- Email: [email protected]
Extracolonic findings at CT colonography: Evaluation of prevalence and cost in a screening population
GASTROENTEROLOGY 2003;124:911–916
Extracolonic Findings at CT Colonography: Evaluation of Prevalence and Cost in a Screening Population THOMAS M. GLUECKER,* C. DANIEL JOHNSON,* LYNN A. WILSON,* ROBERT L. MACCARTY,* TIMOTHY J. WELCH,* DAVID J. VANNESS,‡ and DAVID A. AHLQUIST§ *Department of Radiology, ‡Department of Health Care Policy and Research, and the §Department of Internal Medicine, Mayo Clinic Rochester, Rochester, Minnesota
Background & Aims: To assess the prevalence and spectrum of extracolonic findings in a screening population undergoing computed tomography colonography (CTC), and to evaluate the short-term direct medical costs incurred from subsequent radiologic follow-up evaluation. Methods: Six hundred and eighty-one asymptomatic patients undergoing colonoscopy screening consented to a CTC examination. Extracolonic CT findings were classified into high, medium, and low importance. Clinical and radiologic follow-up, missed lesions, and outcomes were assessed by chart review (time interval, 410 –1513 days; median, 913 days). Short-term direct medical costs of radiologic follow-up were determined based on Medicare 2002 reimbursement rates. Results: Extracolonic findings were found commonly. These were categorized as high clinical importance in 71 (10%) individuals, as medium importance in 183 individuals (27%), and as low importance in 341 individuals (50%). Subsequent medical or surgical interventions resulted from these findings in 9 of the 681 patients (1.3%). Costs of subsequent radiologic follow-up studies were calculated as $23,380.59 (average added costs per CTC examination $34.33). Conclusions: CTC commonly detects extracolonic findings that can be considered clinically important when applied to an asymptomatic screening population. Although such incidental findings add benefit to the screening intervention, moderate incremental costs are incurred based on additional radiologic procedures generated during short-term follow-up.
omputed tomography colonography (CTC) is a new radiologic technique for examination of the colorectum. The examination usually is performed in the prepared colon by using a low-dose CT technique in both supine and prone positions.1–3 By using advanced imaging software (axial and multiplanar 2-dimensional reformatted and 3-dimensional– endoluminal view), images of the colon are reviewed to provide a thorough and noninvasive evaluation of the entire colorectum.1– 4 Recent studies indicate CTC likely will be competitive with other full structural examinations of the colorectum.1,2,5
C
A unique capability of CTC over other colorectal examinations is the display of the entire abdominal and pelvic contents.6 Detection of important abdominal and pelvic extracolonic abnormalities could potentially benefit patients undergoing colorectal screening. At risk is the possible increase in the cost of this examination by the discovery and work-up of clinically insignificant findings. Several reports of extracolonic findings at CTC have been published, but none exist in a screening population that most likely might benefit from the procedure.6,7 The aims of this study were (1) to describe the spectrum of extracolonic findings in a large asymptomatic screening population undergoing CTC, and (2) to evaluate the short-term direct medical costs incurred from subsequent radiologic follow-up during that patient care visit.
Materials and Methods Seven hundred and nineteen consecutive asymptomatic patients who were prescheduled for colonoscopic screening were recruited and enrolled in this study. All patients signed an informed consent form, and the study was approved by our institutional review board. Inclusion criteria were asymptomatic patients with a family history of colorectal cancer or polyps (first-degree relative), a prior personal history of colon polyps or colorectal cancer, or the new onset of asymptomatic anemia. Exclusion criteria included patients with melena, hematochezia, inflammatory bowel disease, familial polyposis, pregnancy, recent bowel resection, or recent polyp removal within 2 months before the examination. Thirty-eight patients (a subset of patients with prior partial colon resection requiring protocol administration of intravenous contrast material and standard CT dose settings) underwent CTC and were excluded from this study. A total of 681 patients formed the basis for the study. The patients’ ages ranged from 41 to 80 years, with a median Abbreviation used in this paper: CTC, computed tomography colonography. © 2003 by the American Gastroenterological Association 0016-5085/03/$30.00 doi:10.1053/gast.2003.50158
912
GLEUCKER ET AL.
age of 64 years. Only 6 patients were under the age of 50. The male-female ratio was 426:255. Patients were placed in a decubitus position and a small enema tip was placed in the rectum. The colon was inflated gently with carbon dioxide to patient tolerance (feeling full without pain). CT scouts were obtained before each acquisition to ensure full distention and anatomic coverage of the abdomen and pelvis. Additional carbon dioxide was added before the second acquisition to ensure full colon distention. Examinations were performed on either a single (GE Highspeed Advantage; GE Medical Systems, Milwaukee, WI) (107 patients) or multidetector CT scanner (GE Lightspeed, Lightspeed Plus; GE Medical Systems) (574 patients). Single-detector scans were performed using a collimation of 5 mm, pitch of 1.3, 3-mm reconstruction interval, 70 mA, 120 kVp, 512 ⫻ 512 matrix, and 180° interpolation algorithm. Single-detector scans were performed using 20-second breath holds (3– 4) with a 3-cm overlap between acquisitions. Multidetector scans were performed using a collimation of 5 mm (4 ⫻ 5 mm detector configuration), table speed of 15 mm per rotation, high quality mode, 2.5-mm reconstruction interval, 40 mA (determined to match image noise of single-slice technique), 120 kVp, 512 ⫻ 512 matrix, and a single 20-second breath hold. Both supine and prone acquisitions were obtained in all patients. The image data was networked to a workstation using customized software developed at our institution. One of 3 board-certified and experienced (with 10 years of practice) radiologists was assigned randomly to detect and report the extracolonic findings. Image review included review of the supine axial images using lung, soft-tissue, and bone windows. None of the observers were aware of the patient’s medical history, colonoscopic findings, or the findings of the other reviewers. The extracolonic findings were separated into 3 categories of clinical importance: high, medium, and low. Findings of high importance included those requiring surgical treatment, medical intervention, and/or further investigation during that patient care visit. Examples of these included indeterminate solid organ masses, previously unknown abdominal aortic aneurysms 3 cm or larger, aneurysms of the splenic or renal arteries, indeterminate chest nodule, adenopathy (lymph nodes ⬎1 cm), pancreatic masses, and small bowel parasites. Findings of medium importance included conditions that did not require immediate treatment but would likely require investigation, recognition, or treatment at a later time. Examples of medium importance included calculi of various organs, previously known abdominal aortic aneurysms, adrenal masses (⬍2 cm in diameter), pancreatic pseudocysts, indeterminate cysts of various organs, uterine enlargement in postmenopausal women, and coronary artery calcifications. Findings of low importance were considered benign and unlikely to require further medical treatment or additional work-up. Examples of these included vascular calcifications (other than coronary artery calcification), granulomas, diverticulosis, simple solid organ cysts, small- to medium-sized hiatal hernias, and abdominal wall hernias containing fat.
