| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ultrasonography |
1 From the Section for Clinical Epidemiology, Division of General Medicine (J.C., S.D.P.) and the Department of Radiology (J.F.P.), Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115; the Department of Internal Medicine, University Hospital, Lausanne, Switzerland (J.C.); and the Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard Pilgrim Health Care, Boston, Mass (S.D.P.). Received April 7, 1998; revision requested June 29; revision received November 30; accepted January 4, 1999. J.C. supported by Research Fellowship grant no. 832B-036980 from the Swiss National Science Foundation. Address reprint requests to J.F.P.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: A retrospective cohort study of the prevalence of DVT included 977 consecutive patients with possible DVT but without known risk factors for DVT. Color flow and compression US were performed. The outcome was the frequency of overlooked, clinically important DVT after negative initial results from bilateral venous US of above- and below-the-knee veins. Patient records (904 patients), mailed questionnaires (61 patients), and telephone contacts (12 patients) were used to establish a diagnosis of clinically relevant DVT.
RESULTS: The prevalence of DVT was 15% (142 of 977), with DVT in 15% (21 of 142) restricted to the below-the-knee veins. Follow-up (median, 34 months) in 835 patients with negative US findings showed one occurrence of venous thrombosis (superficial thrombophlebitis) during the first 3 months of follow-up. Three other cases of venous thrombosis occurred at 17, 18, and 48 months of follow-up but were believed not to be linked to the initial complaint. The incidence of overlooked DVT was 0% (95% CI, 0.0%, 0.4%).
CONCLUSION: In patients without risk factors for DVT, a negative venous US study can help exclude the presence of clinically important DVT if the examination includes careful evaluation of the calf veins.
Index terms: Veins, extremities • Veins, thrombosis, 935.458, 935.751 • Veins, US, 935.12981, 935.12983
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Because venous US may not demonstrate below-the-knee DVT, alternate diagnostic strategies are adopted when a negative above-the-knee US study has been obtained in patients suspected of having DVT. One strategy is to perform venography to help detect calf vein DVT; another is to repeat the US examination to help exclude spread of nondetected calf vein thrombi (7). These strategies are, however, costly. The number of venographic studies can be large because the prevalence of isolated calf vein DVT usually is low, around 15% (6,8). Follow-up US may need to be repeated twice (9).
To our knowledge, an alternate strategy, performance of calf vein US in symptomatic patients suspected of having lower extremity DVT, has not been evaluated in any outcome study. We are not aware of any studies in which the outcome in patients with a negative complete venous US study—that is, evaluation above the knee as well as below the knee—has been examined. We report on the outcome in a large cohort of patients suspected of having DVT at the time of presentation and in whom negative above- and below-the-knee venous US studies were obtained. The purpose of this study was to determine the diagnostic value of below-the-knee venous US in patients suspected of having DVT but without known risk factors for DVT.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
10 days) immobilization, preexistent known cancer, pregnancy or postpartum state (within 2 weeks after delivery), use of oral contraceptives (within 1 month), or documented hypercoagulable state (presence of lupus anticoagulant or a protein S, protein C, or antithrombin III deficiency). Patients with risk factors for DVT were excluded because, for the purpose of our study, an episode of DVT detected at follow-up had to represent an episode that was overlooked at baseline US and was not a de novo episode related to one or more underlying risk factors for venous thrombosis. We did not exclude patients with previous DVT, because DVT in such patients could represent instances of "recurrent idiopathic DVT" (10). We excluded patients who had been examined for the presence of a malignancy, because the presence of a possible malignancy, which is a risk factor for DVT, was not known at the time of venous US. We also excluded patients in whom venous US was part of the work-up to evaluate for possible pulmonary embolism, because the clinical presentation was not that of possible symptomatic DVT but of possible pulmonary embolism. In patients in whom US was performed more than once, only the first US study was considered to be the baseline examination.
