| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ultrasonography |
1 From the Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Ave, Boston, MA 02215. Received October 2, 1998; revision requested November 19; revision received December 17; accepted January 27, 1999. Address reprint requests to R.G.S. (e-mail: rsheiman@caregroup.harvard.edu).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: One hundred eighty-two consecutive patients referred for bilateral lower extremity US as the first examination for suspected PE were evaluated prospectively for predisposing factors and symptoms of DVT. Patients were placed into four groups: group 1, no symptoms or risk factors; group 2, both symptoms and risk factors; group 3, only risk factors; group 4, only symptoms. The prevalence of DVT detected at lower extremity US in each group was determined.
RESULTS: There were 89 patients in group 1, 12 in group 2, 43 in group 3, and 38 in group 4, with a DVT prevalence of 0%, 25%, 14%, and 24%, respectively. There was no significant difference in DVT prevalence between groups with symptoms or risk factors but a significant difference between these groups and the group lacking both symptoms and risk factors.
CONCLUSION: Lower extremity US as the initial examination in patients suspected of having PE should be used only in those patients who have symptoms or risk factors for DVT. This would substantially decrease the number of examinations performed without a decline in DVT detection.
Index terms: Embolism, pulmonary, 60.72 • Extremities, thrombosis, 93.751 • Thrombosis, US, 93.1298, 93.12983, 93.751 • Ultrasound (US), Doppler studies, 93.12983 • Veins, thrombosis, 93.751
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
At our institution (Beth Israel Deaconess Medical Center) and others (6,7), US assessment for lower extremity DVT is becoming the initial study requested for clinically suspected PE. This is likely owing to the well-documented association of lower extremity DVT and PE, the high sensitivity and specificity of US for DVT, and the availability and noninvasive nature of this test. However, when used as a screening study for all patients suspected of having a PE, lower extremity US no longer is being applied in the preselected population with an indeterminate ventilation-perfusion scan. Initial assessment for lower extremity DVT in all patients suspected of having PE, therefore, includes patients who otherwise would not have undergone DVT examination because of a ventilation-perfusion scan with normal or high-probability results.
Spiral computed tomographic (CT) angiography of the pulmonary arteries has become another noninvasive tool for PE assessment and demonstrates sensitivities above 90% and a specificity of 78% or greater for PE up to the segmental level (8,9). In a recent theoretic evaluation of multiple algorithms for the work-up for PE (10), which included pulmonary CT angiography and a D-dimer test, lower extremity US, followed by pulmonary CT angiography, was identified as the most cost-effective initial approach for the evaluation for PE. Other investigators (11) have advocated lower extremity US as the initial screening examination for PE when concomitant lower extremity symptoms of DVT exist. On the basis of these recent publications and our experience, it is expected that the use of US for DVT detection as the first-line examination for PE will increase.
To our knowledge, the application of lower extremity US for DVT in patients who are suspected of having PE and who have not undergone any preselection has not been evaluated prospectively. Therefore, we sought to determine the prevalence of DVT in this patient population and whether DVT symptoms or risk factors may be applied as selection criteria to improve the pretest probability of lower extremity US when used as the initial examination for PE assessment.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
At the time US was performed, each patient was examined by a resident or staff physician for symptoms of acute DVT, which included lower extremity pain, swelling, erythema, or any combination of these, and each patient's clinical history was reviewed for risk factors for DVT. Leg swelling was determined subjectively and on the basis of direct comparison of both legs when swelling was unilateral or was considered bilateral if thought to be present by the patient or the referring physician. Risk factors were elicited directly from the patient and from chart review and included underlying malignancy, hypercoagulable state, prior DVT, pregnancy, prolonged travel or immobility, and recent surgery. When present, symptoms were documented as unilateral, including which extremity, or bilateral. For each patient, data were recorded on a work sheet completed before leaving the US suite and were verified later at the time of chart review.