GASTROENTEROLOGY Vol. 124, No. 4
A chart review was performed on all patients with an extracolonic finding(s) to determine additional radiologic follow-up for each finding. At least 12 months of follow-up was possible for each patient with a range of 410 –1513 days (mean, 931 days). The type, number, and results of these examinations were tabulated, including surgical or medical treatment outcomes relating to the detected findings. The medical records also were assessed for the development of subsequent conditions that were not detected at CTC during the same time period. The short-term direct medical cost of the follow-up radiologic tests was estimated using 2002 Medicare average reimbursement rates for professional services, and Medicare-reported nationwide mean technical and facility costs for providing those services in a hospital outpatient setting. The Resource Based Relative Value System on which Medicare reimbursement is based provides a reasonable proxy to the true opportunity cost of resources used in providing medical services.8 Patients with findings considered to be of low clinical importance had no follow-up costs because each finding with a suggested or performed follow-up examination was considered to be of either moderate or high importance. If a patient had a significant extracolonic finding, and a follow-up examination was suggested either by the referring physician or the radiologist but no further clinical information could be obtained, the cost of the suggested follow-up examination was imputed.
Results There were 858 extracolonic findings in 469 of 681 (69%) patients. No extracolonic findings were detected in 212 of 681 (31%) patients. Findings were categorized as high, medium, or low importance. Table 1 delineates the highly important extracolonic findings, follow-up radiologic examinations, and the associated costs. A total of 88 lesions in 71 of 681 (10%) patients classified as highly important were discovered. These included 34 lesions within the kidney, 27 in the chest, 8 in the liver, 6 in the ovary, 4 in the renal or splenic arteries, 3 in the retroperitoneum (other than the aorta), and one in the pancreas. A total of 94 follow-up procedures were performed to evaluate these lesions, predominantly consisting of ultrasound and CT examinations. Total follow-up radiologic costs in this group were $19,915.40. Nine of the 71 (13%) patients received treatment as a result of the CT findings. These included repair of an unsuspected abdominal aortic aneurysm, resection of a squamous cell carcinoma of the lung, chemotherapy for unsuspected thyroid metastases to the lung, resection of a renal adenocarcinoma, resection of a renal oncocytoma, resection of a serous cystadenoma of the ovary in 3 different patients, and medical treatment for ileal ascariasis. Three patients received additional radiographic follow-up of the pelvis to evaluate a cystic
April 2003
EXTRACOLONIC FINDINGS AT CT COLONOGRAPHY
913
Table 1. Highly Significant Findings, Follow-Up, and Costs
Findings at CTC
No. lesions
Follow-up radiologic examination
4
US Doppler aorta CT abdomen ⫹ pelvis/c US renal CT abdomen/c KUB CT chest w/o CT chest w/o Chest radiograph CT abdomen/c US abdomen CT abdomen/c US abdomen CT abdomen/c US abdomen US pelvis ⫹ endovaginal US abdomen CT abdomen/c US Doppler renal arteries CT abdomen/c None
Aorta
Unknown aneurysm
Kidney
Mass
34
Chest
Retrocrural LN Lung nodule
1 26
Liver
Low-attenuation area Cystic lesion Mass
1 5 2
Retroperitoneum
Adenopathy Cystic nodule Cystic lesion Cystic lesion
1 2 6 1
Ovary Pancreas Renal artery Splenic artery Small Bowel Total
Suspicious aneurysm Aneurysm Parasites
1 3 1 88
No. of follow-up procedures
Cost
Total follow-up costs
1 2 22 2 2 2 31 7 2 5 1 1 2 1 9 1 1 1 1
200.93 1154.46 169.12 584.49 219.46 466.66 466.66 71.71 584.49 223.91 584.49 223.91 1154.46 223.91 361.88 223.91 584.49 200.93 584.49
200.93 2308.92 3720.64 1168.98 438.92 933.32 14,466.46 501.97 1168.98 1119.55 584.49 223.91 2308.92 223.91 3256.92 223.91 584.49 200.93 584.49
94
34,220.64
/c, with intravenous contrast material administration; KUB, plain abdominal radiograph; LN, lymph node; US, ultrasound; w/o, without intravenous contrast material.
ovarian lesion, and one patient had a planned repeat CT scan to follow a cystic pancreatic neoplasm. In addition, there were lesions that were not followed-up at our institution, but were to be evaluated by their local physicians (indeterminate renal mass [n ⫽ 7], indeterminate lung nodules [n ⫽ 5], cystic lesion in the liver [n ⫽ 1]). For these patients, the following investi-
gations were suggested: renal ultrasound for the indeterminate renal lesions, non– contrast-enhanced CT of the chest for the indeterminate lung lesions, and ultrasound of the liver for the cystic liver lesion. The estimated cost of these follow-up examinations would be $2,247.85. Table 2 summarizes the 196 findings in 183 of 681 (27%) patients that were classified as moderately signif-
Table 2. Moderately Significant Findings, Follow-Up, and Costs
Findings at CTC Adrenal Bladder Liver Lung
Mesentery Pancreas Kidney
Aorta Gallbladder Gastric Other Total
No. lesions
Adenoma Calculi Focal lesion Fibrosis
10 3 14 1
Tiny nodules Known lung fibrosis Lymph nodes Pseudocyst Calculi Angiomyolipoma Renal calcifications Horseshoe Known aneurysm Unknown Ectasia Calculi Large hiatal hernia
3 3 1 8 66
4 19 40 8 25 197
Follow-up radiologic examination CT abdomen/c KUB CT abdomen/c CT chest w/o Chest x-ray None None CT abdomen/c CT abdomen ⫹ pelvis w/o KUB None None None US Doppler aorta US Doppler aorta None None None
No. of follow-up procedures
Cost
Total follow-up costs
1 2 2 1 1
584.49 219.49 584.49 466.66 71.71
584.49 438.98 1168.98 466.66 71.71
1 1 1
584.49 941.21 219.49
584.49 941.21 219.49
1
200.93 200.93
200.93 803.72
0
16
5704.57
914
GLEUCKER ET AL.
