D-Dimer in Venous Thromboembolism
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Wells and colleagues (Sept. 25 issue)1 report on the use of D-dimer testing in patients with suspected deep-vein thrombosis. Patients with a low pretest probability of deep-vein thrombosis and a negative result on the D-dimer test did not undergo confirmatory testing. This is problematic. Current recommendations for the evaluation of new diagnostic tests suggest an independent, blinded comparison with a reference standard and application of that standard independently of the test being evaluated.2 The gold standard for deep-vein thrombosis is either duplex ultrasonography or venography, neither of which was performed in this study group. The "gold standard" for these patients was the absence of a clinical event at three months. A false negative result would be identified if a patient had a symptomatic recurrence. It is unclear whether there would be a significant number of such recurrences in just three months. These limitations make it likely that the study overestimates the sensitivity and clinical utility of D-dimer testing. In addition, other studies have suggested that the SimpliRED assay has only 77 percent sensitivity for detecting pulmonary embolism.3 Confirmation of the utility of D-dimer testing awaits trials that uniformly apply diagnostic standards.
Scott D. Stern, M.D.
University of Chicago
Chicago, IL 60637
References
Wells PS, Anderson DR, Rodger M, et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003;349:1227-1235.
Jaeschke R, Guyatt GH, Sackett DL. Users' guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients. JAMA 1994;271:703-707.
Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001;135:98-107.
To the Editor: It is encouraging to see three articles in the September 25 issue — by Wells et al.,1 Fedullo and Tapson,2 and Bockenstedt3 — that highlight the need to assign a pretest probability of disease, which is consistent with Bayes' theorem, before test results are interpreted.1,2,3 However, the liberal use of D-dimer testing in a busy emergency department creates two potential problems.4 The test has a low specificity, so if it is used as a screening tool in patients with a very low pretest probability of disease (e.g., 1 to 5 percent), a positive test result only marginally increases the post-test probability, while at the same time the use of the test generates many false positive results and thus incurs the numerous incremental costs associated with tests required to rule out the disease definitively. Moreover, if the pretest probability of disease is high, a negative result on the D-dimer test should not lower the post-test probability of disease below the threshold at which definitive testing is required, given the dire scenario of missing the diagnosis. The D-dimer test should be used only in patients who have more than a very low pretest probability, and less than a moderate pretest probability, of disease. For patients with a very low pretest probability, the most appropriate decision may be to omit the screening test because of the high number of false positive results. For those who have a moderate or higher pretest probability, the decision should be to proceed directly to a definitive test rather than a screening test.
Douglas A. Propp, M.D.
Alan M. Kumar, M.D.
Advocate Lutheran General Hospital
Park Ridge, IL 60068
douglas.propp-md@advocatehealth.com
References
Wells PS, Anderson DR, Rodger M, et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003;349:1227-1235.
Fedullo PF, Tapson VF. The evaluation of suspected pulmonary embolism. N Engl J Med 2003;349:1247-1256.
Bockenstedt P. D-Dimer in venous thromboembolism. N Engl J Med 2003;349:1203-1204.
Propp D. Caveat emptor: the positive D-dimer slippery slope. Ann Emerg Med 2003;42:309-310.
To the Editor: In their evaluation of D-dimer testing in the diagnosis of suspected lower-extremity deep-vein thrombosis, Wells and coworkers propose incorporating such testing into the strategy for diagnosing deep-vein thrombosis. They argue that doing so would allow deep-vein thrombosis to be ruled out in some patients (those with a negative result on the test) and, consequently, that ultrasound imaging would not be required in such patients.
Although we agree that this algorithm fits well with the diagnosis of deep-vein thrombosis, we disagree with its routine usefulness, for a very simple reason not discussed in the article. When a patient has symptoms in his or her legs that make the diagnosis of deep-vein thrombosis a possibility but does not in fact have deep-vein thrombosis, the physician must identify the cause of the disability that resembles deep-vein thrombosis. Calf hematoma, partial muscle rupture, and popliteal cyst are the most frequent conditions that mimic deep-vein thrombosis. In all patients with these conditions, ultrasound imaging is as necessary as it is in patients with deep-vein thrombosis. That is why, in many patients with suspected deep-vein thrombosis, a negative result on the D-dimer test does not eliminate the need for ultrasonography.
