| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Emergency Radiology |
1 From the Depts of Radiology (D.S.D., S.M.B., J.D.D., P.D.R., D.A.K.) and Surgery (R.C.M.), Div of Emergency Medicine (B.H.), University of Colorado Health Sciences Center, Box A030, 4200 E 9th Ave, Denver, CO 80262; Depts of Surgery (E.E.M.) and Radiology (M.F.M., M.J.H., D.L.S., J.E.), Div of Emergency Medicine (E.J.R.), Denver Health Medical Center, Colo; and Dept of Biostatistics, National Jewish Medical and Research Center, Denver (D.N.I.). Received Jul 20, 1998; revision requested Aug 28; revision received Dec 14; accepted Mar 26, 1999. Supported in part by a grant from the Society for Thoracic Radiology Research and Education Fund. Address reprint requests to D.S.D. (e-mail: Deb.Dyer@UCHSC.edu).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Patients in whom there was very high suspicion of traumatic aortic injury were examined with aortography only. Other patients were examined with contrast material–enhanced CT. Follow-up aortography was performed in all patients with moderate to high suspicion of traumatic aortic injury and in all patients with CT scans that were positive for traumatic aortic injury. CT scans were regarded as positive when they showed mediastinal hematoma or direct findings of aortic injury. During a 4
-year period, 1,009 patients (263 female, 746 male; age range, 3–90 years) were evaluated for possible traumatic aortic injury.
RESULTS: Of the 207 patients who underwent aortography directly without CT, 10 had traumatic aortic injury. Of the 802 patients who were examined with CT, 382 underwent follow-up aortography. In this group, there were 10 true-positive and no false-negative CT scans. CT had 100% sensitivity and a 100% negative predictive value for the detection of traumatic aortic injury.
CONCLUSION: Chest CT can be used to exclude aortic injury.
Index terms: Aorta, CT, 948.12912, 948.12915 • Aorta, injuries, 948.412, 948.413, 948.732 • Aortography, 948.1211 • Thorax, CT, 60.12112, 60.12115 • Trauma
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Evaluation for traumatic aortic injury begins with an assessment of the mechanism of injury, a physical examination, and chest radiography. An acute deceleration injury will lead the trauma physician to initially suspect traumatic aortic injury. The most common settings for traumatic aortic injury include motor vehicle accidents, motorcycle accidents, auto-pedestrian accidents, and falls. The physical examination is not very useful in the hemodynamically stable patient. Although portable chest radiography performed with the patient in the supine position is sensitive for the detection of mediastinal hemorrhage, it is not specific for the detection of traumatic aortic injury and leads to a large number of negative aortograms. Although aortography, which has been the reference-standard procedure for the diagnosis of aortic injury, is relatively safe, it is expensive, time-consuming, resource intensive, and invasive.
In recent years, computed tomography (CT) of the chest has been promoted as a useful screening and diagnostic tool. Proponents of this modality have emphasized its wide availability, speed, and ability to depict other injuries, and the advantage of being able to perform other examinations (eg, head CT) while the patient is in the scanner (3–9). CT is considered to be less invasive, faster, safer, simpler, cheaper, and less resource intensive than aortography. CT is often more readily accessible because most trauma centers and larger hospitals now have in-house CT technologists that are available 24 hours a day (10). The role of CT in recognizing aortic injury has been described, but few studies have addressed its accuracy, positive predictive value, and negative predictive value. Many investigators (3–9) have reported that CT is an effective screening tool and that it can help reduce the need for aortography. Yet none of these studies involved consistent angiographic follow-up of the reported "negative" (ie, "normal") chest CT scans.
To be relied on to help exclude traumatic aortic injury, chest CT must be shown to be highly sensitive and to have a high negative predictive value in the detection of this condition. The purpose of this study was to determine whether chest CT can be relied on to help exclude traumatic aortic injury in blunt chest trauma.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
The study design and protocol were approved by the Colorado Multiple Institutional Review Board. Written informed consent for participation in the study was obtained from all able patients after the CT procedure was fully explained. If the patient was not able to provide informed consent, then it was obtained from a family member, or "emergency consent" was documented on the chart by the trauma physician.