GASTROENTEROLOGY Vol. 124, No. 4
Table 3. Low Significant Findings Findings
No. Patients
Vascular calcifications Diverticulosis Simple renal cysts Simple liver cysts Small- or medium-sized hiatal hernia Lung granuloma Prostate calcifications Liver granuloma Calcified uterine fibroid Spleen granuloma Mild fibrotic changes in lung bases Accessory spleen Calcifications mitral valve Benign bony island Other Total
107 81 56 55 55 37 36 33 28 26 21 14 8 7 10 574
icant: abnormalities of the kidneys (n ⫽ 66), 64 of these owing to renal calculi; abnormalities of the abdominal aorta (known aneurysm ⬎3 cm or ectasia 2–3 cm in diameter) (n ⫽ 23); liver lesions (n ⫽ 14); adrenal masses (n ⫽ 10); pancreatic pseudocysts (n ⫽ 8); lung abnormalities (n ⫽ 7); and bladder calculi (n ⫽ 3). One patient with a large hiatal hernia causing significant reflux esophagitis had a laparoscopic repair of the hernia. Many patients received treatment for renal and bladder calculi; however, imaging follow-up was not long enough to assess the outcome of these therapies, and many patients received further care by their home physicians without follow-up records. Only 15 of the 196 (7.7%) findings classified as moderate significance were followed-up with additional documented radiologic imaging tests at a total cost of $3,465.19. It is difficult to estimate the number of additional tests that would be ordered as a result of these findings, but certainly many of the patients with renal and bladder calculi, abdominal aortic aneurysms, and tiny nodules in the lung bases would, in all likelihood, receive additional imaging tests in the future. Five hundred and seventy-four findings in 341 of 681 (50%) patients were designated to be of low significance (Table 3) and by definition had no additional follow-up testing and associated costs. The most common findings included vascular calcifications, diverticulosis, benignappearing cysts within the liver and kidneys, small- and medium-sized hiatal hernias, and calcified granulomas. A total of 44 lesions were not reported at CTC examination but described in subsequent radiologic examinations or in the clinical charts. Among them, 11 of 44 (25%) could be detected retrospectively, 33 (75%) not. Among these 44 lesions were 3 findings of high importance: 1 pancreatic adenocarcinoma, 1 splenic mass, and
1 ovarian mass. The pancreatic and the splenic lesion could not be detected retrospectively. The ovarian mass could be detected retrospectively at CTC. Twenty lesions of medium importance were not described at CTC, among them 8 renal stones, 1 ileal lipoma in proximity of the ileocecal valve, 1 adrenal adenoma, 2 angiomyolipomas of the kidney, 5 fatty livers, 1 fatty pancreas, 1 complicated renal cyst, and 1 gallstone. Among them, 2 renal stones, 1 ileal lipoma, and 1 adrenal adenoma could be found retrospectively. The remaining 17 lesions of medium importance in 17 patients could not be identified retrospectively. Twenty-one subsequently identified lesions were of low importance, including 10 renal cysts (3 detected retrospectively), 8 liver cysts (2 detected retrospectively), and 3 hepatic hemangiomas (1 detected retrospectively). Fifteen of these 21 lesions (71%) were not identified retrospectively.
Discussion As CTC undergoes scrutiny as a candidate approach to colorectal cancer screening, the balance between potential benefits and potential harms or costs of this new technology must be considered critically. Incidentally discovered extracolonic findings represent a unique feature of CTC that sets it apart from other approaches to colorectal cancer screening. This feature may lead to added benefit via early detection of unsuspected health-threatening pathology but may do so at the cost incurred from downstream interventions and unanticipated iatrogenic morbidity. The present study was undertaken to explore such benefits and costs of extracolonic lesion detection by CTC in an asymptomatic population being screened for colorectal cancer. Based on our study, approximately two thirds of asymptomatic individuals undergoing screening CTC will have unsuspected extracolonic lesions detected. Most of these findings will be inconsequential and not require additional imaging studies or other interventions. However, approximately 10% of our patients had a finding considered to be important clinically and 13% of these patients (1.3% of the study population) received surgery or medical treatment as a result of the radiologic discovery. Extracolonic abnormalities were found in nearly every organ of the abdomen and pelvis. Abnormalities in the kidneys, ovaries, lung, and aorta constituted the organs with the most prevalent findings of significance. It is not possible to determine if mortality reduction actually occurred. Following the approach of Sonnenberg et al.,9 we estimated the average cost of CTC based on 2002 Medicare professional reimbursement and mean technical and
April 2003
facility cost for CT of the abdomen with and without contrast plus 3-dimensional reconstruction, totaling $560.31. The overall short-term cost for the additional radiologic examinations was $23,380.59 for the 681 patients evaluated, thus adding $34.33 (6%) to the average cost of the examination. By examining only short-term radiologic costs, we are necessarily adopting a limited perspective for the cost analysis. Other direct and indirect resource costs and savings10 were not studied. Such costs and savings would include the value of patient and informal caregiver time, and resources devoted to subsequent evaluation and treatment or other nonradiologic procedures incurred or foregone as the result of the extracolonic CTC findings. Nonmonetary benefits and harms (including short-term patient anxiety and concern, as well as long-term morbidity, mortality, and quality-of-life outcomes) associated with follow-up and treatment of extracolonic findings also were not considered. The awareness of findings of even medium or low importance might create distress for the patient such that the referring physician must initiate further work-up for these findings. This could result in additional costs and morbidity. In one study addressing psychologic side effects of breast cancer screening, women with suspicious mammography findings showed significantly elevated mammography-related anxiety, despite the fact that a malignant lesion was ruled out by follow-up studies.11 Investigating this issue was outside of the scope of this study. Because colorectal screening is now advocated for all adults (average-risk patients over age 50, and higher-risk patients over age 40), CTC presents a unique opportunity to assess the abdomen and pelvis for other diseases that affect this older population. Successful treatment of important asymptomatic conditions could provide unexpected cures and reduced morbidity in selected patients. Anticipated prevalence of large colorectal adenomas (ⱖ1 cm) is only 5% to 8% in a screening population.12–14 In this study, 10% of patients were identified with extracolonic findings considered highly significant. A total of 1.5% of all patients underwent treatment (either surgical or medical) for these unsuspected findings. In view of the low prevalence of significant colorectal adenomas in this patient group, these additional findings and their modest incremental costs may be beneficial and potentially could add considerable value to the examination. This study did not address the cost effectiveness of CTC in light of these findings. Additional outcomes-related studies with longitudinal follow-up of these patients would be helpful in the future.
EXTRACOLONIC FINDINGS AT CT COLONOGRAPHY
915
Several studies have been published regarding the prevalence of incidental findings at CTC, including one from our institution.6 Hara et al.6 found extracolonic findings in 41% of 264 patients studied. Eleven percent of the patients had findings considered highly clinically significant, including 6 patients (2.3%) who underwent surgery for these unsuspected abnormalities. The majority of the patients in this study had previously diagnosed colon lesions and, therefore, were not representative of a screening population. Edwards et al.7 reported on extracolonic findings in 100 patients over the age of 55 with an overall prevalence of extracolonic abnormalities of 15%. Two of these 15 (1.3%) patients underwent surgery for benign ovarian cysts. Swensen et al.15,16 reported the results of low-dose CT scanning (screening) of the chest and abdomen in patients at high risk for lung cancer (cigarette smokers). Two hundred and ten of 1520 (14%) of these patients had significant findings other than lung lesions, including abdominal aortic aneurysms in 3.3%.15 Because cigarette smoking predisposes to a number of extrapulmonary diseases, it is difficult to compare these findings with those in our study. Hara et al.6 in 2000 also found that the additional imaging studies added approximately $28 per patient to the cost of the examination. Our average cost of $34.33 is slightly higher, in part owing to a higher overall prevalence of extracolonic findings (69% vs. 41%). Limitations of this technology include the failure to detect important abnormalities at CTC. The low-dose technique used at CTC capitalizes on the high contrast that exists between the air-filled colon lumen and the soft-tissue density wall. Polyps protruding into the airfilled lumen can be detected using these low-dose techniques. Solid-organ contrast requires higher radiation doses for optimal lesion detection. In addition, lesionorgan contrast differences (and lesion detection) can be improved with the use of intravascular iodinated contrast material. None of the patients in this study received intravenous contrast material, and all examinations were performed using a low-dose technique. As a result, it is likely that some solid organ lesions were not detected. Follow-up of our patients for an average time interval of 21 months revealed additional lesions that were detected with other subsequent examinations. Among them were 3 lesions of high importance, 1 adenocarcinoma of the pancreas, 1 splenic, and 1 ovarian mass. The pancreatic and the splenic mass could not be detected retrospectively. Radiologists and referring physicians need to be aware that assessment of solid organs is a potential limitation of this procedure. To decrease patient dose, there are currently studies investigating if the radiation
916
GLEUCKER ET AL.
