Multidetector-Row Computed Tomography in Suspected Pulmonary Embolism
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Complex diagnostic algorithms are in vogue for the diagnosis of venous thromboembolism. In yet another management study, Perrier et al. (April 28 issue)1 propose a combination of multislice computed tomography (CT) and D-dimer measurement as a strategy to diagnose pulmonary embolism, as discussed in an editorial by Goldhaber.2
I am uncomfortable with the suggestion that a strategy that results in a 1-in-50 risk of pulmonary embolism at follow-up is hailed as a breakthrough.
Alexander G.G. Turpie, M.D.
McMaster University
Hamilton, ON L8L 2X2, Canada
turpiea@mcmaster.ca
References
Perrier A, Roy P-M, Sanchez O, et al. Multidetector-row computed tomography in suspected pulmonary embolism. N Engl J Med 2005;352:1760-1768.
Goldhaber SZ. Multislice computed tomography for pulmonary embolism -- a technological marvel. N Engl J Med 2005;352:1812-1814.
To the Editor: Perrier et al. show that among patients who had multidetector CT scans showing no pulmonary embolism, only 0.9 percent were found to have proximal deep venous thrombosis on ultrasonography. On the basis of these results, the authors suggest that future studies be done to determine whether ultrasonography can be safely eliminated from the diagnostic workup of patients with a negative CT scan. Pending a true cost-effectiveness analysis, however, I wonder how many physicians would be comfortable missing even 0.9 percent of cases of deep-vein thrombosis, considering the life-threatening potential of untreated venous thromboembolism. I also wonder whether the authors would have been comfortable forgoing therapy in the three patients who had a negative CT and then received a diagnosis of deep venous thrombosis. If their answer is yes, then perhaps the study they suggest is warranted.
Mark D. Siegel, M.D.
Yale University School of Medicine
New Haven, CT 06520-8057
mark.siegel@yale.edu
The authors reply: We thank Dr. Turpie and Dr. Siegel for their interest in our work. Both express concern about the rate of false negative results associated with our diagnostic strategy. The overall failure rate of our algorithm was 1.5 percent. This rate included the thromboembolic events that would have been missed if ultrasonography had not been included in the initial workup and the thromboembolic events during the three-month follow-up period among patients in whom pulmonary embolism had been ruled out and, therefore, had not been treated. This 1.5 percent risk of thromboembolic events is similar to the risk among patients who were left untreated on the basis of a negative pulmonary angiogram and might reflect recurrence of an initially undetected thromboembolic event or the occurrence of a new, unrelated event in patients with continuing risk factors for venous thromboembolism.
Dr. Turpie judges our strategy to be complex. In fact, we believe our results allow a simplification of existing strategies, since they suggest that a negative result on venous-compression ultrasonography of the lower limbs is no longer required to rule out pulmonary embolism in a patient with a negative multidetector CT scan. However, Dr. Siegel is uncomfortable with that simplification because of the 0.9 percent risk of missing an acute venous thromboembolic event if lower-limb ultrasonography is not performed. Again, that risk was included in the 1.5 percent overall failure rate of a diagnostic strategy that did not include ultrasonography. We agree that a formal cost-effectiveness analysis should be performed, but it is difficult to imagine that a strategy requiring more than 100 ultrasonographic studies to pick up one additional deep venous thrombosis would be cost-effective. Moreover, ultrasonography also has its limitations, and at least some of the deep-vein thromboses detected by systematic examination of patients who had thoracic symptoms but no lower-limb symptoms might be false positives, caused by absence of recanalization of a previous clot.
In summary, no single test or strategy allows the detection of all pulmonary emboli. The discomfort expressed by the commentators should probably be attributed to the difficulties inherent in the diagnosis of pulmonary embolism rather than to our diagnostic scheme.
Arnaud Perrier, M.D.
Geneva University Hospital
CH-1211 Geneva 14, Switzerland
arnaud.perrier@medecine.unige.ch
Pierre-Marie Roy, M.D.
Centre Hospitalier Universitaire d'Angers
49933 Angers, France
Guy Meyer, M.D.
H?pital Européen Georges-Pompidou
75908 Paris, France
The editorialist replies: Dr. Turpie overlooks some key points. The diagnostic algorithm presented by Perrier et al. is simple, not complex. Clinical likelihood assessment (which can be done by "gestalt") and D-dimer enzyme-linked immunosorbent assays can quickly rule out pulmonary embolism in many patients who otherwise would needlessly undergo imaging. This approach is reliable and cost-effective.
The 1-in-50 risk of pulmonary embolism after three months of follow-up with chest CT scanning as the principal imaging test matches the follow-up results obtained with the invasive alternative: classic pulmonary angiography, which increases discomfort, risk, and cost.1
With respect to multislice chest CT,2 this approach has led to four changes in diagnostic approach: venous ultrasonography of the legs is no longer necessary when multislice chest CT scanning rules out pulmonary embolism; the size and accessibility during surgery or catheterization of the pulmonary embolism can be immediately ascertained; detection of right ventricular enlargement identifies high-risk patients with ominous prognoses3; and if pulmonary embolism is ruled out, chest CT may detect alternative diagnoses that explain the presenting symptoms and signs.
Samuel Z. Goldhaber, M.D.
Brigham and Women's Hospital
Boston, MA 02115
References
Quiroz R, Kucher N, Zou KH, et al. Clinical validity of a negative computed tomography scan in patients with suspected pulmonary embolism: a systematic review. JAMA 2005;293:2012-2017.
Schoepf UJ, Goldhaber SZ, Costello P. Spiral computed tomography for acute pulmonary embolism. Circulation 2004;109:2160-2167.