From an original list of 5,797 candidates, we excluded 3,905 patients because of the presence of risk factor(s) or because the indication for venous US was possible pulmonary embolism. An additional 830 were excluded because the US study in these patients was performed as a follow-up examination. Fourteen of the remaining 1,062 patients were excluded because of a concurrent investigation for possible malignancy or because they did not respond to a mailed questionnaire. After these exclusions, the cohort consisted of 1,048 members. Seventy-one patients were excluded because of a lack of follow-up information. However, among these 71 patients, 47 (66%) were still alive 3 months after venous US—that is, they had not died of any condition related to thromboembolic disease. In the other 24 patients (2.3% of the 1,048 patients), no reliable clinical information was available. The cohort in our study, therefore, included 977 patients who were identified as not having any risk factor for the development of DVT.
The study was approved by the human research committee of the hospital. The approval included permission to send a postcard to the patient to ask to conduct a phone interview.
Venous US consisted of a combination of color flow US and compression US (11). The diagnostic criterion for the presence of above-the-knee DVT was loss of compressibility in the femoral or popliteal veins. Compression US was conducted with a 5-MHz linear-array transducer. The full length of the femoral and popliteal veins was evaluated with compression applied to the venous segments every 1–2 cm.
All examinations were performed with the patient in the supine position; the popliteal vein was evaluated with the knee flexed. Color flow US and Doppler waveform analysis were used as a complement to compression US (11). Both extremities were examined unless there was a clinical contraindication (ie, a postoperative or traumatic wound). Evaluation of the calf veins was performed with color flow imaging to help identify the location of the deep veins of the calf (12). The knee was flexed, and the patient was in the supine position. The head of the stretcher was slightly raised to help distend the calf veins. The transducer was placed on the calf just posterior and medial to the tibia. The deep veins were imaged through a window between the tibia and the medial head of the gastrocnemius muscle. This imaging plane was used for most of the examination. The examination consisted of color flow transverse imaging and augmentation to verify patency of the posterior tibial and peroneal veins at the upper, middle, and lower portions of the calf. The color scale was set to a value of 11 cm/sec, and the color sensitivity was adjusted to just below the threshold where color signals were seen in the soft tissues of the calf. In addition, gray-scale compression US with direct visualization of the calf veins was performed. In cases where color flow signals were decreased or absent, additional effort was made to optimize gray-scale US, and a senior sonographer or sonologist was called to perform repeat scanning.
DVT was confirmed, in all cases, on the basis of the presence of a noncompressible venous segment on gray-scale US images. Superficial thrombophlebitis was diagnosed when thrombus involved only the superficial veins. Chronic venous disease was defined either as a partly compressible vein with only circumferential thickening of the wall or as a small noncompressible vein with contiguous collateral venous channels.
All sonographers were certified to perform venous US after a 6-month training course. Patients with symptoms localized to the calf underwent repeat US, as needed, performed by a senior sonologist (J.F.P.). Evaluation of calf veins typically took 15 minutes, as compared with the 30 minutes needed for bilateral lower extremity (above-the-knee) US.
A standardized form was used to collect the following data from patient charts: date of birth, sex, residence, date of procedure, history of DVT, and current or previous use of tobacco. Information about the occurrence of clinically important DVT in the 1,048 patients was obtained by the following means: (a) medical diagnosis after review of either in-hospital medical records in 384 (37%) patients or computerized records in 520 (50%) patients who were members of a health maintenance organization in the Boston (Mass) area; (b) self-reported diagnosis either on the mailed questionnaire in 61 (6%) patients or at the phone interview in 12 (1%). Limited information was obtained in 47 (4%) patients, and no information was obtained in 24 (2%). Information about survival status was obtained either from in-hospital medical records or from the death registry at the Massachusetts Bureau of Vital Statistics. We chose 3 months as an appropriate follow-up to establish a link between a potential diagnosis of venous thrombosis and the initial complaint (13).