dose could be further decreased without compromising the detection of polyps larger or equal to 5 mm.17 One study showed that by lowering the dose to 30 mAs, the detection of colonic lesions remains unimpaired.17 However, further reduction of the radiation dose would increase imaging noise. This is likely to compromise the evaluation of extracolonic parenchymal organs (i.e., liver, pancreas, spleen). It is possible that an abdominal CT scanner with further dose reduction might be insufficient for the detection of extracolonic findings and should be subject to further studies. Clinical problems related to solid organs are studied optimally using conventional CT technique. The CTC examination is optimized for the detection of colonic lesions but, according to its inherent low-dose technique, not designed for the detection of extracolonic lesions. The medicolegal implications of this limitation have not yet been explored fully. The results of this study should not be construed as an endorsement for whole-body screening. As previously stated, the low-dose technique (without intravenous contrast material administration) has limitations for the detection of lesions in solid organs. The authors are not aware of any national organization recommending wholebody scanning, primarily because sufficient evidence supporting this practice change is lacking. We hope this study contributes to a better understanding of abdominopelvic screening in an asymptomatic older population. In conclusion, extracolonic abnormalities are exceedingly common in asymptomatic populations undergoing screening by CTC using a low-dose technique. These incidental extracolonic findings can be categorized according to their clinical relevance; and, based on our observations, roughly 10% could be considered clinically important and would require further evaluation or intervention. The average additional radiologic cost of added imaging studies is approximately $34 per patient across the whole study group, although this likely does not capture the true resource cost or benefit from a societal perspective. The potential benefits of detecting extracolonic abnormalities at CTC include the discovery of important asymptomatic disease with the possibility of early treatment and cure. Radiologists and clinicians must be cognizant of the high prevalence of extracolonic findings found at CTC, and make efforts to triage patients appropriately for additional tests in a fiscally responsible fashion.
GASTROENTEROLOGY Vol. 124, No. 4
2. Fletcher JG, Johnson CD, Welch TJ, MacCarty RL, Ahlquist DA, Reed JE, Harmsen WS, Wilson LA. Optimization of CT colonography technique: prospective trial in 180 patients. Radiology 2000; 216:704 –711. 3. Hara AK, Johnson CD, MacCarty RL, Welch TJ, McCollough CH, Harmsen WS. CT colonography: single- versus multi-detector row imaging. Radiology 2001;219:461– 465. 4. Reed J, Johnson C. Automatic segmentation, tissue characterization, and rapid diagnosis enhancements to the computed tomographic colonography analysis workstation. J Digit Imaging 1997; 10(Suppl):70 –73. 5. Yee J, Akerkar GA, Hung RK, Steinauer-Gebauer A, Wall SD, McQuaid KR. Colorectal neoplasia: performance characteristics of CT colonography for detection in 300 patients. Radiology 2001;219:685– 692. 6. Hara AK, Johnson CD, MacCarty RL, Welch TJ. Incidental extracolonic findings at CT colonography. Radiology 2000;215:353– 357. 7. Edwards JT, Wood CJ, Mendelson RM, Forbes GM. Extracolonic findings at virtual colonoscopy: implications for screening programs. Am J Gastroenterol 2001;96:3009 –3012. 8. Lave JR, Pashos CL, Anderson GF, Brailer D, Bubolz T, Conrad D, Freund DA, Fox SH, Keeler E, Lipscomb J, Luft HS, Provenzano G. Costing medical care: using Medicare administrative data. Med Care 1994;32:JS77–JS89. 9. Sonnenberg A, Delco F, Bauerfeind P. Is virtual colonoscopy a cost-effective option to screen for colorectal cancer? Am J Gastroenterol 1999;94:2268 –2274 10. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. Oxford: Oxford University Press, 1996. 11. Lerman C, Trock B, Rimer BK, Jepson C, Brody D, Boyce A. Psychological side effects of breast cancer screening. Health Psychol 1991;10:259 –267. 12. O’Brian MJ, Winawer SJ, Zauber AG, Gottlieb LS, Sternberg SS, Diaz B, Dickersin GR, Ewing S, Geller S, Kasimian D, Komorowski R, Szoporn A. The national polyp study. Patient and polyp characteristics associated with high-grade dysplasia in colorectal adenomas. Gastroenterology 1990;98:371–379 13. DiSario JA, Fouth PG, Mai HD, Pardy K, Manne RK. Prevalence and malignant potential of colorectal polyps in asymptomatic, average-risk men. Am J Gastroenterol 1991;86:941–945. 14. Cannon-Albright LA, Bishop DT, Samowitz W, DiSario JA, Lee R, Burt RW. Colonic polyps in an unselected population: prevalence, characteristics, and associations. Am J Gastroenterol 1994;89: 827– 831. 15. Swensen SJ, Viggiano RW, Midthun DE, Mueller NL, Sherrick A, Yamashita K, Naidich DP, Patz EF, Hartman TE, Muhm JR, Weaver AL. Lung nodule enhancement at CT: multicenter study. Radiology 2000;214:73– 80. 16. Swensen SJ, Jett JR, Sloan JA, Midthun DE, Hartman TE, Sykes AM, Augenbaugh GL, Zink FE, Hillman SL, Noetzel GR, Marks RS, Clayton AC, Pairolero PC. Screening for lung cancer with low-dose spiral computed tomography. Am J Respir Crit Care Med 2002; 165:508 –513. 17. vanGelder RE, Venema HW, Serlie IWO, Yung Nio C, Determann RM, Tipker CA, Vos FM, Glas AS, Bartelsman JFW, Bossuyt PMM, Larneris JS, Stocker J. CT colonography at different radiation dose levels: feasibility of dose reduction. Radiology 2002;224: 25–33.
References 1. Fenlon HM, Nunes DP, Schroy III PC, Barish MA, Clarke PD, Ferrucci JT. A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps. N Engl J Med 1999;341: 1496 –1503.
Received August 27, 2001. Accepted January 9, 2003. Address requests for reprints to: C. Daniel Johnson, M.D., Department of Radiology, Mayo Clinic Rochester, 200 First Street SW, Rochester, Minnesota 55902.
