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