Ultrasound for detection of deep venous thrombosis
ACP J Club. 1991 Nov-Dec;115:87. doi:10.7326/ACPJC-1991-115-3-087
Chance JF, Abbitt PL, Tegtmeyer CJ, Powers RD. Real-time ultrasound for the detection of deep venous thrombosis. Ann Emerg Med. 1991 May;20:494-6.
To assess the accuracy of real-time ultrasonography among patients presenting to the emergency department with clinically suspected deep-vein thrombosis.
Real-time, B-mode ultrasonography findings were compared with independent interpretations of contrast venography.
A university hospital emergency department with an annual volume of 40 000 patients, over a 1-year period.
70 adult patients presenting with clinically suspected deep venous thrombosis of the leg were referred by the emergency department attending physician for ultrasonography and venography. Patients who had a history of contrast reactions or who were pregnant were excluded.
Description of test and criterion standard
An ultrasound examination of the symptomatic leg was done by the radiology resident or ultrasonographer on call, and an initial report was made for the emergency department physician. The common femoral and popliteal veins were evaluated for compressibility; calf veins were not examined. Contrast venography (the criterion standard) was then obtained and interpreted by the resident or staff radiologist on call who was blinded to the ultrasound result. The ultrasound records and venograms were reviewed later by a faculty ultrasonographer and vascular radiologist, respectively, who (blinded to all results of the other test but with access to the initial reading of ultrasonography or venography, respectively) provided a final interpretation.
Main outcome measures
Diagnosis of proximal deep-vein thrombosis on initial and final readings.
The initial and final venogram interpretations agreed in all 70 patients. 20 patients (29%) had venographic evidence of deep-vein thrombosis; 14 patients (20%) had proximal-vein thrombosis; and 6 (9%) had calf-vein thrombosis only. All 14 patients with proximal thrombosis were identified on initial and final ultrasound readings (sensitivity, 100%; 95% CI, 79% to 100%). The final reading of the ultrasonograms identified all 56 patients without proximal deep vein thrombosis (specificity, 100%; CI, 94% to 100%), but the initial ultrasound readings (some by less experienced interpreters) incorrectly diagnosed 4 of these patients (specificity, 93%; CI, 84% to 97%).
The diagnosis of proximal deepvein thrombosis in symptomatic emergency department patients can reliably be excluded if a realtime ultrasound examination is negative.
Source of funding: Not stated.
Address for article reprint: Dr. J.F. Chance, Division of Emergency Medicine, Box 523-21, Charlottesville, VA 22908.
This paper adds to the growing number of studies that validate the use of ultrasonography as a diagnostic test in the management of patients with clinically suspected deep venous thrombosis. When the results of previous studies comparing ultrasonography with venography are pooled, real-time, Bmode ultrasound has been shown to have a sensitivity of 97% (range, 83% to 100%) and a specificity of 97% (range, 86% to 100%) (1).
The investigators of this study did not describe how the study subjects were selected from the larger pool of patients presenting to the emergency department with leg symptoms; thus we are unable to determine their similarity to patients that other clinicians see. However, the main conclusion that real-time ultrasonography is almost as accurate as venography for femoral and popliteal thrombosis can be accepted.
In this study, 11% of patients with negative ultrasound had calf deep venous thrombosis at venography. This observation, combined with the low sensitivity of ultrasound for calf deep venous thrombosis, supports the need for repeat testing to detect extending calf deep venous thrombosis in symptomatic patients with negative initial ultrasound results.
Finally, although real-time ultrasound has been shown to be accurate for proximal deep venous thrombosis, a clinical management trial still needs to be reported before this method can be recommended for widespread use (2). In such a study, patients with a positive ultrasound examination would have the result confirmed by contrast venography, whereas patients with negative results on serial testing would be followed for at least 3 months to see if an excess of venous thromboembolic events occurred.
William H. Geerts, MD
University of Toronto Toronto, Ontario