is unclear. It was also unclear whether the reduction in procedure volume and average radiation dose were due to increased scrutiny of appropriateness or to other factors, including differences in ordering protocols or technology. These unanswered questions are promising areas for future research. Last, the cancer risk predictions were based on the linear no-threshold model, which has substantial supporting evidence [22] but has not been proved. TAKE-HOME POINTS
A 15% net decrease in total annual CT procedures from 2008 to 2012 at an urban tertiary-care hospital was driven primarily by the 37% decrease in abdominal imaging. CT volumes for the head, sinus, and lumbar spine fluctuated slightly from year to year but remained relatively unchanged after the 5-year period. Reductions in CT rate from 2008 to 2012 per patient seem to indicate a more conservative prescribing behavior among referring physicians and radiologists. Dose reduction strategies, including new protocols and technology enhancements, resulted in 30% to 52% dose reductions. The combination of CT volume reduction and dose reduction for each procedure reduced the predicted risk for induced cancer and cancer deaths by 63%. The percentage of patients receiving an annual effective radiation dose of 50 mSv was reduced from 10% to 0.2%. Investments in new protocols and technology enhancements were more effective than reduced CT volume in reducing patients’ effective radiation dose and associated future cancer risk.
REFERENCES 1. Levin DC, Rao VM, Parker L, Frangos AJ, Sunshine JH. Bending the curve: the recent marked slowdown in growth of noninvasive diagnostic imaging. AJR Am J Roentgenol 2011;196: W25-9. 2. Levin DC, Rao VM, Parker L. The recent downturn in utilization of CT: the start of a new trend? J Am Coll Radiol 2012;9:795-8. 3. Vartanians VM, Sistrom CL, Weilburg JB, Rosenthal DI, Thrall JH. Increasing the appropriateness of outpatient imaging: effects of a barrier to ordering low-yield examinations. Radiology 2010;255:842-9. 4. Blackmore CC, Mecklenburg RS, Kaplan GS. Effectiveness of clinical decision support in controlling inappropriate imaging. J Am Coll Radiol 2011;8:19-25. 5. Goldman AR, Maldjian PD. Reducing radiation dose in body CT: a practical approach to optimizing CT protocols. AJR Am J Roentgenol 2013;200:748-54. 6. Maldjian PD, Goldman AR. Reducing radiation dose in body CT: a primer on dose metrics and key CT technical parameters. AJR Am J Roentgenol 2013;200:741-7. 7. Willemink MJ, Leiner T, Jong PA, et al. Iterative reconstruction techniques for computed tomography part 2: initial results in dose reduction and image quality. Eur Radiol 2013;23:1632-42. 8. Han BK, Grant KLR, Garberich R, Sedlmair M, Lindberg J, Lesser JR. Assessment of an iterative reconstruction algorithm (SAFIRE) on image quality in pediatric cardiac CT datasets. J Cardiovasc Comput Tomogr 2012;6:200-4. 9. Brenner DJ, Hall EJ. Computed tomography—an increasing source of radiation exposure. N Engl J Med 2007;357:2277-84. 10. Berrington de González A, Mahesh M, Kim K-P, et al. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med 2009;169:2071-7. 11. Smith-Bindman R, Lipson J, Marcus R, et al. Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Arch Intern Med 2009;169:2078-86. 12. Cameron JR. Link between diagnostic X-rays and cancer uncertain. Evidence-Based Healthc 2004;8:205-6. 13. Mezrich R. Are CT scans carcinogenic? J Am Coll Radiol 2008;5: 691-3. 14. Mezrich RS. Cancer risk from multiple imaging tests. JAMA 2011;305: 887-8. 15. Budoff MJ, Gupta M. Radiation exposure from cardiac imaging procedures: do the risks outweigh the benefits? J Am Coll Cardiol 2010;56: 712-4. 16. Gerber TC, Gibbons RJ. Weighing the risks and benefits of cardiac imaging with ionizing radiation. JACC Cardiovasc Imaging 2010;3: 528-35. 17. Guthrie C. How dangerous are CT scans? Time. June 27, 2008. Available at: http://content.time.com/time/health/article/0,8599,181852 0,00.html. Accessed December 27, 2013. 18. Shulman M. Facing a CT scan? Think about radiation. US News and World Report. May 30, 2008. Available at: http://health. usnews.com/health-news/managing-your-healthcare/cancer/articles/ 2008/05/30/facing-a-ct-scan-think-about-radiation. Accessed December 27, 2013. 19. Knox R. Radiation from CT scans may raise cancer risk. National Public Radio. December 15, 2009. Available at: http://www.npr.org/ templates/story/story.php?storyId¼121436092. Accessed December 27, 2013. 20. Robb-Nicholson C. A doctor talks about: radiation risk from medical imaging. Harvard Health Publications. October 2010. Available at:
6 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
http://www.health.harvard.edu/newsletters/Harvard_Womens_Health_ Watch/2010/October/radiation-risk-from-medical-imaging. Accessed December 27, 2013.