Michel Vayssairat, M.D.
H?pital Tenon
75020 Paris, France
michel.vayssairat@tnn.ap-hop-paris.fr
To the Editor: In their articles on the evaluation of suspected pulmonary embolism or deep-vein thrombosis, Wells et al., Fedullo and Tapson, and Bockenstedt do not mention pulmonary hypertension as a cause of a positive result on the D-dimer assay. Pulmonary hypertension is accompanied by active microvascular thrombosis, which explains the sustained elevation of D-dimer levels in the absence of hypercoagulability and deep-vein thrombosis.1,2 Patients with pulmonary hypertension present with many of the confounding clinical symptoms of acute pulmonary embolism, such as dyspnea, systemic hypotension, tachycardia, cardiac-rhythm disturbances, and anxiety. In addition to ultrasound imaging (to rule out deep-vein thrombosis) and ventilation–perfusion lung scanning or spiral CT scanning (to rule out pulmonary embolism), patients with a positive result on the D-dimer assay should be evaluated with electrocardiography or echocardiography so that pulmonary hypertension can be ruled out as a cause of the D-dimer elevation.
Lanfranco de Clari, M.D.
Libera Università di Scienze Umane e Technologiche
CH-6900 Paradiso, Switzerland
declari@bluewin.ch
References
Shitrit D, Rudensky B, Zimran A, Elstein D. D-dimer assay in Gaucher disease: correlation with severity of bone and lung involvement. Am J Hematol 2003;73:236-239.
Shitrit D, Bendayan D, Bar-Gil-Shitrit A, et al. Significance of a plasma D-dimer test in patients with primary pulmonary hypertension. Chest 2002;5:1674-1678.
The authors reply: We thank Drs. Stern, Propp and Kumar, Vayssairat, and de Clari for their comments on our study. Dr. Stern is concerned about a gold-standard test for D-dimer. The D-dimer test is not a new diagnostic test. It has been evaluated in a number of studies, and its negative predictive value and potential utility have been demonstrated. It is for this reason that we performed a management study in which patient care was based on the algorithms under study and in which success was defined by the absence of thromboembolic end points in patients in whom the diagnosis of deep-vein thrombosis was initially considered to be ruled out. In our study, we showed that a strategy according to which deep-vein thrombosis is ruled out with a negative D-dimer test result and a clinical assessment suggesting that deep-vein thrombosis is unlikely is not problematic.
We agree with Drs. Propp and Kumar that in patients with a high pretest clinical probability of deep-vein thrombosis, ultrasonography needs to be performed. We have previously shown that at least 15 percent of patients in this probability subgroup have false negative D-dimer results.1 In our current study, all the patients categorized as having a high pretest probability underwent ultrasound imaging. D-dimer testing was used only to determine the need for follow-up ultrasound assessments in those with initially normal ultrasound images. We share the correspondents' concern about the risk of overuse of the D-dimer test. It must be emphasized that it is not a screening test and should not be part of a screening algorithm for patients at risk who present with leg or pulmonary symptoms. If it were used for screening, we would run the risk of actually increasing the number of diagnostic tests performed to rule out deep-vein thrombosis, whereas decreasing the number of tests was one of the goals of our research.
Dr. Vayssairat is correct in stating that ultrasound imaging is capable, in some cases, of identifying an alternative cause of symptoms in patients who present with lower-extremity swelling or discomfort. However, the treatment is almost always the same for the majority of these problems — namely, antiinflammatory medication, rest, and local measures, such as application of heat or ice. We have managed quite nicely for several years now without always knowing exactly what the cause of the symptoms are, and we have avoided the cost of performing ultrasound studies in thousands of patients. If a patient's symptoms do not resolve with basic measures, then further evaluation may be indicated.
We thank Dr. de Clari for reminding readers that D-dimer tests may be positive in patients with chronic pulmonary hypertension. Indeed, D-dimer tests may be positive in many other conditions. The point is that a negative D-dimer test in conjunction with an unlikely clinical probability rules out the diagnosis of deep-vein thrombosis. We are not arguing against diagnostic imaging if chronic pulmonary hypertension is suspected.
