Role of venous duplex scanning in patients with suspected pulmonary embolism

We examined the use of venous duplex scanning (VDS) in the diagnosis of pulmonary embolism (PE) at our institution.

Patients undergoing lower extremity VDS from October 1988 through June 1995 were cross-referenced with those who underwent ventilation perfusion (V/Q) scans and pulmonary angiography (PA) for PE.

A total of 664 of 3534 VDS were for "rule out PE." Deep venous thrombosis was found in 13%. A total of 256 VDS were in conjunction with V/Q scans in 249 patients, with only 8% undergoing PA. Deep venous thrombosis was present in 18% for those with both V/Q and VDS compared with 10% (p < 0.01) for those with VDS as the sole study. The order in which V/Q, VDS, and PA were obtained and the relationship of positive studies was examined.

We found no pattern to the sequence of tests ordered. V/Q scan rather than VDS should be the first study in the evaluation of PE. PE was diagnosed or excluded in nearly one third of patients based on V/Q as the initial study. A total of 29% of VDS could have been avoided. Treatment could be determined on the basis of VDS as the initial study in only 13%. We found only 14% incidence of positive PA in patients with nondiagnostic V/Q scans. We advocate judicious use of diagnostic tests in a stepwise fashion to appropriately treat patients with suspected PE.