25. Nuclear Regulatory Commission. Occupational dose limits for adults (10 CFR 20.1201). Available at: http://www.nrc.gov/reading-rm/doccollections/cfr/part020/part020-1201.html. Accessed July 17, 2013.
21. Levin DC, Rao VM, Parker L, Frangos AJ. The sharp reductions in Medicare payments for noninvasive diagnostic imaging in recent years: will they satisfy the federal policymakers? J Am Coll Radiol 2012;9:643-7.
26. Sharpe RE Jr, Levin DC, Parker L, Rao VM. The recent reversal of the growth trend in MRI: a harbinger of the future? J Am Coll Radiol 2013;10:599-602.
22. McCollough C, Cody D, Edyvean S, Geise R, Gould B. The measurement, reporting, and management of radiation dose in CT. College Park, Maryland: American Association of Physicists in Medicine; 2008. 23. National Research Council. BEIR VII: health risks from exposure to low levels of ionizing radiation. Washington, District of Columbia: National Academies Press; 2005. 24. McCollough CH, Primak AN, Braun N, Kofler J, Yu L, Christner J. Strategies for reducing radiation dose in CT. Radiol Clin North Am 2009;47:27-40.
27. American College of Radiology. ACR Appropriateness Criteria. Available at: http://www.acr.org/Quality-Safety/Appropriateness-Criteria. Accessed July 17, 2013. 28. Brink JA, Amis ES. Image Wisely: a campaign to increase awareness about adult radiation protection. Radiology 2010;257:601-2. 29. Miglioretti DL. The use of computed tomography in pediatrics and the associated radiation exposure and estimated cancer risk. JAMA Pediatr 2013;167:700-7. 30. Rehani MM. Challenges in radiation protection of patients for the 21st century. AJR Am J Roentgenol 2013;200:762-4.