All examinations were performed on a model 128XP/10 (Acuson Computed Sonography, Mountain View, Calif) or Ultramark 9 or HDI 3000 (Advanced Technology Laboratories, Bothell, Wash), all with color Doppler and duplex capabilities and with the appropriate 5-, 7-, or 10-MHz linear array transducer. Examinations included gray-scale, color Doppler, and spectral evaluation of the common femoral, saphenous, superficial femoral, popliteal, and proximal calf veins in the longitudinal and transverse planes. Vessels were assessed for compressibility and intraluminal filling defects on gray-scale scans, presence or absence of flow voids on color Doppler scans, and spectral evaluation of blood flow characteristics, which included augmentation with calf compression. Patients were examined by staff sonographers and then reexamined by a staff radiologist to confirm all findings before patient discharge from the vascular laboratory.
Positive diagnosis of thrombus was based on lack of complete lumen obliteration with compression, flow void on color Doppler scans, and lack of flow detection at spectral analysis. The last two criteria alone were applied to the adductor canal because assessment for compressibility cannot be performed in this area. Intraluminal filling defects on gray-scale scans and luminal expansion, if present, helped to confirm the finding of thrombus. Chronic thrombus, as defined by vessel noncompressibility, wall thickening, and luminal contraction with or without flow at spectral analysis and color Doppler, was not considered a positive study result for the sake of statistical analysis.
Patients were placed into one of four groups according to the presence or absence of DVT symptoms and risk factors as follows: group 1, no symptoms or risk factors; group 2, both symptoms and risk factors; group 3, only DVT risk factors; group 4, only symptoms. Groups were then compared for the prevalence of DVT by using the Fleiss method of multiple pairwise comparison of proportions. A P value less than .05 was considered to indicate a statistically significant difference. This allowed us to determine the potential utility of using DVT symptoms and risk factors as selection criteria for using lower extremity US for screening in patients suspected of having a PE. In addition, the prevalence of US-detectable DVT in the entire patient population in our study was calculated. Finally, we noted if patients underwent any additional evaluation for PE; however, we made no attempt to determine the final disposition or prevalence of PE in the population of patients in our study, as this was not the focus of our study.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
Of the nine patients with DVT who presented with only symptoms, all presented with unilateral symptoms and had DVT only in the symptomatic leg. No patient with bilateral symptoms was found to have DVT.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Multiple retrospective studies support our results, although none focused formally on the use of both DVT symptoms and risk factors as selection criteria for the application of lower extremity US to screen for PE. Eze et al (6) found lower extremity DVT in only 10% (31 of 306) of patients referred for US for the work-up for PE, and this declined to 7% (22 of 306) when patients with isolated calf clot were excluded. However, when patients were stratified according to the presence or absence of unilateral symptoms of DVT, the prevalence of DVT was 40% (10 of 25) and 5% (15 of 281), respectively. Unlike in our study, these investigators did not define DVT symptoms or discuss how they assessed for DVT symptoms. This study also differs from ours in that no attempt was made to assess for risk factors for DVT, which may explain why DVT was found in the 5% of patients without lower extremity symptoms. Of interest was the lack of further evaluation for PE in 66% of their patients, similar to our findings (70%).
Bendick et al (12) retrospectively evaluated 507 patients who underwent lower extremity US for DVT assessment who were referred for suspected PE. Seventy-six percent (384) of these patients had no symptoms or risk factors for DVT, and clot above the knee was identified in only one patient (0.3% prevalence). Conversely, 44 (36%) of 123 patients with symptoms or risk factors were found to have DVT. This study, however, included all patients who underwent lower extremity US as part of their work-up for PE and did not focus solely on patients in whom lower extremity US was the initial examination for suspected PE.
Finally, Lipski et al (13) reviewed the records of 450 patients suspected of having PE who were referred for lower extremity testing for DVT. No stratification concerning lower extremity symptoms or risk factors for DVT was performed because these investigators were concerned with the effect of lower extremity US on the management of suspected PE rather than the development of selection criteria for the use of lower extremity US in this clinical scenario. In 171 of the patients in their study, US was the first-line examination, and DVT was identified in only 13 for a prevalence of 8%.