Philip S. Wells, M.D.
Ottawa Hospital
Ottawa, ON K1Y 4E9, Canada
pwells@ohri.ca
David Anderson, M.D.
Queen Elizabeth II Health Science Centre
Halifax, NS B3H 2Y9, Canada
References
Wells PS, Anderson DR, Bormanis J, Guy F, Mitchell M, Lewandowski B. SimpliRED D-dimer can reduce the diagnostic tests in suspected deep vein thrombosis. Lancet 1998;351:1405-1406.
Scott D. Stern, M.D.
University of Chicago
Chicago, IL 60637
References
Wells PS, Anderson DR, Rodger M, et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003;349:1227-1235.
Jaeschke R, Guyatt GH, Sackett DL. Users' guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients. JAMA 1994;271:703-707.
Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001;135:98-107.
To the Editor: It is encouraging to see three articles in the September 25 issue — by Wells et al.,1 Fedullo and Tapson,2 and Bockenstedt3 — that highlight the need to assign a pretest probability of disease, which is consistent with Bayes' theorem, before test results are interpreted.1,2,3 However, the liberal use of D-dimer testing in a busy emergency department creates two potential problems.4 The test has a low specificity, so if it is used as a screening tool in patients with a very low pretest probability of disease (e.g., 1 to 5 percent), a positive test result only marginally increases the post-test probability, while at the same time the use of the test generates many false positive results and thus incurs the numerous incremental costs associated with tests required to rule out the disease definitively. Moreover, if the pretest probability of disease is high, a negative result on the D-dimer test should not lower the post-test probability of disease below the threshold at which definitive testing is required, given the dire scenario of missing the diagnosis. The D-dimer test should be used only in patients who have more than a very low pretest probability, and less than a moderate pretest probability, of disease. For patients with a very low pretest probability, the most appropriate decision may be to omit the screening test because of the high number of false positive results. For those who have a moderate or higher pretest probability, the decision should be to proceed directly to a definitive test rather than a screening test.
Douglas A. Propp, M.D.
Alan M. Kumar, M.D.
Advocate Lutheran General Hospital
Park Ridge, IL 60068
douglas.propp-md@advocatehealth.com
References
Wells PS, Anderson DR, Rodger M, et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003;349:1227-1235.
Fedullo PF, Tapson VF. The evaluation of suspected pulmonary embolism. N Engl J Med 2003;349:1247-1256.
Bockenstedt P. D-Dimer in venous thromboembolism. N Engl J Med 2003;349:1203-1204.
Propp D. Caveat emptor: the positive D-dimer slippery slope. Ann Emerg Med 2003;42:309-310.
To the Editor: In their evaluation of D-dimer testing in the diagnosis of suspected lower-extremity deep-vein thrombosis, Wells and coworkers propose incorporating such testing into the strategy for diagnosing deep-vein thrombosis. They argue that doing so would allow deep-vein thrombosis to be ruled out in some patients (those with a negative result on the test) and, consequently, that ultrasound imaging would not be required in such patients.
Although we agree that this algorithm fits well with the diagnosis of deep-vein thrombosis, we disagree with its routine usefulness, for a very simple reason not discussed in the article. When a patient has symptoms in his or her legs that make the diagnosis of deep-vein thrombosis a possibility but does not in fact have deep-vein thrombosis, the physician must identify the cause of the disability that resembles deep-vein thrombosis. Calf hematoma, partial muscle rupture, and popliteal cyst are the most frequent conditions that mimic deep-vein thrombosis. In all patients with these conditions, ultrasound imaging is as necessary as it is in patients with deep-vein thrombosis. That is why, in many patients with suspected deep-vein thrombosis, a negative result on the D-dimer test does not eliminate the need for ultrasonography.
Michel Vayssairat, M.D.