Purpose: Cancer risks associated with radiation from CT procedures have recently received increased attention. An important question is whether the combined impact of CT volume and dose reduction strategies has reduced radiation exposure to adult patients undergoing CT examinations. The aim of this study was to determine differences in radiation exposure from 2008 to 2012 to patients receiving CT scans of the abdomen, head, sinus, and lumbar spine at a midwestern academic medical center that implemented dose reduction strategies. Methods: Data were collected from two internal data sets from 2008 to 2012 for general medicine and intensive care unit patients. These data were used to calculate annual CT volume, rate, average effective dose, radiation exposure, and estimated cancer risk. Results: A 37% reduction in abdominal CT volume was found from 2008 to 2012. However, no volume reductions were found for CT examinations of the head or lumbar spine, and the decrease in sinus imaging was minimal. Dose reduction strategies resulted in 30% to 52% decreases in radiation exposure for the targeted body areas. The combined reduction in volume and dose per procedure reduced estimated induced cancers by 63%. Conclusions: Exposure to ionizing radiation from these examinations was reduced at one institution because of reduced volumes of procedures and the reduction of each procedure’s effective dose through new protocols and technologies. Although both the volume reduction and dose reduction strategies contributed to the reduced exposure, it seems that investments in implementing the protocols and new technology had the greatest effect on future cancer risk. Key Words: Radiation, CT, volume, dose reduction strategies, cancer J Am Coll Radiol 2014;-:---. © 2014 Published by Elsevier Inc. on behalf of American College of Radiology
INTRODUCTION
The use of CT imaging in the United States steadily increased from 1998 through 2008. Studies of the Medicare population demonstrated a 10.1% increase in CT volume from 1998 to 2005 [1,2]. Although the increase slowed to 5.1% from 2005 to 2008, a decrease in CT volume was not noted until 2009. From 2009 to 2010, studies of the Medicare population demonstrated a 1.7% decrease in CT scans. The decrease for hospital inpatients on Medicare during this time period was even
a
School of Health and Rehabilitation Sciences, College of Medicine, The Ohio State University, Columbus, Ohio.
b
College of Medicine, The Ohio State University, Columbus, Ohio.
Corresponding author and reprints: Michael F. Rayo, PhD, The Ohio State University, College of Medicine, School of Health and Rehabilitation Sciences, 37 W Kenworth Road, Columbus, OH 43214; e-mail: rayo.3@osu. edu. ª 2014 Published by Elsevier Inc. on behalf of American College of Radiology 1546-1440/14/$36.00 http://dx.doi.org/10.1016/j.jacr.2013.12.005
more pronounced, with CT volumes reduced by 4.5% [1,2]. In an effort to reduce patient radiation dose, health care providers have invested in technological interventions such as computer-based programs to reduce the ordering of inappropriate CT imaging [3,4]. In addition, CT equipment innovations such as tube current modulation [5,6], dual-energy imaging [6], and adaptive statistical iterative reconstruction techniques [6-8] and protocols such as weight-based variable kilovoltage selection and focused collimation produce comparable or better image quality with a reduced effective radiation dose [5,6]. The increased effort to decrease effective radiation dose is commonly attributed to 3 articles published between 2007 and 2009 [9-11]. Brenner and Hall [9] estimated that as many as 2% of all US cancers could be attributed to CT radiation exposure. Berrington de González et al [10] predicted that 29,000 future US cancers could be caused by the annual CT imaging 1
2 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
