Perioperative asymptomatic venous thrombosis: role of duplex scanning versus venography

We compared combined B-mode/Doppler (duplex ultrasonic scanning and venography in routine preoperative and postoperative screening for major proximal deep vein thrombosis in 78 patients undergoing total hip or knee arthroplasty. Of 309 extremity examinations, duplex scanning had an overall sensitivity of 85.7% (12/14) and a specificity of 97.3% (287/295). The preoperative prevalence and postoperative incidence of major deep vein thrombosis were 2.5% and 14.1% of patients, respectively, despite intensive mechanical and pharmacologic prophylaxis. In addition, venography documented a preoperative prevalence and postoperative incidence of isolated calf deep vein thrombosis in 2.5% and 16.7% of patients, respectively. Whereas such disease extended proximally even in the absence of anticoagulation in only 18% of patients studied by serial duplex scans, calf deep vein thrombosis accounted for the only two instances of pulmonary embolism in this study. There were no deaths related to pulmonary embolism. This study suggests that duplex scanning is useful in screening for perioperative deep vein thrombosis in patients undergoing total hip or knee arthroplasty, which carries a significant risk of venous thromboembolism despite routine prophylaxis.