Despite imperfect sensitivity, ultrasound thrombosis detection following arthroplasty is useful

Abstract

We sought to determine whether ultrasound may be used to screen for deep venous thrombosis 7 days following arthroplasty (despite its known imperfect sensitivity), to select a population at high risk of pulmonary embolism from day 7 to 90 after surgery. Complete bilateral triplex ultrasound imaging was scheduled for 656 patients one week after primary total hip or knee replacement. Prophylactic low molecular weight heparin administration was replaced by anticoagulant treatment if a thrombosis was detected. Venous thrombosis, pulmonary embolism and haemorrhage were recorded at 3 months, as were deaths. Thrombo-embolic events were observed in 83 of 656 patients at one week (mainly asymptomatic). Age > 60 years, knee (versus hip) arthroplasty or a history of varicose veins were associated with a significantly increased risk of thrombosis at day 7 (p < 0.0001). Three of these 83 patients (3.6%) suffered from pulmonary embolism (1 fatal) between days 7 and 90, despite efficient anticoagulant treatment. Treatment of venous thrombosis in these 83 patients was associated with one re-admission for haemorrhagic complications (1.2%). In the remaining 573 patients between day 7 and 90, two patients suffered from pulmonary embolism, 12 others developed a venous thrombosis and 6 were re-admitted for haemorrhage (1 fatal). Neither age ≥ 60 years, varicoses veins, nor knee (vs. hip) arthroplasty significantly increased the risk of delayed pulmonary embolism. Ultrasound is useful for the detection of asymptomatic deep venous thrombosis following arthroplasty. A positive echo-doppler (and no other clinical index) increases by 10 times (p < 0.01) the risk of delayed pulmonary embolism despite anticoagulant therapy as compared to subjects with no evidence of thrombosis receiving prophylaxis.