volume, and Smith-Bindman et al [11] suggested that 1 in 270 women undergoing CT coronary angiography may develop cancer from that procedure. Although the assumptions underlying these estimates were disputed by members of the imaging community [12-16], these articles caught the attention of the national news media [17-19], prominent health institutions such as Harvard Medical Center [20], and the FDA. Another potential cause of reduced ionizing radiation exposure was the Deficit Reduction Act of 2005 and other similar federal legislation that began reducing Medicare payments for advanced imaging beginning in 2007 [21]. Because this legislation affects only Medicare patients, it is unclear whether the observed reductions are predictive of a decrease in CT volume in the overall US patient population. In addition, it is unclear whether recent protocol and technology changes implemented for both Medicare and non-Medicare patients have affected individual and population cancer risk. As such, the purpose of this study was to determine the impact of changes in CT procedure volumes and dose reduction interventions on possible future cancer risk for the inpatient population at an academic medical center. METHODS
CT procedure volumes were obtained from an electronic database capable of providing deidentified patient information available for research purposes. Data were collected for calendar years 2008 to 2012. This study focused primarily on the inpatient population, so data were collected only from patients located in the medical intensive care unit and general medicine units. Procedure volumes for CT of the abdomen and pelvis, head, sinus, and lumbar spine were collected. Current Procedural Terminology, version 4, codes corresponding to the abdominal and pelvic codes were 72192, 72193, 72194, 74150, 74160, 74170, 74176, 74177, and 74178. Head CT codes were 70450, 70460, and 70470. Sinus CT codes were 70486, 70487, and 70488. Lumbar spinal CT codes were 72131 and 72132. Annual CT procedure volumes for each body area and for all body areas combined were
reported. The CT imaging rate was then determined by dividing CT imaging volume by the number of patient admissions for the medical intensive care unit and general medicine units. Because the data analyzed included the entire population under study for the selected years, comparative statistics were not used to determine differences. A subsequent analysis of inpatient abdominal ultrasound procedure volumes from 2008 to 2012 was conducted using the same methods and sources to help explain the findings on abdominal CT procedure volume. To calculate average effective dose in order to determine the effectiveness of the dose reduction strategies, the average dose-length products (DLP) for 2010 and 2012 were multiplied by tissue weighting factors supplied by the American Association of Physics in Medicine [22]. The average DLP and associated standard deviation for each body area was calculated for 2010 and 2012; those years were specifically selected to represent preintervention and postintervention data because all dose reduction strategies were implemented in 2011. To calculate the DLP, patient records from the institution’s PACS were randomly selected for each desired year and procedure. Data from 100 patients in each year were recorded for abdominal and pelvic CT. Eighty-seven patients in each year were sampled for head CT. For lumbar spinal CT, a smaller sample size of 10 patients was used to be proportional to the smaller annual procedure volume relative to the first 2 body areas. Insufficient data were available to perform this analysis for sinus CT. In addition, predicted induced cancers and related mortalities from these 3 body areas were calculated on the basis of estimates published in the BEIR VII report [23]. This study was approved by the institutional review board. RESULTS
Table 1 and Figure 1 show the inpatient CT procedure volumes from 2008 to 2012 for the 4 targeted body areas. Total imaging volume for CT imaging of these 4 body areas increased by 21% from 2008 to 2010 and then decreased by 30% from 2010 to 2012, for a net
Table 1. Annual inpatient CT imaging volumes and rates for the abdomen and pelvis, head, sinus, and lumbar spine from 2008 to 2012 Annual Volume (Rate per 100 Admitted Patients) Body Area 2008 2009 2010 2011 2012 Abdomen/pelvis Head Sinus Lumbar Total volume Total admissions
2,481 (39) 1,606 (25) 84 (1) 69 (1) 4,240 6,333
2,824 (42) 2,116 (31) 79 (1) 72 (1) 5,091 6,749
3,099 (42) 1,846 (25) 91 (1) 91 (1) 5,127 7,314
2,362 (30) 2,005 (25) 107 (1) 68 (1) 4,542 7,949
1,565 (17) 1,866 (20) 83 (1) 75 (1) 3,589 9,104
Rayo et al/Rate of Change in Exposure to Ionizing Radiation 3
Fig 1. Annual inpatient CT imaging volumes and rates for the abdomen and pelvis, head, sinus, and lumbar spine from 2008 to 2012.