Statistical analyses were performed with the 
2 and Wilcoxon rank sum statistics when appropriate. The 95% CIs were determined on the basis of the binomial distribution (exact CIs). The SAS statistical package (SAS Institute, Cary, NC) was used for all analyses.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The prevalence of acute DVT in the 977 patients was 15% (142 patients). The venous thrombi were principally distributed in above-the-knee popliteal or femoral veins in 121 (85%) of these 142 patients; the distribution was restricted to the below-the-knee veins in 21 (15%) patients. In comparison with the patients in whom DVT was not diagnosed (Table 1), patients with DVT were older (56 years ± 17 vs 52 years ± 17, P = .01), more likely to be male (55% vs 32%, P = .01), and current or former smokers (51% vs 41%, P = .06).
|
Of the 835 patients in whom a negative venous US study was obtained, two died in the first 3 months of follow-up (Table 2). Neither was suspected of having DVT or pulmonary embolism. In the 833 patients in whom a negative venous US study and at least 3 months of follow-up information were obtained, the median follow-up was 34 months (range, 3–60 months), for a total of 2,361 person years. Acute thrombophlebitis was diagnosed after initial negative US studies in four patients, but this diagnosis was established in only one of them during the first 3 months of follow-up. This patient had superficial thrombophlebitis and returned 25 days after the first examination with extension to the proximal portion of the greater saphenous vein. Thus, at 3-month follow-up, there were no cases of clinically important DVT in any of the 835 patients suspected of having DVT, for an incidence of 0% (95% CI, 0.0%, 0.4%).
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This study was not the first in which the outcome of a negative diagnostic test for the presence of acute DVT of the leg was examined. Follow-up in patients with a negative lower extremity venographic study has been shown (13,14) to yield a false-negative rate of 1%–2%. The false-negative rates for the results of noninvasive examinations for above-the-knee venous thrombosis also are available (3,15,16,17). Serial noninvasive venous examinations can help detect above-the-knee venous thrombosis at follow-up but at the cost of performing the examinations up to three times in all patients suspected of having DVT (3,9).
Our study differs in that the calf veins were evaluated. We observed a 15% prevalence of calf venous thrombosis. This rate was similar to that reported in studies with venography (8). Our rate of detection with the aid of US of isolated muscular venous thrombi is similar to that reported (6,18,19) for detection with venography. Studies (3,8,17) of the outcome after negative venous US results have not included an evaluation of the deep veins of the calf. The authors of these reports relied on a diagnostic strategy with follow-up examinations to help detect calf vein thrombi that may spread and become a risk for pulmonary embolism (1,9). This strategy is costly, however, because negative results are likely in most patients who present for evaluation of the deep veins.
The diagnostic sensitivity of calf vein imaging is a point of controversy, although values ranging from 88% to 95% have been reported (3–6). The results of our study cannot help clarify the issue of diagnostic accuracy as defined in comparison with venographic results. However, calf vein imaging in this cohort helped in the detection of all clinically relevant thrombi. Furthermore, the distribution of disease in our cohort was similar to that reported in studies with venography (6).
Our results can be compared with those from other studies and may cast light on the possible utility of evaluation of the calf veins. By using the results of follow-up after a diagnosis of DVT as the reference standard, we can assume a prevalence of at least 21 cases of calf vein DVT in our cohort (which was the observed number of cases) if the sensitivity was 100% and up to 42 cases if the sensitivity was 50%. On the basis of these numbers and an expected 20% rate of proximal propagation of the thrombus (1), four (20% of 21 cases) to eight (20% of 42 cases) potential cases of calf vein thrombi could have propagated above the knee and been detected at serial testing. Thus, if we had not performed imaging of calf veins but opted to perform serial US of the above-the-knee veins, we would have obtained five to 10 positive serial US studies in 142 patients, for a rate of 4%–7%. This estimate is within the prevalence rates of proximal propagation (per number of positive cases) estimated on the basis of serial US results by Vaccaro et al (17) (nine [2.1%] of 421) and Heijboer et al (3) (six [7%] of 84). Our estimate also is supported by the results of a recently published study (20) in which an incidence of proximal propagation was seen in seven (10%) of 70 patients examined at 5–7-day follow-up.