During the first half of the study (1991–1993), conventional CT scans were obtained by using GE9800 scanners (GE Medical Systems, Milwaukee, Wis). During the second half of the study (1994–1995), helical CT scans were obtained by using CT HiSpeed Advantage scanners (GE Medical Systems). All CT examinations were performed with adherence to a strict protocol. Ionic (Hypaque 60 [diatrizoate meglumine 60%]; Nycomed Amersham, Princeton, NJ) or nonionic (Omnipaque 300 [iohexol 60%]; Nycomed Amersham or Optiray 240 [ioversol 51%]; Mallinckrodt, St. Louis, Mo) intravenous contrast material was administered. For conventional CT scanning, 150 mL of contrast material was injected; for helical CT scanning, 100 mL was injected. If abdominal CT was required, then 150 mL of contrast material was administered intravenously. For conventional CT, the chest was scanned in three sets as follows: 10 x 10 mm, from the neck to the top of the arch; 5 x 5 mm, from the top of the arch to the bottom of the left hilum; and 10 x 10 mm, from the bottom of the left hilum through the hemidiaphragms. For helical CT, the chest was scanned at a pitch of 1:1 by using three helical sets—that is, with 10-mm collimation, 5-mm collimation, and 10-mm collimation—that were obtained through the same regions as those scanned with conventional CT. The interval between the beginning of the injection and the start of scanning was usually between 15–25 seconds, but it was shorter when tachycardia was present.
The CT scans were read prospectively. The CT scans obtained after hours were read by the radiology resident on call (S.M.B. was one of several on-call radiology residents) and reviewed by an attending radiologist (D.S.D., M.F.M., M.J.H., P.D.R., D.L.S., and J.E. were among the several attending radiologists) the next morning. The CT scans obtained during regular hours were read by an attending radiologist. Radiology residents received annual training in how to perform CT and interpret the findings for potential aortic injury.
The CT scans were scored as positive if any of the following criteria were present: poorly defined fat planes, mediastinal hemorrhage, perivascular hemorrhage, periaortic hematoma, change in caliber of the aorta (ie, acute coarctation), intraluminal irregularity (ie, intimal flap), abnormal contour of the aorta, or abnormal contour of the proximal great vessels. The first four signs are indicative of mediastinal hematoma but not direct evidence of aortic injury. Caliber change of the aorta, intraluminal irregularity, and abnormal contour of the aorta are considered to be direct signs of aortic injury. This scoring system was designed with considerable overlap in the findings to optimize the identification of any abnormalities. The CT scans were also evaluated for image quality. The final CT-based diagnosis was one of the following: suboptimal, equivocal for aortic injury, positive for aortic injury, or no evidence of aortic injury.
Aortography was performed by using a standard intraarterial digital subtraction technique. By using a 6- or 7-F pigtail catheter placed into the ascending aorta, 50–60 mL of ionic (Hypaque 60 or Hypaque 76 [diatrizoate meglumine 76%]; Nycomed Amersham) or nonionic (Omnipaque 300, Omnipaque 350 [iohexol 75.5%]; Nycomed Amersham, or Optiray 320 [ioversol 68%]; Mallinckrodt) contrast material was injected intravenously at a rate of 30–40 mL/sec, and the images were obtained in the right anterior oblique and left anterior oblique projections.
The CT, aortographic, and operative findings (when available) were compared. Statistical analyses were performed by using two-by-two contingency tables. Test properties were described by using proportions with 95% CIs.
A cost analysis based on the technical costs and professional fees for the CT and aortographic examinations performed in this study was performed retrospectively. The University Hospital 1997 charge master (technical costs) and professional fee schedule (charges) were used to determine the various costs. The cost of performing follow-up aortography in all the patients who were examined with CT was compared with the cost of performing CT and follow-up aortography in all the patients who had positive CT scans. In addition, the cost of performing CT plus follow-up aortography in only those patients with direct findings of aortic injury at CT was compared with the cost of performing CT plus follow-up aortography in the patients with periaortic hematoma and/or direct findings of aortic injury at CT.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In the 802 patients who underwent CT initially, the CT findings were negative in 638 (80%), and positive in 152 (19%) patients. Suboptimal scans were obtained in 140 patients, but only 12 (2%) scans were considered to be nondiagnostic. Nondiagnostic scans were obtained in seven patients because they were allergic to iodine and thus received no intravenous contrast material. Five other patients had nondiagnostic scans because of poor timing of the contrast material bolus. Ten of the 12 patients with nondiagnostic scans underwent aortography, and the results were negative; two underwent magnetic resonance imaging and the results were negative. With the exception of the costs of the five nondiagnostic, technically poor scans, the results of the nondiagnostic scanning examinations are excluded from the CT data. The frequency of CT abnormalities on the positive CT scans are shown in Table 2. Anterior mediastinal hemorrhage was the only abnormality observed in 103 patients. Mediastinal hematoma with periaortic hematoma was identified in 40 patients. Direct signs of aortic injury were observed in 25 patients.