decrease of 15% over the 5-year period. This same trend was seen individually for abdominal and pelvic imaging, which also increased from 2008 to 2010, then decreased from 2010 to 2012 for a 37% net decrease. The CT procedure volumes for imaging of the head, sinus, and lumbar spine all fluctuated year to year from 2008 to 2012. Over the 5-year span, sinus CT imaging decreased by 1%, yet head and lumbar spinal CT imaging increased by 16% and 9%, respectively. The CT imaging rate per admitted patient is shown in Table 1. Rates for abdominal and pelvic and head imaging rose from 2008 to 2009, then remained stable or decreased each year through 2012, for net decreases of 56% and 19%, respectively. Sinus and lumbar spinal CT imaging rates remained relatively stable at 1% for the 5-year period. Inpatient abdominal ultrasound imaging volume and rate are shown in Table 2. Procedure volume increased by 71% from 2008 to 2009 (from 316 to 542 procedures) and remained relatively stable through 2012. The rate of ultrasound procedures increased by 61% from 2008 to 2009 and declined from 2009 to 2012. Table 3 shows the average effective doses for 2010 and 2012 for CT procedures of the abdomen, head, and lumbar spine. The effective radiation dose was reduced from 14.7 mSv (95% confidence interval [CI], 13.1e16.3 mSv) to 7.1 mSv (95% CI, 6.4e7.7 mSv) for abdominal CT procedures. The effective dose attributed to head CT procedures was reduced from Table 2. Annual inpatient abdominal ultrasound imaging volume and rate from 2008 to 2012 Annual Volume Year (Rate per 100 Admitted Patients) 2008 2009 2010 2011 2012
316 542 484 574 603
(5.0) (8.0) (6.6) (7.2) (6.6)
2.8 mSv (95% CI, 2.6e2.9 mSv) to 2.0 mSv (95% CI, 2.0e2.0 mSv). Lumbar spinal procedure effective dose was also reduced from 20.2 mSv (95% CI, 15.8e24.6 mSv) to 14.2 mSv (95% CI, 10.6e17.8 mSv). During this time period, the academic medical center implemented 4 different dose reduction strategies. These strategies included the creation of enterprise-wide standardized imaging protocols enforcing dose uniformity on the basis of the recent literature [5,6], the use of tube current modulation [5,6,24], the use of variable weight-based kilovoltage [5,6], and the use of adaptive statistical iterative reconstruction techniques [6,8]. Table 4 compares 2010 and 2012 CT imaging of the abdomen, head, and lumbar spine. Effective dose was calculated as the product of the average effective dose and CT volumes reported. Between those 2 years, the average effective dose per patient was reduced by >60% (from 23.2 to 8.6 mSv). The number of CT procedures for patients undergoing at least one CT scan was reduced from 2.2 to 1.4. The number of patients who received a cumulative annual dose of 50 mSv, which is the limit for occupational exposure [25], decreased from 10% to 0.2%. The total exposure of the population was reduced from 53,281 to 19,971 mSv. After applying the findings from the BEIR VII report [23], this dose reduction resulted in decreases of both estimated induced cancer cases (from 10.1 to 3.8) and resultant mortalities (from 5.1 to 1.9). DISCUSSION
Recent studies [2,26] have shown that the volume and rate of advanced imaging studies have recently been decreasing. This is true for both CT and MR imaging among Medicare recipients. This study shows a trend similar to those studies. This is notable because the scope of this study is different: its focus is not just Medicare recipients but all general medicine and intensive care unit patients at an academic medical center receiving 1 of these 4 procedures. The same rationale noted by the previous studies [2,26] can be used to explain this study’s overall findings. Increased attention to radiation concerns from the news media and medical journals may have led referring physicians to use more stringent criteria for when to order advanced imaging. Cost-conscious physicians may have opted for less expensive but sufficient imaging alternatives, or no imaging at all, to aid their diagnoses. Physicians may have paid more attention to the appropriateness criteria published by the ACR [27] and other organizations. The publicity generated from programs such as Image Wisely [28] may have reduced overuse. Commercial payers’ requirement of prior authority for advanced imaging may have reduced the number of inappropriate tests. Last, the recession may have affected utilization, as many commercial payers require some out-of-pocket
4 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
Table 3. Average effective dose for abdominal and pelvic, head, sinus, and lumbar spinal CT at an academic medical center for 2010 and 2012 Average Effective Dose (mSv) (95% CI) Body Area 2010 2012 Interventions Abdomen/pelvis Head Lumbar spine
14.7 (13.1e16.3) 2.8 (2.6e2.9) 20.2 (15.8e24.6)
7.1 (6.4e7.7) 2.0 (2.0e2.0) 14.2 (10.6e17.8)
DU, WB, IR DU, IR DU, WB, IR
Note: CI ¼ confidence interval; DU ¼ dose uniformity between scanners; IR ¼ iterative reconstruction of resultant imaging; WB ¼ weight-based current and voltage modifications.