Although we know of no other study that prospectively evaluated the frequency of DVT in patients clinically suspected of having a PE, a theoretic calculation of this can be estimated. In their study modeling the cost-effectiveness of spiral CT angiography for suspected PE, van Erkel et al (10) combined data from multiple published reports to derive a pretest probability of 24% for PE in patients clinically suspected of having PE and a probability of 65% for detectable DVT in patients with documented PE. On the basis of these values, theoretically only 15.6% of all patients suspected of having PE should have demonstrable DVT. This is slightly higher than our findings and is likely owing to their use of the upper limit quoted in the literature for the probability of DVT when PE is documented. Still, this implies, as is exemplified by our data, that the majority of examinations for DVT performed for suspected PE are negative. We believe that the number of studies with normal results can be substantially reduced by using DVT symptoms and risk factors as selection criteria.
As mentioned, 70% of the patients in our study had no additional work-up for PE, and within this subgroup 96% had normal examination results. This would indicate that the suspicion for PE in the patient population in our study was low. It is likely that the combination of low clinical suspicion for PE and normal bilateral US results allayed the fears of PE. However, this type of logic, which was put forth in a recent publication (14), seems neither safe nor cost-effective. We and others (11,13,15) point out the shortcoming of terminating the pursuit of PE after normal lower extremity US results owing to the frequent absence of DVT in cases of documented PE. As stated by Cronan (16), ". . .the absence of a positive US study should not be used to exclude the diagnosis of PE."
Also, we point out that quantifying the clinical suspicion for PE, once the question of PE has been raised, is difficult. This is exemplified by the Prospective Investigation of Pulmonary Embolism Diagnosis, or PIOPED, study (17) in which clinical suspicion for PE was graded as noncommittal in the majority (64% [569 of 887]) of cases. This is why we made no attempt to formally quantify the level of clinical suspicion for PE in this study and do not believe that the lack of doing so led to any selection bias. We believe, however, that on the basis of our findings, if such an approach is to persist, the selection criteria put forth here for the use of lower extremity US should be applied.
Within the group of patients with only lower extremity symptoms, all cases of DVT were unilateral, with thrombus identified only in the symptomatic extremity. No DVT was found in any asymptomatic extremity or in any patient with bilateral symptoms. This is in accord with our previous findings (18,19). These data support performance of a larger prospective study focused on the utility of performing a bilateral examination in the patient who has unilateral DVT symptoms and who is undergoing lower extremity US for suspected PE.
In conclusion, if lower extremity US is contemplated as the initial examination in patients suspected of having PE, it should be applied only in those patients who also have DVT symptoms or risk factors. In this way, DVT detection can be optimized and the maximum number of patients eliminated from further work-up for PE. In patients who are suspected of having PE and who have no symptoms or risk factors, our data indicate that screening US is of no clinical utility, and an alternative noninvasive examination such as pulmonary CT arteriography should be contemplated. In addition, our results indicate that the diagnosis of PE is often not pursued once an initial lower extremity US examination shows normal results. Because DVT is frequently absent with documented PE, lower extremity US should not be ordered as a first-line study for PE if a normal result will terminate additional work-up.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, R.G.S.; study concepts and design, R.G.S.; definition of intellectual content, R.G.S., C.R.M.; literature research, R.G.S.; clinical studies, R.G.S., C.R.M.; data acquisition, R.G.S.; data analysis, R.G.S., C.R.M.; manuscript preparation, R.G.S.; manuscript editing and review, R.G.S., C.R.M.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. K. Pahade, D. Litmanovich, I. Pedrosa, J. Romero, A. A. Bankier, and P. M. Boiselle Quality Initiatives: Imaging Pregnant Patients with Suspected Pulmonary Embolism: What the Radiologist Needs to Know RadioGraphics, May 1, 2009; 29(3): 639 - 654. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Bernstein, N. Haim, B. Brenner, G. Sarig, G. Bar-Sela, and D. Gaitini Venous Sonography for the Diagnosis of Asymptomatic Deep Vein Thrombosis in Patients With Cancer Undergoing Chemotherapy J. Ultrasound Med., May 1, 2004; 23(5): 655 - 658. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Mayo, L. H. Ketai, P. A. Loud, Z. D. Grossman, D. S. Katz, and A. M. Gittleman Invited Commentary * Authors' Response RadioGraphics, October 1, 2002; 22(90001): S20 - 24. [Full Text] [PDF]
|
|
|
|
 |
|
 Lower-Extremity Ultrasound as an Initial Test for PE? Journal Watch (General), July 23, 1999; 1999(723): 1 - 1. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