Schoepf UJ, Kucher N, Kipfmueller F, Quiroz R, Costello P, Goldhaber SZ. Right ventricular enlargement on chest computed tomography: a predictor of early death in acute pulmonary embolism. Circulation 2004;110:3276-3280.
I am uncomfortable with the suggestion that a strategy that results in a 1-in-50 risk of pulmonary embolism at follow-up is hailed as a breakthrough.
Alexander G.G. Turpie, M.D.
McMaster University
Hamilton, ON L8L 2X2, Canada
turpiea@mcmaster.ca
References
Perrier A, Roy P-M, Sanchez O, et al. Multidetector-row computed tomography in suspected pulmonary embolism. N Engl J Med 2005;352:1760-1768.
Goldhaber SZ. Multislice computed tomography for pulmonary embolism -- a technological marvel. N Engl J Med 2005;352:1812-1814.
To the Editor: Perrier et al. show that among patients who had multidetector CT scans showing no pulmonary embolism, only 0.9 percent were found to have proximal deep venous thrombosis on ultrasonography. On the basis of these results, the authors suggest that future studies be done to determine whether ultrasonography can be safely eliminated from the diagnostic workup of patients with a negative CT scan. Pending a true cost-effectiveness analysis, however, I wonder how many physicians would be comfortable missing even 0.9 percent of cases of deep-vein thrombosis, considering the life-threatening potential of untreated venous thromboembolism. I also wonder whether the authors would have been comfortable forgoing therapy in the three patients who had a negative CT and then received a diagnosis of deep venous thrombosis. If their answer is yes, then perhaps the study they suggest is warranted.
Mark D. Siegel, M.D.
Yale University School of Medicine
New Haven, CT 06520-8057
mark.siegel@yale.edu
The authors reply: We thank Dr. Turpie and Dr. Siegel for their interest in our work. Both express concern about the rate of false negative results associated with our diagnostic strategy. The overall failure rate of our algorithm was 1.5 percent. This rate included the thromboembolic events that would have been missed if ultrasonography had not been included in the initial workup and the thromboembolic events during the three-month follow-up period among patients in whom pulmonary embolism had been ruled out and, therefore, had not been treated. This 1.5 percent risk of thromboembolic events is similar to the risk among patients who were left untreated on the basis of a negative pulmonary angiogram and might reflect recurrence of an initially undetected thromboembolic event or the occurrence of a new, unrelated event in patients with continuing risk factors for venous thromboembolism.
Dr. Turpie judges our strategy to be complex. In fact, we believe our results allow a simplification of existing strategies, since they suggest that a negative result on venous-compression ultrasonography of the lower limbs is no longer required to rule out pulmonary embolism in a patient with a negative multidetector CT scan. However, Dr. Siegel is uncomfortable with that simplification because of the 0.9 percent risk of missing an acute venous thromboembolic event if lower-limb ultrasonography is not performed. Again, that risk was included in the 1.5 percent overall failure rate of a diagnostic strategy that did not include ultrasonography. We agree that a formal cost-effectiveness analysis should be performed, but it is difficult to imagine that a strategy requiring more than 100 ultrasonographic studies to pick up one additional deep venous thrombosis would be cost-effective. Moreover, ultrasonography also has its limitations, and at least some of the deep-vein thromboses detected by systematic examination of patients who had thoracic symptoms but no lower-limb symptoms might be false positives, caused by absence of recanalization of a previous clot.
In summary, no single test or strategy allows the detection of all pulmonary emboli. The discomfort expressed by the commentators should probably be attributed to the difficulties inherent in the diagnosis of pulmonary embolism rather than to our diagnostic scheme.
Arnaud Perrier, M.D.
Geneva University Hospital
CH-1211 Geneva 14, Switzerland
arnaud.perrier@medecine.unige.ch
Pierre-Marie Roy, M.D.
Centre Hospitalier Universitaire d'Angers
49933 Angers, France
Guy Meyer, M.D.
H?pital Européen Georges-Pompidou
75908 Paris, France
The editorialist replies: Dr. Turpie overlooks some key points. The diagnostic algorithm presented by Perrier et al. is simple, not complex. Clinical likelihood assessment (which can be done by "gestalt") and D-dimer enzyme-linked immunosorbent assays can quickly rule out pulmonary embolism in many patients who otherwise would needlessly undergo imaging. This approach is reliable and cost-effective.
The 1-in-50 risk of pulmonary embolism after three months of follow-up with chest CT scanning as the principal imaging test matches the follow-up results obtained with the invasive alternative: classic pulmonary angiography, which increases discomfort, risk, and cost.1
With respect to multislice chest CT,2 this approach has led to four changes in diagnostic approach: venous ultrasonography of the legs is no longer necessary when multislice chest CT scanning rules out pulmonary embolism; the size and accessibility during surgery or catheterization of the pulmonary embolism can be immediately ascertained; detection of right ventricular enlargement identifies high-risk patients with ominous prognoses3; and if pulmonary embolism is ruled out, chest CT may detect alternative diagnoses that explain the presenting symptoms and signs.
Samuel Z. Goldhaber, M.D.
Brigham and Women's Hospital
Boston, MA 02115
References
Quiroz R, Kucher N, Zou KH, et al. Clinical validity of a negative computed tomography scan in patients with suspected pulmonary embolism: a systematic review. JAMA 2005;293:2012-2017.
Schoepf UJ, Goldhaber SZ, Costello P. Spiral computed tomography for acute pulmonary embolism. Circulation 2004;109:2160-2167.
Schoepf UJ, Kucher N, Kipfmueller F, Quiroz R, Costello P, Goldhaber SZ. Right ventricular enlargement on chest computed tomography: a predictor of early death in acute pulmonary embolism. Circulation 2004;110:3276-3280.