costs for advanced imaging. In addition to these explanations, a change in the default order sets in the institution’s electronic health record reduced the number of imaging procedures automatically ordered for most clinical indications. However, changes in head, sinus, and lumbar spinal CT volumes did not follow the trends noted in these national studies, and changes in CT imaging rates varied widely among the 4 body areas analyzed in this study. These findings may be better explained by the relative opportunity to decrease inappropriateness imaging, as documented by the ACR [27], for these body areas. The increase in inpatient head CT volume and smaller rate decrease relative to abdominal imaging is likely due to CT’s being the most appropriate modality for imaging head traumas, hearing loss and vertigo, and severe and sudden headache symptoms. In addition, CT was found to be the preferred modality at this institution because of the speed and accuracy of the test. The insubstantial volume changes and static imaging rates of sinus and lumbar spinal CT may be explained by the lack of opportunity to reduce inappropriate imaging for these procedures. For some time, imaging of the paranasal sinuses has been discouraged unless the patient is immunodeficient, has a neurologic deficit, or experiences chronic, recurrent sinusitis. Lumbar spinal imaging is also often discouraged, with MR as a consistently preferred modality when patient comorbidities make advanced imaging appropriate. Surprisingly, abdominal CT volume decreases were not associated with meaningful abdominal ultrasound procedure volume increases. We expected to see evidence of prescribing behaviors changing in favor of abdominal ultrasound, as it are less expensive, portable, Table 4. Comparison of 2010 and 2012 CT imaging statistics for patients undergoing at least one abdominal, head, sinus, or lumbar spinal CT scan Variable 2010 2012 Number of patients undergoing CT Average effective dose/patient (mSv) Number of CT studies/patient Percentage of patients receiving an effective dose >50 mSv Total radiation exposure (mSv)
2,321 23.2 2.2 10
2,340 8.6 1.4 0.2
53,281
19,971
and often the most appropriate imaging alternative to CT for the abdomen [27]. What we found was that, after both abdominal CT and ultrasound volumes increased from 2008 to 2009, abdominal ultrasound volumes plateaued, fluctuating year to year for a net increase of 61 procedures (11%) between 2009 and 2012, which is less than one-tenth of the decrease in abdominal CT volume during the same time period. It is likely that advanced abdominal imaging volume in general decreased for this patient population. The implemented dose-reduction strategies seem to be the main contributor to the reductions in effective dose per patient (from 23.2 to 8.6 mSv), the percentage of patients with annual effective dose >50 mSv (from 10% to 0.2%), predicted induced cancers (from 10.1 to 3.8), and predicted cancer deaths (from 5.1 to 1.9). These improvements were also primarily responsible for the almost 100-fold decrease in patients receiving a cumulative annual radiation dose of >50 mSv. This is significant as dose reductions for these patients have been predicted to have a disproportionately high effect on the overall cancer risk [29]. The finding that the dose restriction strategies had more impact than CT volume decreases in reducing the patients’ effective dose is consistent with the conclusion drawn by others that imaging technology manufacturers may play the largest role in making patients safer [30]. These new technologies provide new imaging opportunities [5-8] and facilitate the implementation of new protocols [5,6] that have been found to reduce the effective radiation dose to patients without compromising image quality. It is unclear whether the decrease in observed abdomen CT volume, the relatively static volumes of sinus and lumbar spinal CT volume, or the year-toyear fluctuations in the head CT data provide adequate foreshadowing for future CT volume trends. Another possible predictor, decreasing CT imaging rate for abdomen and head CT, seems to indicate increasingly conservative prescribing behavior for referring physicians and radiologists. Whatever the CT volume and rate may be in the future, however, the dose reduction strategies used at this institution have greatly reduced the predicted cancer risk to this patient population.