Extracolonic Findings at CT Colonography: Evaluation of Prevalence and Cost in a Screening Population THOMAS M. GLUECKER,* C. DANIEL JOHNSON,* LYNN A. WILSON,* ROBERT L. MACCARTY,* TIMOTHY J. WELCH,* DAVID J. VANNESS,‡ and DAVID A. AHLQUIST§ *Department of Radiology, ‡Department of Health Care Policy and Research, and the §Department of Internal Medicine, Mayo Clinic Rochester, Rochester, Minnesota
Background & Aims: To assess the prevalence and spectrum of extracolonic findings in a screening population undergoing computed tomography colonography (CTC), and to evaluate the short-term direct medical costs incurred from subsequent radiologic follow-up evaluation. Methods: Six hundred and eighty-one asymptomatic patients undergoing colonoscopy screening consented to a CTC examination. Extracolonic CT findings were classified into high, medium, and low importance. Clinical and radiologic follow-up, missed lesions, and outcomes were assessed by chart review (time interval, 410 –1513 days; median, 913 days). Short-term direct medical costs of radiologic follow-up were determined based on Medicare 2002 reimbursement rates. Results: Extracolonic findings were found commonly. These were categorized as high clinical importance in 71 (10%) individuals, as medium importance in 183 individuals (27%), and as low importance in 341 individuals (50%). Subsequent medical or surgical interventions resulted from these findings in 9 of the 681 patients (1.3%). Costs of subsequent radiologic follow-up studies were calculated as $23,380.59 (average added costs per CTC examination $34.33). Conclusions: CTC commonly detects extracolonic findings that can be considered clinically important when applied to an asymptomatic screening population. Although such incidental findings add benefit to the screening intervention, moderate incremental costs are incurred based on additional radiologic procedures generated during short-term follow-up.
omputed tomography colonography (CTC) is a new radiologic technique for examination of the colorectum. The examination usually is performed in the prepared colon by using a low-dose CT technique in both supine and prone positions.1–3 By using advanced imaging software (axial and multiplanar 2-dimensional reformatted and 3-dimensional– endoluminal view), images of the colon are reviewed to provide a thorough and noninvasive evaluation of the entire colorectum.1– 4 Recent studies indicate CTC likely will be competitive with other full structural examinations of the colorectum.1,2,5
C
A unique capability of CTC over other colorectal examinations is the display of the entire abdominal and pelvic contents.6 Detection of important abdominal and pelvic extracolonic abnormalities could potentially benefit patients undergoing colorectal screening. At risk is the possible increase in the cost of this examination by the discovery and work-up of clinically insignificant findings. Several reports of extracolonic findings at CTC have been published, but none exist in a screening population that most likely might benefit from the procedure.6,7 The aims of this study were (1) to describe the spectrum of extracolonic findings in a large asymptomatic screening population undergoing CTC, and (2) to evaluate the short-term direct medical costs incurred from subsequent radiologic follow-up during that patient care visit.
Materials and Methods Seven hundred and nineteen consecutive asymptomatic patients who were prescheduled for colonoscopic screening were recruited and enrolled in this study. All patients signed an informed consent form, and the study was approved by our institutional review board. Inclusion criteria were asymptomatic patients with a family history of colorectal cancer or polyps (first-degree relative), a prior personal history of colon polyps or colorectal cancer, or the new onset of asymptomatic anemia. Exclusion criteria included patients with melena, hematochezia, inflammatory bowel disease, familial polyposis, pregnancy, recent bowel resection, or recent polyp removal within 2 months before the examination. Thirty-eight patients (a subset of patients with prior partial colon resection requiring protocol administration of intravenous contrast material and standard CT dose settings) underwent CTC and were excluded from this study. A total of 681 patients formed the basis for the study. The patients’ ages ranged from 41 to 80 years, with a median Abbreviation used in this paper: CTC, computed tomography colonography. © 2003 by the American Gastroenterological Association 0016-5085/03/$30.00 doi:10.1053/gast.2003.50158
912
GLEUCKER ET AL.
age of 64 years. Only 6 patients were under the age of 50. The male-female ratio was 426:255. Patients were placed in a decubitus position and a small enema tip was placed in the rectum. The colon was inflated gently with carbon dioxide to patient tolerance (feeling full without pain). CT scouts were obtained before each acquisition to ensure full distention and anatomic coverage of the abdomen and pelvis. Additional carbon dioxide was added before the second acquisition to ensure full colon distention. Examinations were performed on either a single (GE Highspeed Advantage; GE Medical Systems, Milwaukee, WI) (107 patients) or multidetector CT scanner (GE Lightspeed, Lightspeed Plus; GE Medical Systems) (574 patients). Single-detector scans were performed using a collimation of 5 mm, pitch of 1.3, 3-mm reconstruction interval, 70 mA, 120 kVp, 512 ⫻ 512 matrix, and 180° interpolation algorithm. Single-detector scans were performed using 20-second breath holds (3– 4) with a 3-cm overlap between acquisitions. Multidetector scans were performed using a collimation of 5 mm (4 ⫻ 5 mm detector configuration), table speed of 15 mm per rotation, high quality mode, 2.5-mm reconstruction interval, 40 mA (determined to match image noise of single-slice technique), 120 kVp, 512 ⫻ 512 matrix, and a single 20-second breath hold. Both supine and prone acquisitions were obtained in all patients. The image data was networked to a workstation using customized software developed at our institution. One of 3 board-certified and experienced (with 10 years of practice) radiologists was assigned randomly to detect and report the extracolonic findings. Image review included review of the supine axial images using lung, soft-tissue, and bone windows. None of the observers were aware of the patient’s medical history, colonoscopic findings, or the findings of the other reviewers. The extracolonic findings were separated into 3 categories of clinical importance: high, medium, and low. Findings of high importance included those requiring surgical treatment, medical intervention, and/or further investigation during that patient care visit. Examples of these included indeterminate solid organ masses, previously unknown abdominal aortic aneurysms 3 cm or larger, aneurysms of the splenic or renal arteries, indeterminate chest nodule, adenopathy (lymph nodes ⬎1 cm), pancreatic masses, and small bowel parasites. Findings of medium importance included conditions that did not require immediate treatment but would likely require investigation, recognition, or treatment at a later time. Examples of medium importance included calculi of various organs, previously known abdominal aortic aneurysms, adrenal masses (⬍2 cm in diameter), pancreatic pseudocysts, indeterminate cysts of various organs, uterine enlargement in postmenopausal women, and coronary artery calcifications. Findings of low importance were considered benign and unlikely to require further medical treatment or additional work-up. Examples of these included vascular calcifications (other than coronary artery calcification), granulomas, diverticulosis, simple solid organ cysts, small- to medium-sized hiatal hernias, and abdominal wall hernias containing fat.