H?pital Tenon
75020 Paris, France
michel.vayssairat@tnn.ap-hop-paris.fr
To the Editor: In their articles on the evaluation of suspected pulmonary embolism or deep-vein thrombosis, Wells et al., Fedullo and Tapson, and Bockenstedt do not mention pulmonary hypertension as a cause of a positive result on the D-dimer assay. Pulmonary hypertension is accompanied by active microvascular thrombosis, which explains the sustained elevation of D-dimer levels in the absence of hypercoagulability and deep-vein thrombosis.1,2 Patients with pulmonary hypertension present with many of the confounding clinical symptoms of acute pulmonary embolism, such as dyspnea, systemic hypotension, tachycardia, cardiac-rhythm disturbances, and anxiety. In addition to ultrasound imaging (to rule out deep-vein thrombosis) and ventilation–perfusion lung scanning or spiral CT scanning (to rule out pulmonary embolism), patients with a positive result on the D-dimer assay should be evaluated with electrocardiography or echocardiography so that pulmonary hypertension can be ruled out as a cause of the D-dimer elevation.
Lanfranco de Clari, M.D.
Libera Università di Scienze Umane e Technologiche
CH-6900 Paradiso, Switzerland
declari@bluewin.ch
References
Shitrit D, Rudensky B, Zimran A, Elstein D. D-dimer assay in Gaucher disease: correlation with severity of bone and lung involvement. Am J Hematol 2003;73:236-239.
Shitrit D, Bendayan D, Bar-Gil-Shitrit A, et al. Significance of a plasma D-dimer test in patients with primary pulmonary hypertension. Chest 2002;5:1674-1678.
The authors reply: We thank Drs. Stern, Propp and Kumar, Vayssairat, and de Clari for their comments on our study. Dr. Stern is concerned about a gold-standard test for D-dimer. The D-dimer test is not a new diagnostic test. It has been evaluated in a number of studies, and its negative predictive value and potential utility have been demonstrated. It is for this reason that we performed a management study in which patient care was based on the algorithms under study and in which success was defined by the absence of thromboembolic end points in patients in whom the diagnosis of deep-vein thrombosis was initially considered to be ruled out. In our study, we showed that a strategy according to which deep-vein thrombosis is ruled out with a negative D-dimer test result and a clinical assessment suggesting that deep-vein thrombosis is unlikely is not problematic.
We agree with Drs. Propp and Kumar that in patients with a high pretest clinical probability of deep-vein thrombosis, ultrasonography needs to be performed. We have previously shown that at least 15 percent of patients in this probability subgroup have false negative D-dimer results.1 In our current study, all the patients categorized as having a high pretest probability underwent ultrasound imaging. D-dimer testing was used only to determine the need for follow-up ultrasound assessments in those with initially normal ultrasound images. We share the correspondents' concern about the risk of overuse of the D-dimer test. It must be emphasized that it is not a screening test and should not be part of a screening algorithm for patients at risk who present with leg or pulmonary symptoms. If it were used for screening, we would run the risk of actually increasing the number of diagnostic tests performed to rule out deep-vein thrombosis, whereas decreasing the number of tests was one of the goals of our research.
Dr. Vayssairat is correct in stating that ultrasound imaging is capable, in some cases, of identifying an alternative cause of symptoms in patients who present with lower-extremity swelling or discomfort. However, the treatment is almost always the same for the majority of these problems — namely, antiinflammatory medication, rest, and local measures, such as application of heat or ice. We have managed quite nicely for several years now without always knowing exactly what the cause of the symptoms are, and we have avoided the cost of performing ultrasound studies in thousands of patients. If a patient's symptoms do not resolve with basic measures, then further evaluation may be indicated.
We thank Dr. de Clari for reminding readers that D-dimer tests may be positive in patients with chronic pulmonary hypertension. Indeed, D-dimer tests may be positive in many other conditions. The point is that a negative D-dimer test in conjunction with an unlikely clinical probability rules out the diagnosis of deep-vein thrombosis. We are not arguing against diagnostic imaging if chronic pulmonary hypertension is suspected.
Philip S. Wells, M.D.
Ottawa Hospital
Ottawa, ON K1Y 4E9, Canada
pwells@ohri.ca
David Anderson, M.D.
Queen Elizabeth II Health Science Centre
Halifax, NS B3H 2Y9, Canada
References
Wells PS, Anderson DR, Bormanis J, Guy F, Mitchell M, Lewandowski B. SimpliRED D-dimer can reduce the diagnostic tests in suspected deep vein thrombosis. Lancet 1998;351:1405-1406.