Rayo et al/Rate of Change in Exposure to Ionizing Radiation 5
These findings suggest that although efforts should continue in trying to reduce inappropriate diagnostic imaging for both cost and patient safety reasons, implementing protocol and technology improvements may be more effective in reducing patients’ effective radiation dose and estimated future cancer risk. This study was limited because it was constrained to the inpatient patient population of one midwestern tertiary care academic medical center. The DLPs reported were the result of an average taken by randomly sampling procedures, not of the actual DLP recorded for each procedure in the data set. Effective doses reported were the product of these DLPs and an estimated tissue weighting, not actual cell damage to the patient. We feel that our method of estimation is justified, however, given the small confidence intervals that resulted and the large reductions in dosage after the interventions. Another limitation is that because of the data available for this study, the relative effect size of each dose reduction strategy is unclear. It was also unclear whether the reduction in procedure volume and average radiation dose were due to increased scrutiny of appropriateness or to other factors, including differences in ordering protocols or technology. These unanswered questions are promising areas for future research. Last, the cancer risk predictions were based on the linear no-threshold model, which has substantial supporting evidence [22] but has not been proved. TAKE-HOME POINTS
A 15% net decrease in total annual CT procedures from 2008 to 2012 at an urban tertiary-care hospital was driven primarily by the 37% decrease in abdominal imaging. CT volumes for the head, sinus, and lumbar spine fluctuated slightly from year to year but remained relatively unchanged after the 5-year period. Reductions in CT rate from 2008 to 2012 per patient seem to indicate a more conservative prescribing behavior among referring physicians and radiologists. Dose reduction strategies, including new protocols and technology enhancements, resulted in 30% to 52% dose reductions. The combination of CT volume reduction and dose reduction for each procedure reduced the predicted risk for induced cancer and cancer deaths by 63%. The percentage of patients receiving an annual effective radiation dose of 50 mSv was reduced from 10% to 0.2%. Investments in new protocols and technology enhancements were more effective than reduced CT volume in reducing patients’ effective radiation dose and associated future cancer risk.
REFERENCES 1. Levin DC, Rao VM, Parker L, Frangos AJ, Sunshine JH. Bending the curve: the recent marked slowdown in growth of noninvasive diagnostic imaging. AJR Am J Roentgenol 2011;196: W25-9. 2. Levin DC, Rao VM, Parker L. The recent downturn in utilization of CT: the start of a new trend? J Am Coll Radiol 2012;9:795-8. 3. Vartanians VM, Sistrom CL, Weilburg JB, Rosenthal DI, Thrall JH. Increasing the appropriateness of outpatient imaging: effects of a barrier to ordering low-yield examinations. Radiology 2010;255:842-9. 4. Blackmore CC, Mecklenburg RS, Kaplan GS. Effectiveness of clinical decision support in controlling inappropriate imaging. J Am Coll Radiol 2011;8:19-25. 5. Goldman AR, Maldjian PD. Reducing radiation dose in body CT: a practical approach to optimizing CT protocols. AJR Am J Roentgenol 2013;200:748-54. 6. Maldjian PD, Goldman AR. Reducing radiation dose in body CT: a primer on dose metrics and key CT technical parameters. AJR Am J Roentgenol 2013;200:741-7. 7. Willemink MJ, Leiner T, Jong PA, et al. Iterative reconstruction techniques for computed tomography part 2: initial results in dose reduction and image quality. Eur Radiol 2013;23:1632-42. 