GASTROENTEROLOGY Vol. 124, No. 4
A chart review was performed on all patients with an extracolonic finding(s) to determine additional radiologic follow-up for each finding. At least 12 months of follow-up was possible for each patient with a range of 410 –1513 days (mean, 931 days). The type, number, and results of these examinations were tabulated, including surgical or medical treatment outcomes relating to the detected findings. The medical records also were assessed for the development of subsequent conditions that were not detected at CTC during the same time period. The short-term direct medical cost of the follow-up radiologic tests was estimated using 2002 Medicare average reimbursement rates for professional services, and Medicare-reported nationwide mean technical and facility costs for providing those services in a hospital outpatient setting. The Resource Based Relative Value System on which Medicare reimbursement is based provides a reasonable proxy to the true opportunity cost of resources used in providing medical services.8 Patients with findings considered to be of low clinical importance had no follow-up costs because each finding with a suggested or performed follow-up examination was considered to be of either moderate or high importance. If a patient had a significant extracolonic finding, and a follow-up examination was suggested either by the referring physician or the radiologist but no further clinical information could be obtained, the cost of the suggested follow-up examination was imputed.
Results There were 858 extracolonic findings in 469 of 681 (69%) patients. No extracolonic findings were detected in 212 of 681 (31%) patients. Findings were categorized as high, medium, or low importance. Table 1 delineates the highly important extracolonic findings, follow-up radiologic examinations, and the associated costs. A total of 88 lesions in 71 of 681 (10%) patients classified as highly important were discovered. These included 34 lesions within the kidney, 27 in the chest, 8 in the liver, 6 in the ovary, 4 in the renal or splenic arteries, 3 in the retroperitoneum (other than the aorta), and one in the pancreas. A total of 94 follow-up procedures were performed to evaluate these lesions, predominantly consisting of ultrasound and CT examinations. Total follow-up radiologic costs in this group were $19,915.40. Nine of the 71 (13%) patients received treatment as a result of the CT findings. These included repair of an unsuspected abdominal aortic aneurysm, resection of a squamous cell carcinoma of the lung, chemotherapy for unsuspected thyroid metastases to the lung, resection of a renal adenocarcinoma, resection of a renal oncocytoma, resection of a serous cystadenoma of the ovary in 3 different patients, and medical treatment for ileal ascariasis. Three patients received additional radiographic follow-up of the pelvis to evaluate a cystic
April 2003
EXTRACOLONIC FINDINGS AT CT COLONOGRAPHY
913
Table 1. Highly Significant Findings, Follow-Up, and Costs
Findings at CTC
No. lesions
Follow-up radiologic examination
4
US Doppler aorta CT abdomen ⫹ pelvis/c US renal CT abdomen/c KUB CT chest w/o CT chest w/o Chest radiograph CT abdomen/c US abdomen CT abdomen/c US abdomen CT abdomen/c US abdomen US pelvis ⫹ endovaginal US abdomen CT abdomen/c US Doppler renal arteries CT abdomen/c None
Aorta
Unknown aneurysm
Kidney
Mass
34
Chest
Retrocrural LN Lung nodule
1 26
Liver
Low-attenuation area Cystic lesion Mass
1 5 2
Retroperitoneum
Adenopathy Cystic nodule Cystic lesion Cystic lesion
1 2 6 1
Ovary Pancreas Renal artery Splenic artery Small Bowel Total
Suspicious aneurysm Aneurysm Parasites
1 3 1 88
No. of follow-up procedures
Cost
Total follow-up costs
1 2 22 2 2 2 31 7 2 5 1 1 2 1 9 1 1 1 1
200.93 1154.46 169.12 584.49 219.46 466.66 466.66 71.71 584.49 223.91 584.49 223.91 1154.46 223.91 361.88 223.91 584.49 200.93 584.49
200.93 2308.92 3720.64 1168.98 438.92 933.32 14,466.46 501.97 1168.98 1119.55 584.49 223.91 2308.92 223.91 3256.92 223.91 584.49 200.93 584.49
94
34,220.64
/c, with intravenous contrast material administration; KUB, plain abdominal radiograph; LN, lymph node; US, ultrasound; w/o, without intravenous contrast material.
ovarian lesion, and one patient had a planned repeat CT scan to follow a cystic pancreatic neoplasm. In addition, there were lesions that were not followed-up at our institution, but were to be evaluated by their local physicians (indeterminate renal mass [n ⫽ 7], indeterminate lung nodules [n ⫽ 5], cystic lesion in the liver [n ⫽ 1]). For these patients, the following investi-
gations were suggested: renal ultrasound for the indeterminate renal lesions, non– contrast-enhanced CT of the chest for the indeterminate lung lesions, and ultrasound of the liver for the cystic liver lesion. The estimated cost of these follow-up examinations would be $2,247.85. Table 2 summarizes the 196 findings in 183 of 681 (27%) patients that were classified as moderately signif-
Table 2. Moderately Significant Findings, Follow-Up, and Costs
Findings at CTC Adrenal Bladder Liver Lung
Mesentery Pancreas Kidney
Aorta Gallbladder Gastric Other Total
No. lesions
Adenoma Calculi Focal lesion Fibrosis
10 3 14 1
Tiny nodules Known lung fibrosis Lymph nodes Pseudocyst Calculi Angiomyolipoma Renal calcifications Horseshoe Known aneurysm Unknown Ectasia Calculi Large hiatal hernia
3 3 1 8 66
4 19 40 8 25 197
Follow-up radiologic examination CT abdomen/c KUB CT abdomen/c CT chest w/o Chest x-ray None None CT abdomen/c CT abdomen ⫹ pelvis w/o KUB None None None US Doppler aorta US Doppler aorta None None None
No. of follow-up procedures
Cost
Total follow-up costs
1 2 2 1 1
584.49 219.49 584.49 466.66 71.71
584.49 438.98 1168.98 466.66 71.71
1 1 1
584.49 941.21 219.49
584.49 941.21 219.49
1
200.93 200.93
200.93 803.72
0
16
5704.57
914
GLEUCKER ET AL.
