Upper-Extremity DVT


Synopsis: The history and physical examination are poor predictors of the presence of acute venous thrombosis.

Source: Prandoni P, et al. Arch Intern Med 1997;157:57-62.

A prospective survey of patients presenting with suspected upper-extremity deep venous thrombosis (UEDVT) was undertaken by Prandoni et al to determine the diagnostic accuracy of ultrasonographic techniques (compression ultrasonography [CUS], color-flow Doppler imaging [CFDI], and Doppler ultrasonography [DUS]) compared with venography; the importance of risk factors; and complications. Fifty-eight patients were evaluated; 27 ultimately had UEDVT confirmed by venography. Significant risk factors included central venous catheters, thrombophilic states, and past lower-extremity venous thrombosis. Sensitivity and specificity of the various ultrasonographic studies were noted: CFDI (100% and 93%); CUS (96% and 94%); and DUS (81% and 77%). Thirty-six percent of patients experienced pulmonary embolism (PE) while, after a mean of two years’ surveillance, two patients suffered recurrent thromboembolic events and four patients had post-thrombotic sequelae. Among the 31 patients with an entirely negative evaluation, alternate diagnoses were identified in eight, including superficial venous thrombosis (4), lymphedema (2), neoplastic compression (1), and muscle injury (1).


Patients presenting with upper-extremity edema, pain, or both may have an UEDVT; however, it was ultimately diagnosed in fewer than half of the patients referred with symptoms and signs in this study. This finding reinforces the point that the history and physical examination are poor predictors of the presence of acute venous thrombosis. Ultimately, therapeutic decisions should be made after performance of a reliable diagnostic study.

Many of the "classically" reported risk factors for UEDVT—such as anatomic abnormality, strenuous exercise, and IV drug abuse—were not noted to occur individually to a significant extent. Several of these factors, however, did occur in conjunction with thrombophilic states, suggesting perhaps that these "less important" risk factors may assume a greater significance when combined with an underlying predisposition. Of interest, 26% of patients were noted to have a thrombophilic state, suggesting that patients presenting with UEDVT should have blood samples obtained to uncover such underlying conditions prior to initiation of anticoagulant therapy.

The authors noted that both CUS and CFDI were of equal acceptable value in their ability to diagnose UEDVT, while DUS was of considerably less assistance. They recommend the use of CUS because this technique is both less expensive and less time-consuming compared to CFDI. These three techniques should not be considered as separate, exclusive investigations; rather, they complement one another and are often used together in most applications of upper-extremity ultrasonography. The physician must also consider the local expertise of the radiologic consultant in the choice of study; venography may be preferable if the examining radiologist does not have sufficient experience with non-invasive imaging of the upper extremity.

Finally, PE was encountered in 36% of patients with UEDVT. This relatively high rate is significantly more frequent than the expected range of 10-25%. PE does occur in the setting of UEDVT; UEDVT represents a syndrome with the potential for significant morbidity and mortality. (Dr. Brady is Assistant Professor of Emergency Medicine and Internal Medicine, Medical Director, Chest Pain Center, University of Virginia, Charlottesville.)