8. Han BK, Grant KLR, Garberich R, Sedlmair M, Lindberg J, Lesser JR. Assessment of an iterative reconstruction algorithm (SAFIRE) on image quality in pediatric cardiac CT datasets. J Cardiovasc Comput Tomogr 2012;6:200-4. 9. Brenner DJ, Hall EJ. Computed tomography—an increasing source of radiation exposure. N Engl J Med 2007;357:2277-84. 10. Berrington de González A, Mahesh M, Kim K-P, et al. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med 2009;169:2071-7. 11. Smith-Bindman R, Lipson J, Marcus R, et al. Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Arch Intern Med 2009;169:2078-86. 12. Cameron JR. Link between diagnostic X-rays and cancer uncertain. Evidence-Based Healthc 2004;8:205-6. 13. Mezrich R. Are CT scans carcinogenic? J Am Coll Radiol 2008;5: 691-3. 14. Mezrich RS. Cancer risk from multiple imaging tests. JAMA 2011;305: 887-8. 15. Budoff MJ, Gupta M. Radiation exposure from cardiac imaging procedures: do the risks outweigh the benefits? J Am Coll Cardiol 2010;56: 712-4. 16. Gerber TC, Gibbons RJ. Weighing the risks and benefits of cardiac imaging with ionizing radiation. JACC Cardiovasc Imaging 2010;3: 528-35. 17. Guthrie C. How dangerous are CT scans? Time. June 27, 2008. Available at: http://content.time.com/time/health/article/0,8599,181852 0,00.html. Accessed December 27, 2013. 18. Shulman M. Facing a CT scan? Think about radiation. US News and World Report. May 30, 2008. Available at: http://health. usnews.com/health-news/managing-your-healthcare/cancer/articles/ 2008/05/30/facing-a-ct-scan-think-about-radiation. Accessed December 27, 2013. 19. Knox R. Radiation from CT scans may raise cancer risk. National Public Radio. December 15, 2009. Available at: http://www.npr.org/ templates/story/story.php?storyId¼121436092. Accessed December 27, 2013. 20. Robb-Nicholson C. A doctor talks about: radiation risk from medical imaging. Harvard Health Publications. October 2010. Available at:
6 Journal of the American College of Radiology/Vol.
-
No.
-
Month 2014
http://www.health.harvard.edu/newsletters/Harvard_Womens_Health_ Watch/2010/October/radiation-risk-from-medical-imaging. Accessed December 27, 2013.
25. Nuclear Regulatory Commission. Occupational dose limits for adults (10 CFR 20.1201). Available at: http://www.nrc.gov/reading-rm/doccollections/cfr/part020/part020-1201.html. Accessed July 17, 2013.
21. Levin DC, Rao VM, Parker L, Frangos AJ. The sharp reductions in Medicare payments for noninvasive diagnostic imaging in recent years: will they satisfy the federal policymakers? J Am Coll Radiol 2012;9:643-7.
26. Sharpe RE Jr, Levin DC, Parker L, Rao VM. The recent reversal of the growth trend in MRI: a harbinger of the future? J Am Coll Radiol 2013;10:599-602.
22. McCollough C, Cody D, Edyvean S, Geise R, Gould B. The measurement, reporting, and management of radiation dose in CT. College Park, Maryland: American Association of Physicists in Medicine; 2008. 23. National Research Council. BEIR VII: health risks from exposure to low levels of ionizing radiation. Washington, District of Columbia: National Academies Press; 2005. 24. McCollough CH, Primak AN, Braun N, Kofler J, Yu L, Christner J. Strategies for reducing radiation dose in CT. Radiol Clin North Am 2009;47:27-40.
27. American College of Radiology. ACR Appropriateness Criteria. Available at: http://www.acr.org/Quality-Safety/Appropriateness-Criteria. Accessed July 17, 2013. 28. Brink JA, Amis ES. Image Wisely: a campaign to increase awareness about adult radiation protection. Radiology 2010;257:601-2. 29. Miglioretti DL. The use of computed tomography in pediatrics and the associated radiation exposure and estimated cancer risk. JAMA Pediatr 2013;167:700-7. 30. Rehani MM. Challenges in radiation protection of patients for the 21st century. AJR Am J Roentgenol 2013;200:762-4.