|
|
|
|
|
|
|
|
|
Cost Analyses
The cost analysis for aortic imaging is shown in Table 4. The calculations are based on a cost of $1,786 for thoracic aortography ($964 in technical costs + $822 in professional fees) and a cost of $468 for contrast material–enhanced helical CT ($299 in technical costs + $169 in professional fees). Performing aortography in all 1,009 patients would have cost $1,802,074; performing aortography in the 795 patients who underwent CT would have cost $1,419,870. The total cost of CT scanning is based on 795 patients rather than 802 patients because the nondiagnostic nonenhanced CT scans obtained in the seven patients who were allergic to the contrast material were excluded; these patients did not meet the study protocol and the examinations should not have been included in the study. The cost of follow-up aortography in the five patients with nondiagnostic, poor-quality CT scans (because of poor bolus timing) was $8,930. The cost of performing CT in all 795 patients who underwent CT was $372,060. Performing follow-up aortography in all 152 of the patients who had positive CT scans would have cost $271,472; in only the 25 patients who had direct CT findings of aortic injury, $44,650; and in the 49 patients with CT scans showing periaortic hematoma and/or direct CT findings, $87,514.
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In this study, at least 638 (80%) of 795 patients (Table 5) would have been spared aortography by undergoing CT first. One hundred ninety-three (93%) of the 207 patients in whom the suspicion of traumatic aortic injury was high and who were examined with only aortography had normal aortograms. Performing CT probably would have enabled aortography to be avoided in a number of these patients as well. These results are quite different from those of Fisher et al (13), who found that about 25% (18 of 72) of the patients who were examined with CT could have avoided undergoing follow-up aortography. Possible explanations for this difference include the use, in the Fisher et al study, of a conventional CT scanner with 1-cm collimation, CT scan scoring system that did not specify mediastinal hematoma by location or focus on the direct signs of aortic injury, and study population that may have had a higher prevalence of disease. The results of our study compare favorably with those of others (3–9,14) who have found that CT can obviate aortography in a substantial number of patients.
Although there were no false-negative CT scans in our study, they have been reported in previously published articles (15–17). Close scrutiny of these studies, however, revealed inconsistent or variable scanning techniques. In the study by McLean et al (15), two false-negative CT scans were reported; the procedures were performed at outside hospitals, and there was no mention of the scanning technique or of the criteria used to interpret the scans. In the study by Miller et al (16), five false-negative CT studies were described. It is important to note that these scans were obtained by using an older model scanner, 75 mL of contrast material, and 10-mm collimation. In one case, intravenous contrast material was not administered and the images were degraded by artifact. Another case was thought to be positive in retrospect. The other three cases were branch vessel injuries; two of these were vertebral arterial injuries that were not treated, and the other was a subclavian arterial injury that was apparent clinically.
Tomiak et al (17) reported three false-negative CT studies. These "negative" readings, however, were questionable, because the scans were obtained with older model scanners and 50 mL of intravenous contrast material. Two of these scans showed periaortic hemorrhage, and one was degraded by motion and artifact (17). Scans with substantial motion or streak artifacts should not be considered to be of diagnostic value. Unless the scan is minimally compromised, the CT examination should be repeated or the patient should undergo aortography. Certainly, a poor-quality scan cannot be relied on to help exclude aortic injury.
Mirvis et al (8) and others have suggested that localizing the injuries in the mediastinum can be quite helpful in increasing the specificity of CT and decreasing the number of false-positive studies. The results of our study support this concept and suggest that we need only be concerned about the direct signs of aortic injury. However, in the 6 months after the study ended, a patient with a small posterior pseudoaneurysm that was identified at aortography had an initial CT scan that demonstrated periaortic hematoma but no obvious direct signs of aortic injury (Fig 5). Even in retrospect, no pseudoaneurysm or intimal flap could be identified; however, a subtle caliber change (pseudocoarctation) was apparent. As a result of this case, we reanalyzed our study data by using the presence of periaortic hematoma and/or direct signs of aortic injury as the criteria for a positive CT scan. The reanalyzed data are shown in Table 3. Although the inclusion of periaortic hematoma in the criteria caused the number of false-positive CT scans to increase, the specificity to decrease, and the positive predictive value to decrease, it preserved the sensitivity and negative predictive value, both of which remained at 100%. Ishikawa et al (14) also noted the importance of periaortic hematoma, identifying it as the most common and most suggestive finding of aortic rupture.
|
|
|
|
|
|
It may be of some concern that aortograms were not obtained in all patients who underwent CT. At one trauma center, all of the patients underwent follow-up aortography; at the other center, only the patients with higher mechanisms of injury underwent follow-up aortography. This was mainly because at that one center, for institutional review board approval, confirmatory aortography was required for all patients who underwent CT. At the other center, institutional review board approval was not required for the study because CT had become the "standard of care." Because some of the physicians at this center were still uncomfortable with CT, a compromise, in which follow-up aortography was performed only in the patients with higher mechanisms of injury, was reached.