At 3-month follow-up, no clinically important case of DVT was diagnosed in any of the 835 patients suspected of having DVT, for an incidence of 0% (95% CI, 0.0%, 0.4%). This incidence should be compared with the following rates: (a) for serial impedance plethysmography, a rate of 2.5% (3); (b) for limited serial US of selected femoral and popliteal segments, a rate of 1.5% (3); and (c) for a mixture of serial (86 [8%] of 1,111) and single (1,025 [92%] of 1,111) above-the-knee lower extremity compression US, a rate of 0.5% (17) (Table 3). The latter study (17) was somewhat incomplete, however, in that only imaging records, hospital inpatient charts, and death certificates were evaluated in determining final patient outcome. The lower rates observed in our series are likely due to the selection of patients without risk factors, whereas the studies mentioned earlier in this paragraph include a heterogeneous group of patients. In addition, the authors of those studies did not stratify according to risk factors.
|
The likelihood of observing a DVT during follow-up is, therefore, dependent on the strength and number of the risk factors. For example, let us examine the typical postoperative scenario in a patient who has undergone hip replacement. This patient has a 20%–30% chance of the occurrence of a DVT. If US is 100% sensitive, DVT would be detected in 20–30 of 100 patients examined after hip surgery. In the remaining 70–80 patients, there would still be a 5%–10% chance of the occurrence of a new DVT after discharge from the hospital. If these patients were again evaluated with a "perfect" US examination, the outcome would be positive for DVT in 5%–10% of patients in whom a negative US study was originally obtained. This does not mean that the rate of overlooked DVT is 5%–10%. The outcomes are diluted: "positive DVT overlooked at first examination" plus "new DVT due to the presence of risk factors." In our study, we wished to examine only cases of "positive DVT overlooked at first examination."
Our strategy of evaluating the calf veins when performing lower extremity US may be superior to other diagnostic approaches such as the use of venography to help detect possible calf vein DVT or the performance of serial US. Whether our findings in a population of patients without risk factors for DVT can also be applied in patients with risk factors for DVT must be prospectively confirmed. Even if one assumes that the case of superficial thrombophlebitis was an overlooked case of DVT, the prevalence of overlooked DVT still was very low (0.1%; 95% CI, 0.0%, 0.7%).
This study was conducted at one institution and, because of this, the results may not be easily generalizable. The findings may be biased because of referral patterns. Furthermore, because the yield of venous US may be observer dependent, the quality of the sonographer's training may not be the same at other institutions.
In previous publications (12,19), successful demonstration of the calf veins at color Doppler flow US has been achieved in more than 96% of cases. We estimate that our rate of successful demonstration of large thrombi in calf veins is at least equal to this rate (19). The gray-scale component of the US examination is less often successful in the evaluation of the calf veins for smaller, nonobstructing thrombi. We (18) have previously achieved full gray-scale depiction of the calf veins in 40 of 56 (71%) consecutive patients. Failure to depict the calf veins tended to occur in the presence of other pathologic conditions or after surgery in six of eleven patients (18). Despite this limitation, US can help identify thrombi of the gastrocnemius or soleal vein that are not seen at venography (6,18,19).
Retrospective collection of data from medical records could be considered a weakness of the study design, because follow-up was not based on recurrent diagnostic testing. However, the use of follow-up information is more representative of routine clinical care. We excluded 71 patients in whom no reliable clinical information about the occurrence of symptomatic DVT was available. Nevertheless, 47 (66%) of these patients were still alive 3 months after the venous examination, that is, they had not died of any condition related to thromboembolic disease; thus, only 24 (2%) of 1,048 patients were lost to follow-up. However, we cannot exclude the possibility that some of those 71 patients actually had venous thrombosis.
