Algorithm could be safe for suspected upper-extremity deep venous thrombosis

Combining a clinical decision score, D-dimer testing, and ultrasonography allows for safe and effective exclusion of upper-extremity deep venous thrombosis (UEDVT), according to a study published on April 1 by the Annals of Internal Medicine.

“Confirming the presence of UEDVT is important in view of the risk for pulmonary embolism, prevention of the postthrombotic syndrome, and avoidance of unnecessary exposure to anticoagulants,” wrote lead author Ankie Kleinjan, MD, PhD, of the Academic Medical Center in Amsterdam, and colleagues.

Currently, a diagnostic algorithm that combines clinical probability assessment, D-dimer testing and ultrasonography is commonly used for suspected deep venous thrombosis of the leg. The authors suggest the need for simple and effective tools in confirming or refuting diagnosis, and posit the idea of a combined diagnostic algorithm for UEDVT.

Using 406 inpatients and outpatients with suspected UEDVT, the researchers prospectively evaluated the safety and feasibility of an algorithm that used the Constans clinical decision score, D-dimer testing and compression ultrasonography.

Of these patients, 203 had a clinical decision score implying that UEDVT was unlikely and thus underwent D-dimer testing. In 21 percent of the patients, the UEDVT diagnosis was excluded without ultrasonography. None of these patients developed symptomatic venous thromboembolism.

Superficial venous thrombosis (SVT) and UEDVT were diagnosed in 54 and 103 patients, respectively. All 249 patients with normal diagnostic work-up were followed for three months. One patient developed UEDVT during the follow-up period, resulting in an overall failure rate of 0.4 percent.

“This study shows that a noninvasive diagnostic algorithm combining the Constans clinical score, D-dimer testing, and ultrasonography is safe in excluding UEDVT and SVT and feasible in 96% of patients,” wrote Kleinjan and colleagues. “This ap- proach is attractive because it is simple, quick, noninvasive, and similar to the well-established algorithm for diagnosis of suspected DVT of the leg, which could facilitate its implementation in clinical practice.”