GASTROENTEROLOGY Vol. 124, No. 4
Table 3. Low Significant Findings Findings
No. Patients
Vascular calcifications Diverticulosis Simple renal cysts Simple liver cysts Small- or medium-sized hiatal hernia Lung granuloma Prostate calcifications Liver granuloma Calcified uterine fibroid Spleen granuloma Mild fibrotic changes in lung bases Accessory spleen Calcifications mitral valve Benign bony island Other Total
107 81 56 55 55 37 36 33 28 26 21 14 8 7 10 574
icant: abnormalities of the kidneys (n ⫽ 66), 64 of these owing to renal calculi; abnormalities of the abdominal aorta (known aneurysm ⬎3 cm or ectasia 2–3 cm in diameter) (n ⫽ 23); liver lesions (n ⫽ 14); adrenal masses (n ⫽ 10); pancreatic pseudocysts (n ⫽ 8); lung abnormalities (n ⫽ 7); and bladder calculi (n ⫽ 3). One patient with a large hiatal hernia causing significant reflux esophagitis had a laparoscopic repair of the hernia. Many patients received treatment for renal and bladder calculi; however, imaging follow-up was not long enough to assess the outcome of these therapies, and many patients received further care by their home physicians without follow-up records. Only 15 of the 196 (7.7%) findings classified as moderate significance were followed-up with additional documented radiologic imaging tests at a total cost of $3,465.19. It is difficult to estimate the number of additional tests that would be ordered as a result of these findings, but certainly many of the patients with renal and bladder calculi, abdominal aortic aneurysms, and tiny nodules in the lung bases would, in all likelihood, receive additional imaging tests in the future. Five hundred and seventy-four findings in 341 of 681 (50%) patients were designated to be of low significance (Table 3) and by definition had no additional follow-up testing and associated costs. The most common findings included vascular calcifications, diverticulosis, benignappearing cysts within the liver and kidneys, small- and medium-sized hiatal hernias, and calcified granulomas. A total of 44 lesions were not reported at CTC examination but described in subsequent radiologic examinations or in the clinical charts. Among them, 11 of 44 (25%) could be detected retrospectively, 33 (75%) not. Among these 44 lesions were 3 findings of high importance: 1 pancreatic adenocarcinoma, 1 splenic mass, and
1 ovarian mass. The pancreatic and the splenic lesion could not be detected retrospectively. The ovarian mass could be detected retrospectively at CTC. Twenty lesions of medium importance were not described at CTC, among them 8 renal stones, 1 ileal lipoma in proximity of the ileocecal valve, 1 adrenal adenoma, 2 angiomyolipomas of the kidney, 5 fatty livers, 1 fatty pancreas, 1 complicated renal cyst, and 1 gallstone. Among them, 2 renal stones, 1 ileal lipoma, and 1 adrenal adenoma could be found retrospectively. The remaining 17 lesions of medium importance in 17 patients could not be identified retrospectively. Twenty-one subsequently identified lesions were of low importance, including 10 renal cysts (3 detected retrospectively), 8 liver cysts (2 detected retrospectively), and 3 hepatic hemangiomas (1 detected retrospectively). Fifteen of these 21 lesions (71%) were not identified retrospectively.
Discussion As CTC undergoes scrutiny as a candidate approach to colorectal cancer screening, the balance between potential benefits and potential harms or costs of this new technology must be considered critically. Incidentally discovered extracolonic findings represent a unique feature of CTC that sets it apart from other approaches to colorectal cancer screening. This feature may lead to added benefit via early detection of unsuspected health-threatening pathology but may do so at the cost incurred from downstream interventions and unanticipated iatrogenic morbidity. The present study was undertaken to explore such benefits and costs of extracolonic lesion detection by CTC in an asymptomatic population being screened for colorectal cancer. Based on our study, approximately two thirds of asymptomatic individuals undergoing screening CTC will have unsuspected extracolonic lesions detected. Most of these findings will be inconsequential and not require additional imaging studies or other interventions. However, approximately 10% of our patients had a finding considered to be important clinically and 13% of these patients (1.3% of the study population) received surgery or medical treatment as a result of the radiologic discovery. Extracolonic abnormalities were found in nearly every organ of the abdomen and pelvis. Abnormalities in the kidneys, ovaries, lung, and aorta constituted the organs with the most prevalent findings of significance. It is not possible to determine if mortality reduction actually occurred. Following the approach of Sonnenberg et al.,9 we estimated the average cost of CTC based on 2002 Medicare professional reimbursement and mean technical and
April 2003
facility cost for CT of the abdomen with and without contrast plus 3-dimensional reconstruction, totaling $560.31. The overall short-term cost for the additional radiologic examinations was $23,380.59 for the 681 patients evaluated, thus adding $34.33 (6%) to the average cost of the examination. By examining only short-term radiologic costs, we are necessarily adopting a limited perspective for the cost analysis. Other direct and indirect resource costs and savings10 were not studied. Such costs and savings would include the value of patient and informal caregiver time, and resources devoted to subsequent evaluation and treatment or other nonradiologic procedures incurred or foregone as the result of the extracolonic CTC findings. Nonmonetary benefits and harms (including short-term patient anxiety and concern, as well as long-term morbidity, mortality, and quality-of-life outcomes) associated with follow-up and treatment of extracolonic findings also were not considered. The awareness of findings of even medium or low importance might create distress for the patient such that the referring physician must initiate further work-up for these findings. This could result in additional costs and morbidity. In one study addressing psychologic side effects of breast cancer screening, women with suspicious mammography findings showed significantly elevated mammography-related anxiety, despite the fact that a malignant lesion was ruled out by follow-up studies.11 Investigating this issue was outside of the scope of this study. Because colorectal screening is now advocated for all adults (average-risk patients over age 50, and higher-risk patients over age 40), CTC presents a unique opportunity to assess the abdomen and pelvis for other diseases that affect this older population. Successful treatment of important asymptomatic conditions could provide unexpected cures and reduced morbidity in selected patients. Anticipated prevalence of large colorectal adenomas (ⱖ1 cm) is only 5% to 8% in a screening population.12–14 In this study, 10% of patients were identified with extracolonic findings considered highly significant. A total of 1.5% of all patients underwent treatment (either surgical or medical) for these unsuspected findings. In view of the low prevalence of significant colorectal adenomas in this patient group, these additional findings and their modest incremental costs may be beneficial and potentially could add considerable value to the examination. This study did not address the cost effectiveness of CTC in light of these findings. Additional outcomes-related studies with longitudinal follow-up of these patients would be helpful in the future.
EXTRACOLONIC FINDINGS AT CT COLONOGRAPHY
915
Several studies have been published regarding the prevalence of incidental findings at CTC, including one from our institution.6 Hara et al.6 found extracolonic findings in 41% of 264 patients studied. Eleven percent of the patients had findings considered highly clinically significant, including 6 patients (2.3%) who underwent surgery for these unsuspected abnormalities. The majority of the patients in this study had previously diagnosed colon lesions and, therefore, were not representative of a screening population. Edwards et al.7 reported on extracolonic findings in 100 patients over the age of 55 with an overall prevalence of extracolonic abnormalities of 15%. Two of these 15 (1.3%) patients underwent surgery for benign ovarian cysts. Swensen et al.15,16 reported the results of low-dose CT scanning (screening) of the chest and abdomen in patients at high risk for lung cancer (cigarette smokers). Two hundred and ten of 1520 (14%) of these patients had significant findings other than lung lesions, including abdominal aortic aneurysms in 3.3%.15 Because cigarette smoking predisposes to a number of extrapulmonary diseases, it is difficult to compare these findings with those in our study. Hara et al.6 in 2000 also found that the additional imaging studies added approximately $28 per patient to the cost of the examination. Our average cost of $34.33 is slightly higher, in part owing to a higher overall prevalence of extracolonic findings (69% vs. 41%). Limitations of this technology include the failure to detect important abnormalities at CTC. The low-dose technique used at CTC capitalizes on the high contrast that exists between the air-filled colon lumen and the soft-tissue density wall. Polyps protruding into the airfilled lumen can be detected using these low-dose techniques. Solid-organ contrast requires higher radiation doses for optimal lesion detection. In addition, lesionorgan contrast differences (and lesion detection) can be improved with the use of intravascular iodinated contrast material. None of the patients in this study received intravenous contrast material, and all examinations were performed using a low-dose technique. As a result, it is likely that some solid organ lesions were not detected. Follow-up of our patients for an average time interval of 21 months revealed additional lesions that were detected with other subsequent examinations. Among them were 3 lesions of high importance, 1 adenocarcinoma of the pancreas, 1 splenic, and 1 ovarian mass. The pancreatic and the splenic mass could not be detected retrospectively. Radiologists and referring physicians need to be aware that assessment of solid organs is a potential limitation of this procedure. To decrease patient dose, there are currently studies investigating if the radiation
916
GLEUCKER ET AL.