We also excluded patients with well-defined risk factors for DVT, because, for the purpose of our study, an occurrence of DVT detected during follow-up must have represented a DVT that was overlooked at baseline US and not a de novo episode of DVT related to the presence of underlying risk factors for venous disorders.
We conclude that the results of a venous US examination that includes an evaluation of the calf veins in symptomatic patients without risk factors for DVT can help exclude the presence of important thromboembolic disease. Further study is needed to evaluate the effect of risk factors on our findings.
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, J.F.P.; study concepts, S.D.P., J.C., J.F.P.; study design, J.C., S.D.P.; definition of intellectual content, J.C.; literature research, J.C., J.F.P.; clinical studies, J.F.P.; data acquisition and analysis, J.C.; statistical analysis, J.C., J.F.P.; manuscript preparation, J.C., J.F.P.; manuscript editing and review, S.D.P., J.F.P.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. Li, J. Salanitri, S. Tutton, E. E. Dunkle, J. R. Schneider, J. A. Caprini, L. N. Pierchala, P. M. Jacobs, and R. R. Edelman Lower Extremity Deep Venous Thrombosis: Evaluation with Ferumoxytol-enhanced MR Imaging and Dual-Contrast Mechanism--Preliminary Experience Radiology, March 1, 2007; 242(3): 873 - 881. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Patel and E. A. Kazerooni Helical CT for the Evaluation of Acute Pulmonary Embolism Am. J. Roentgenol., July 1, 2005; 185(1): 135 - 149. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Lockhart, H. I. Sheldon, and M. L. Robbin Augmentation in Lower Extremity Sonography for the Detection of Deep Venous Thrombosis Am. J. Roentgenol., February 1, 2005; 184(2): 419 - 422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Schwarz, G. Siegert, W. Oettler, K. Halbritter, J. Beyer, R. Frommhold, S. Gehrisch, F. Lenz, E. Kuhlisch, H.-E. Schroeder, et al. Venous Thrombosis After Long-haul Flights Arch Intern Med, December 8, 2003; 163(22): 2759 - 2764. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. H. Gottlieb, S. L. Voci, L. Syed, C. Shyu, P. J. Fultz, D. J. Rubens, J. G. Strang, N. Carson, W. J. DiGrazio, and C. W. Francis Randomized Prospective Study Comparing Routine Versus Selective Use of Sonography of the Complete Calf in Patients with Suspected Deep Venous Thrombosis Am. J. Roentgenol., January 1, 2003; 180(1): 241 - 245. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Friera, N. R. Gimenez, P. Caballero, P. S. Molini, and C. Suarez Deep Vein Thrombosis: Can a Second Sonographic Examination Be Avoided? Am. J. Roentgenol., April 1, 2002; 178(4): 1001 - 1005. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Schwarz, B. Schmidt, B. Schmidt,, and S. M. Schellong Interobserver Agreement of Complete Compression Ultrasound for Clinically Suspected Deep Vein Thrombosis Clinical and Applied Thrombosis/Hemostasis, January 1, 2002; 8(1): 45 - 49. [Abstract] [PDF]
|
|
|
|
|
|
E. E. Coche, X. L. Hamoir, F. D. Hammer, P. Hainaut, and P. P. Goffette Using Dual-Detector Helical CT Angiography to Detect Deep Venous Thrombosis in Patients with Suspicion of Pulmonary Embolism: Diagnostic Value and Additional Findings Am. J. Roentgenol., April 1, 2001; 176(4): 1035 - 1039. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Garg, J. L. Kemp, D. Wojcik, S. Hoehn, R. J. Johnston, L. C. Macey, and A. E. Baron Thromboembolic Disease: Comparison of Combined CT Pulmonary Angiography and Venography with Bilateral Leg Sonography in 70 Patients Am. J. Roentgenol., October 1, 2000; 175(4): 997 - 1001. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
