| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Medical College of Wisconsin, 9200 W Wisconsin Ave.,
Milwaukee, WI 53226-3596.
2 Department of Research, St. Joseph Mercy Hospital-Oakland, Pontiac, MI.
3 Department of Medicine, Wayne State University, Detroit, MI.
4 Office of the Vice President, Weill Cornell Medical College, New York,
NY.
5 Office of the Vice President, Methodist Hospital, Houston, TX.
6 Department of Surgery, University of Michigan, Ann Arbor, MI.
7 Department of Radiology, Washington University, St. Louis, MO.
8 Department of Radiology, Weill Cornell Medical College, New York, NY.
Received January 5, 2007;
accepted after revision March 28, 2007.
Address correspondence to L. R. Goodman
(lgoodman{at}mcw.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. In PIOPED II, CT venography followed CT angiography (CTA) to detect pulmonary embolus, using 7.5-mm continuous helical imaging from the iliac crest to the tibial plateau. DVT was detected in 105 of 737 patients (14.2%). We randomly chose 54 positive cases and 96 negative cases for our study. The continuous helical images were reformatted as 7.5-mm images and two of every three images were deleted. These images (7.5 mm; skip = 15 mm) were then sent—without identifying information—to the original reviewers. From 1 to 3.5 years had elapsed since the original interpretations. The results of the new interpretations were compared with the original CT venography consensus interpretations of PIOPED II.
RESULTS. There was agreement for the presence of DVT in at least one
leg (same leg) or for the absence of DVT in both legs in 133 of the 150 study
patients (89%). The kappa statistic showed substantial agreement between the
consensus interpretations and the test interpretations (
= 0.75; 95% CI
= 0.64-0.86) per patient.
CONCLUSION. There was good—but not perfect—agreement between continuous helical and discontinuous axial imaging for the detection of DVT. Given the vagaries of interobserver and intraobserver variation, there appears to be little difference between the two approaches. Adopting discontinuous imaging and other dose-reduction strategies can reduce pelvic radiation by more than 75%.
Keywords: CT venography • deep venous thrombosis • emergency radiology • PIOPED II • pulmonary embolism • thromboembolic disease • venous thromboembolism
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Other investigators have reported that B-mode compression sonography showed DVT in 70 of 162 patients (43%) with PE diagnosed on the basis of a high-probability ventilation-perfusion scan and in 35 of 60 patients (58%) diagnosed on the basis of a high-probability ventilation-perfusion scan combined with a high clinical probability [6]. Cham et al. [7] detected DVT using CT in 100 (41%) of 243 patients with a CT-proven PE. In PIOPED II, DVT was diagnosed using CT venous phase imaging (CT venography) in 105 of 192 (55%) patients with a PE diagnosed by a composite reference diagnosis [8]. The apparent limited sensitivity of both compression sonography and CT venography for the detection of DVT may partially reflect the fact that many of these patients may have had small, residual, asymptomatic nonobstructing DVTs, which are difficult to detect, or that the thrombi had already embolized to the lung at the time of venous imaging.
With the increasing use of CT angiography (CTA) to detect pulmonary embolus [9], many investigators have advocated the routine use of CT venous phase imaging to identify lower extremity thrombi in patients with suspected PE [7, 10-15]. The combination of CTA and CT venography provides a set of imaging studies that can be obtained in approximately 20 minutes, thereby considerably simplifying the workup for the clinician, the radiologist, and the patient. This short examination time is especially advantageous in an emergency department where time is of the essence and followup is often difficult [16]. The addition of CT venography in PIOPED II increased the sensitivity for the detection of PE (patients requiring anticoagulation) by 7%, from 83% to 90% [8]. Other estimates of CT venography's contribution to increasing sensitivity for the diagnosis of venous thromboembolism (VTE) vary from 1% to 27% [7, 13, 17-19].
In PIOPED II, the increased sensitivity for VTE detection with the combined use of CT venography with CTA led the PIOPED II investigators to recommend this combination as the imaging protocol of choice for most patients [20]. However, Perrier and Bounameaux [19], in an editorial accompanying the PIOPED II article, came to the opposite conclusion: "CTV [CT venography] does not appear to improve the diagnostic yield of CTA enough to justify the additional radiation."