dose could be further decreased without compromising the detection of polyps larger or equal to 5 mm.17 One study showed that by lowering the dose to 30 mAs, the detection of colonic lesions remains unimpaired.17 However, further reduction of the radiation dose would increase imaging noise. This is likely to compromise the evaluation of extracolonic parenchymal organs (i.e., liver, pancreas, spleen). It is possible that an abdominal CT scanner with further dose reduction might be insufficient for the detection of extracolonic findings and should be subject to further studies. Clinical problems related to solid organs are studied optimally using conventional CT technique. The CTC examination is optimized for the detection of colonic lesions but, according to its inherent low-dose technique, not designed for the detection of extracolonic lesions. The medicolegal implications of this limitation have not yet been explored fully. The results of this study should not be construed as an endorsement for whole-body screening. As previously stated, the low-dose technique (without intravenous contrast material administration) has limitations for the detection of lesions in solid organs. The authors are not aware of any national organization recommending wholebody scanning, primarily because sufficient evidence supporting this practice change is lacking. We hope this study contributes to a better understanding of abdominopelvic screening in an asymptomatic older population. In conclusion, extracolonic abnormalities are exceedingly common in asymptomatic populations undergoing screening by CTC using a low-dose technique. These incidental extracolonic findings can be categorized according to their clinical relevance; and, based on our observations, roughly 10% could be considered clinically important and would require further evaluation or intervention. The average additional radiologic cost of added imaging studies is approximately $34 per patient across the whole study group, although this likely does not capture the true resource cost or benefit from a societal perspective. The potential benefits of detecting extracolonic abnormalities at CTC include the discovery of important asymptomatic disease with the possibility of early treatment and cure. Radiologists and clinicians must be cognizant of the high prevalence of extracolonic findings found at CTC, and make efforts to triage patients appropriately for additional tests in a fiscally responsible fashion.
GASTROENTEROLOGY Vol. 124, No. 4
2. Fletcher JG, Johnson CD, Welch TJ, MacCarty RL, Ahlquist DA, Reed JE, Harmsen WS, Wilson LA. Optimization of CT colonography technique: prospective trial in 180 patients. Radiology 2000; 216:704 –711. 3. Hara AK, Johnson CD, MacCarty RL, Welch TJ, McCollough CH, Harmsen WS. CT colonography: single- versus multi-detector row imaging. Radiology 2001;219:461– 465. 4. Reed J, Johnson C. Automatic segmentation, tissue characterization, and rapid diagnosis enhancements to the computed tomographic colonography analysis workstation. J Digit Imaging 1997; 10(Suppl):70 –73. 5. Yee J, Akerkar GA, Hung RK, Steinauer-Gebauer A, Wall SD, McQuaid KR. Colorectal neoplasia: performance characteristics of CT colonography for detection in 300 patients. Radiology 2001;219:685– 692. 6. Hara AK, Johnson CD, MacCarty RL, Welch TJ. Incidental extracolonic findings at CT colonography. Radiology 2000;215:353– 357. 7. Edwards JT, Wood CJ, Mendelson RM, Forbes GM. Extracolonic findings at virtual colonoscopy: implications for screening programs. Am J Gastroenterol 2001;96:3009 –3012. 8. Lave JR, Pashos CL, Anderson GF, Brailer D, Bubolz T, Conrad D, Freund DA, Fox SH, Keeler E, Lipscomb J, Luft HS, Provenzano G. Costing medical care: using Medicare administrative data. Med Care 1994;32:JS77–JS89. 9. Sonnenberg A, Delco F, Bauerfeind P. Is virtual colonoscopy a cost-effective option to screen for colorectal cancer? Am J Gastroenterol 1999;94:2268 –2274 10. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. Oxford: Oxford University Press, 1996. 11. Lerman C, Trock B, Rimer BK, Jepson C, Brody D, Boyce A. Psychological side effects of breast cancer screening. Health Psychol 1991;10:259 –267. 12. O’Brian MJ, Winawer SJ, Zauber AG, Gottlieb LS, Sternberg SS, Diaz B, Dickersin GR, Ewing S, Geller S, Kasimian D, Komorowski R, Szoporn A. The national polyp study. Patient and polyp characteristics associated with high-grade dysplasia in colorectal adenomas. Gastroenterology 1990;98:371–379 13. DiSario JA, Fouth PG, Mai HD, Pardy K, Manne RK. Prevalence and malignant potential of colorectal polyps in asymptomatic, average-risk men. Am J Gastroenterol 1991;86:941–945. 14. Cannon-Albright LA, Bishop DT, Samowitz W, DiSario JA, Lee R, Burt RW. Colonic polyps in an unselected population: prevalence, characteristics, and associations. Am J Gastroenterol 1994;89: 827– 831. 15. Swensen SJ, Viggiano RW, Midthun DE, Mueller NL, Sherrick A, Yamashita K, Naidich DP, Patz EF, Hartman TE, Muhm JR, Weaver AL. Lung nodule enhancement at CT: multicenter study. Radiology 2000;214:73– 80. 16. Swensen SJ, Jett JR, Sloan JA, Midthun DE, Hartman TE, Sykes AM, Augenbaugh GL, Zink FE, Hillman SL, Noetzel GR, Marks RS, Clayton AC, Pairolero PC. Screening for lung cancer with low-dose spiral computed tomography. Am J Respir Crit Care Med 2002; 165:508 –513. 17. vanGelder RE, Venema HW, Serlie IWO, Yung Nio C, Determann RM, Tipker CA, Vos FM, Glas AS, Bartelsman JFW, Bossuyt PMM, Larneris JS, Stocker J. CT colonography at different radiation dose levels: feasibility of dose reduction. Radiology 2002;224: 25–33.
References 1. Fenlon HM, Nunes DP, Schroy III PC, Barish MA, Clarke PD, Ferrucci JT. A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps. N Engl J Med 1999;341: 1496 –1503.
Received August 27, 2001. Accepted January 9, 2003. Address requests for reprints to: C. Daniel Johnson, M.D., Department of Radiology, Mayo Clinic Rochester, 200 First Street SW, Rochester, Minnesota 55902.