Short versus extended thromboprophylaxis after total knee arthroplasty: a randomized comparison

The optimal duration of thromboprophylaxis after total knee arthroplasty remains uncertain.

We performed a randomized, open trial to determine whether to stop thromboprophylactic therapy at Day 10±2 ('short thromboprophylaxis') was non-inferior to continue thromboprophylactic therapy up to Day 35±5 ('extended thromboprophylaxis') after total knee arthroplasty. At Day 7±2, subjects were screened by ultrasonography for asymptomatic deep-vein thrombosis and randomized. The primary outcome was a composite of proximal deep-vein thrombosis, any symptomatic deep-vein thrombosis, non-fatal symptomatic pulmonary embolism, major bleeding, heparin-induced thrombocytopenia, or all-cause death up to Day 35±5. The secondary outcome was ultrasonographic (extension or new onset) distal deep-vein thrombosis at Day 35±5.

Twenty-one patients (2.4%) were not randomized, because of asymptomatic proximal deep-vein thrombosis on systematic ultrasonography at Day 7±2. Among the 857 randomized patients, mean (SD) duration of anticoagulant treatment was 11.2 (6.7) and 33.9 (3.7) days in the short and extended thromboprophylaxis groups, respectively. The respective rates of the primary outcome were 4.0% (17/420) and 2.4% (10/422), with an absolute difference of 1.7% (90% confidence interval, -0.3 to 3.7). In 285 patients with asymptomatic distal deep-vein thrombosis at Day 7±2, the respective rates of the primary outcome were 7.8% and 2.8% (p=0.067). The rates of the secondary outcome were 14.8% (62/420) and 4.5% (19/422), respectively (p<0.001).

Short thromboprophylaxis was not non-inferior to extended thromboprophylaxis after total knee arthroplasty. In this setting, the thromboembolic risk persisted longer than seven days, notably in patients with asymptomatic distal deep-vein thrombosis at discharge. ClinicalTrials.gov number: NCT00362492.