Although the value of CT venography in this setting is contentious, all agree that it produces increased somatic and gonadal radiation, which is of particular importance in patients of child-bearing age or younger [20, 21]. Loud et al. [10, 11] suggested that discontinuous axial images may have a diagnostic yield equivalent to that of continuous helical images for the detection of DVT yet reduce pelvic radiation. They postulated that DVTs are usually several centimeters long and should be diagnosable with discontinuous imaging. Using CT venography, Cham et al. [7] found that only six of 100 (6%) lower extremity thrombi measured 2 cm, 12 (12%) measured 3 cm, and the remaining thrombi were 4 cm or longer. Similarly, Garg et al. [14] found that 98% of thrombi were more than 2 cm long in a study using a 5-mm collimation and 15-mm skip intervals. Those investigators showed that only 2% of DVTs were visible on only one axial image [14]. However, the accuracy of discontinuous CT for the detection of DVT has not been studied directly, to our knowledge. The present study was undertaken to determine whether CT venography images obtained with a 7.5-mm collimation at 15-mm intervals are equivalent to continuous CT venography images obtained with a 7.5-mm collimation for the diagnosis of DVT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT scans were obtained using 4-MDCT scanners in 691 patients, 8-MDCT scanners in 37 patients, and 16-MDCT scanners in 45 patients. Low-osmolar nonionic contrast material (135 mL) was injected through an arm vein at 4 mL/s. Patients who weighed more than 250 lb (112.5 kg) received 150 mL of contrast material. The CTA techniques have been described [8]. CT venography was performed after a 3-minute delay. The deep leg veins were scanned from the inferior vena cava (IVC) confluence (iliac crest) through the popliteal veins (tibial plateau). The parameters for helical CT venography were a 7.5-mm collimation, 7.5-mm reconstruction, table speed of 30 mm/rotation, and pitch of 1.5. The tube current was 180 mA and the rotation time was 1 second. The peak kilovoltage was 120 kVp, but it was increased to 140 kVp for patients who weighed more than 250 lb (112.5 kg) [8].
In PIOPED II, image interpretations for all diagnostic tests except venous sonography were based on agreement of two PIOPED II-certified reviewers from centers other than the clinical center of origin [8]. An additional reviewer was used if the initial interpretations were discordant. Reviewers were unaware of any clinical information and of the results of other imaging tests except chest radiographs, which were included with ventilation-perfusion scans; the CTA and CT venography images were reviewed together. The criterion for the diagnosis of acute DVT on CT venography was a complete or partial central filling defect. For combined CTA and CT venography, VTE was diagnosed in patients with suspected PE if there was consensus that either CTA or CT venography was positive. Reviewers had to agree about the presence of PE in at least one lobe or about the presence of DVT in the same leg. VTE was excluded if there was consensus that both CTA and CT venography were negative.
The current study used a subset of 737 of 773 patients (95%) for whom complete CTA and CT venography data were available (105 with positive CT venography findings and 632 with negative CT venography findings). In PIOPED II, there were 13 central reviewers of CTA and CT venography. Of those reviewers, six were picked randomly. For each reviewer, all positive cases were included in this study. Control studies for each positive CT venography study were included: The next two (by date) negative CT venography cases for that reviewer were chosen. This resulted in a sample of 54 of 105 CT venography-positive cases and 108 of 632 CT venography-negative cases. The examinations were then reformatted from the original data set so that two of every three 7.5-mm images were deleted. Thus, the reformatted cases consisted of 7.5-mm images with 15-mm gaps.
Each reviewer received only his original CT venography-positive cases reformatted and twice the number of CT venography-negative cases reformatted. Initially, 162 cases were selected, but 12 were not useable because of various technical and archiving problems. The final study group consisted of 54 positive and 96 negative cases. Each reviewer was asked to interpret the studies at the same workstation that was used in PIOPED II and was provided with the same PIOPED II DVT score sheets for grading. The reviewers were informed that these images were PIOPED II images that had been reformatted in a discontinuous format. All identifying information, other than a code name, had been removed. Between 1 and 3.5 years had elapsed between the original interpretations and the interpretations of reformatted images.
PIOPED II used two or more reviewers to determine final CT venography interpretations. These consensus interpretations served as the standard with which the interpretations of the reformatted images were compared. Agreement between the consensus interpretations and the reformatted interpretations for each leg and per patient were then calculated. Kappa statistics were also calculated [22].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our study, the 89% agreement and kappa value of 0.75 per patient for the discontinuous images are similar to interobserver scores reported in the literature for CT venography. Richman et al. [16] reported an interobserver agreement ranging between 84% and 89% for different groups of reviewers, with kappa values varying from 0.65 to 0.84 (10-mm scans, 20-mm increments). Garg et al. [14], using 5-mm-thick axial images every 20 mm, found an 88% interobserver agreement with a kappa value of only 0.59 (95% CI, 0.39-0.78). This result compares with that of Coche et al. [23] who found a kappa statistic of 0.88 using continuous imaging.
Supporting the concept that discontinuous imaging is approximately equal to continuous imaging for the diagnosis of DVT is a review of 14 CT venography studies using discontinuous or continuous helical or mixed techniques by Ghaye and Dondelinger [13]. They stated that "Preliminary results did not show any difference, whether discontinuous or spiral techniques were compared with ultrasound" [13].
