1. Home>
2. Specialties>
3. General Medicine>
4. Summary and Comment

D-dimer in the Diagnosis of DVT -- A Randomized Trial

D-dimer testing and clinical risk assessment could reduce the number of ultrasounds that are performed for suspected DVT.

Although several diagnostic strategies have been proposed for deep-vein thrombosis (DVT), they have not been compared in randomized trials. Now, Canadian investigators report results of a multicenter trial that involved 1096 outpatients with suspected DVT.

First, clinicians determined whether each patient's pretest probability of DVT was high or low, according to a previously validated prediction model. Then, patients were randomized to either D-dimer testing or no D-dimer testing (controls). In the D-dimer group, patients with low clinical probabilities and negative D-dimer tests underwent no further work-up, and the remaining patients underwent ultrasonography. In the control group, all patients underwent ultrasonography. In both groups, high-risk patients with negative ultrasound results underwent repeat ultrasonography at 1 week.

During the initial evaluation, DVT was diagnosed in about 15% of patients in each group. During 3 months of follow-up, 2 additional venous thromboembolic events occurred in the D-dimer group, and 6 occurred in the control group -- a nonsignificant difference. Ultrasound was not performed in 39% of D-dimer patients.

Comment: In this trial, a diagnostic strategy that involved D-dimer testing and selective ultrasonography performed as well as a strategy of mandatory ultrasonography for all patients. The success of the D-dimer strategy elsewhere will depend on at least 2 things -- the ability of clinicians to apply correctly the pretest probability model, and the accuracy of a given D-dimer assay (in this study, the SimpliRED assay and IL-Test were used).

— Allan S. Brett, MD

Published in Journal Watch General Medicine October 14, 2003

Citation(s):

Wells PS et al. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003 Sep 25; 349:1227-35.

• Original article (Subscription may be required)
• Medline abstract (Free)