Only one branch vessel injury was found in the group that underwent aortography directly. No branch vessel injuries were found in the patients who were examined with CT and follow-up aortography during the study period. We have no explanation for this very low prevalence of branch vessel injuries; perhaps some of these injuries were undetected. Considering the lack of branch vessel injuries in our study group, the usefulness of CT in detecting these injuries remains unconfirmed.
Performing CT as the initial imaging study would have resulted in substantial cost savings. The use of CT with follow-up aortography in only those patients who had periaortic hematoma and/or direct signs of aortic injury would have saved nearly $1 million. It is unknown how many of the patients who underwent imaging with only aortography might have had negative CT scans, but it can be assumed that some of them could have avoided undergoing aortography, with the result of additional cost savings.
Furthermore, the increasing availability of helical CT in rural facilities may help to avoid the unnecessary and risky emergent transporting of patients who are suspected of having aortic injury. For example, although several hospitals in many of Colorado's rural and resort communities are not equipped for aortography or thoracic surgery, they offer helical CT scanning 24 hours a day. With our increasing confidence in CT as the definitive screening examination for possible traumatic aortic injury, many patients are being spared from being urgently transported by air on the basis of negative scans. Decreasing the need to transport patients helps to avoid the risks associated with transporting both patients and personnel, particularly in bad weather, improves patient comfort, and has considerable financial advantages.
On the basis of our study results, we recommend the use of chest CT as a screening and diagnostic tool in the evaluation of potential aortic injury in hemodynamically stable patients. In addition, we recommend that CT scans be interpreted with the following guidelines (Fig 8): (a) When the mediastinum and aorta are normal or isolated anterior mediastinal hematoma is present, no further evaluation is necessary because traumatic aortic injury has been reliably excluded; (b) when periaortic hematoma is present, the examination is suspicious for traumatic aortic injury and aortography should be performed; (c) when there are equivocal direct signs of traumatic aortic injury, aortography should be performed; and (d) when unequivocal direct signs of traumatic aortic injury are present, either confirmatory aortography or immediate surgery should be performed.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Sammer, E. Wang, C. C. Blackmore, T. R. Burdick, and W. Hollingworth Indeterminate CT Angiography in Blunt Thoracic Trauma: Is CT Angiography Enough? Am. J. Roentgenol., September 1, 2007; 189(3): 603 - 608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M R Jones and J H Reid Emergency chest radiology: thoracic aortic disease and pulmonary embolism Imaging, September 1, 2006; 18(3): 122 - 138. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Alkadhi, S. Wildermuth, L. Desbiolles, T. Schertler, D. Crook, B. Marincek, and T. Boehm Vascular Emergencies of the Thorax after Blunt and Iatrogenic Trauma: Multi-Detector Row CT and Three-dimensional Imaging RadioGraphics, September 1, 2004; 24(5): 1239 - 1255. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wong, M. B. Gotway, A. D. Sasson, and R. B. Jeffrey Periaortic Hematoma at Diaphragmatic Crura at Helical CT: Sign of Blunt Aortic Injury in Patients with Mediastinal Hematoma Radiology, April 1, 2004; 231(1): 185 - 189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. E. Lee, B. Arslan, R. Queiroz, and D. L. Waldman Assessment of Inter- and Intraobserver Agreement between Intravascular US and Aortic Angiography of Thoracic Aortic Injury Radiology, May 1, 2003; 227(2): 434 - 439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. W. Vane, M. S. Keller, K. H. Sartorelli, and A. P. Miceli Pediatric Trauma: Current Concepts and Treatments J Intensive Care Med, September 1, 2002; 17(5): 230 - 249. [Abstract] [PDF]
|
|
|
|
|
|
M. S. Parker, T. L. Matheson, A. V. Rao, C. D. Sherbourne, K. G. Jordan, M. J. Landay, G. L. Miller, and J. A. Summa Making the Transition: The Role of Helical CT in the Evaluation of Potentially Acute Thoracic Aortic Injuries Am. J. Roentgenol., May 1, 2001; 176(5): 1267 - 1272. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Harris Jr Reflections: Emergency Radiology Radiology, February 1, 2001; 218(2): 309 - 316. [Abstract] [Full Text]
|
|
|
|
|
|
W. W. Mayo-Smith Administration of a CT Division Radiology, February 1, 2002; 222(2): 319 - 326. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