The images in our study were derived from helical data sets. Slice profiles derived from helical data have less sharp edges than those derived from axial data. Thus, our images were of slightly lower resolution than one would expect with true axial images of the same thickness. Whether this lower resolution was a diagnostic problem is not clear, but the results could only improve if true axial images were used.
Several shortcomings of the study should be mentioned. We did not restudy the entire PIOPED II population, but we chose a sample that should be representative. By chance, a disproportionate number of CT venography-negative studies had to be eliminated for technical reasons. PIOPED II reviewers compared CT venography with lower extremity sonography but had no independent reference standard for DVT. Thus, our study lacked an independent reference standard to determine if continuous imaging or discontinuous imaging was correct. For PIOPED II, agreement between CT venography and lower extremity sonography was > 95%.
Another shortcoming was our choice of reviewers. All of our reviewers are experienced CT venography reviewers and perhaps have a level of skill and confidence that yields better results than might be achieved with less experienced reviewers interpreting discontinuous images. Reviewer confidence was not surveyed, so it is possible that in some cases, confidence in diagnosing or excluding DVT would have been higher with continuous imaging.
Another potential limitation is the choice of section intervals. The PIOPED II CT venography studies were continuous images obtained with a 7.5-mm collimation and could be retro-spectively reformatted in only 7.5-mm intervals for this study (7.5 mm; skip = 15 mm). Many CT scanners cannot perform imaging using a 7.5-mm collimation to obtain images at 15-mm intervals. At our hospital, we use a 5-mm collimation for axial images and obtain images every 20 mm (5 mm; skip = 15 mm). This protocol or similar protocols should yield results similar to our study.
We calculated the differences in radiation dose from 5-mm axial discontinuous sections with a 15-mm skip versus continuous scanning from the iliac crest to the knees using our current CT venography protocol. The calculated pelvic radiation dose dropped from 9.1 to 4.5 mSv (Table 5). When automated dose modulation was used, radiation fell to 3.0 mSv. PIOPED II showed thrombi isolated to the IVC and iliac vessels occurred in only fewer than 3% of the cases and that CTA was positive for PE in each case. Others have also shown that thrombi isolated to only the IVC or iliac vessels are rare [12, 23]. Therefore, scanning from the acetabulum—rather than from the iliac crest—halves the region of pelvic coverage, further reducing the radiation dose. By our calculation, the use of discontinuous imaging, automated dose modulation (mA), and decreased anatomic area (acetabulum to knees) reduced the radiation dose from 9.1 to 2.0 mSv—a 78% reduction.
|
In conclusion, the per-patient agreement between continuous helical scanning and discontinuous scanning was 89%. This high level of agreement suggests that discontinuous helical imaging (7.5-mm collimation with images obtained at 15-mm intervals) provides results similar to continuous imaging. Using discontinuous imaging, reduced anatomic coverage, and automated mA adjustments, pelvic radiation dose can be reduced substantially.
Acknowledgments
Thanks to Sylvia Bartz for her secretarial help, Robert Smith for his help
with data tracking, and Mary Thielke for data reformatting. We thank the
following radiologists who served as reviewers for this study: Brannon
Hatfield, Emory University; John MacGregor, University of Calgary; David
Spizarny, Henry Ford Hospital; Conrad Wittram, Harvard University; Richard
Woodcock, Emory University; and David Yankelevitz, Cornell University. We also
thank the other PIOPED II CT reviewers: Claudia Henschke, Cornell Medical
Center; Laura Heyneman, Duke University; Theresa C. McLoud and JoAnn Shepard,
Massachusetts General Hospital; Paul Burrowes, University of Calgary; Ella
Kazerooni and Smita Patel, University of Michigan; and Jay Heiken and Pamela
Woodard, Washington University.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Nazaroglu, C. A. Ozmen, H. O. Akay, I. Kilinc, and A. Bilici 64-MDCT Pulmonary Angiography and CT Venography in the Diagnosis of Thromboembolic Disease Am. J. Roentgenol., March 1, 2009; 192(3): 654 - 661. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. R. Goodman, H. D. Sostman, P. D. Stein, and P. K. Woodard CT Venography: A Necessary Adjunct to CT Pulmonary Angiography or a Waste of Time, Money, and Radiation? Radiology, February 1, 2009; 250(2): 327 - 330. [Full Text] [PDF]
|
|
|
|
|
|
L. R. Goodman, P. D. Stein, F. Matta, H. D. Sostman, T. W. Wakefield, P. K. Woodard, R. Hull, D. F. Yankelevitz, and A. Beemath CT Venography and Compression Sonography Are Diagnostically Equivalent: Data from PIOPED II Am. J. Roentgenol., November 1, 2007; 189(5): 1071 - 1076. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
