Abstract
Introduction
Peak incidence of pulmonary embolism (PE) typically occurs weeks after total hip (THA) or knee (TKA) arthroplasty, long after hospital discharge. We investigated risk factors for acute PE occurring during index hospitalisation.
Methods
Retrospective review of an IRB-approved database identified 329 arthroplasties performed by a single surgeon between 2002 and 2007 at two University teaching hospitals. Warfarin (goal INR 2.0) was standard venous thromboembolism prophylaxis.
Results
There were 126 (38.6%) primary THA, 86 (26.1%) primary TKA and 117 (35.3%) revision arthroplasties. Seven patients (7/329; 2.1%) experienced clinically evident non-fatal pulmonary embolism, including 5 after TKA (5/128; 3.9%) and 2 after THA (2/194; 1.0%). In-hospital PE occurred in 4 (1.2%) patients (3 TKA, 1 THA) at a mean 2.7 days (range 2-4 days) after operation, compared with 3 symptomatic events (2 TKA, 1 THA) occurring after discharge (mean 19.3 days; range 8-27 days). Three of four patients suffering acute in-hospital PE were on pre-operative warfarin for chronic atrial fibrillation. Among all patients on pre-operative warfarin, 3 (12.5%) sustained an early PE, with a relative risk of 38.1 times that of controls not on chronic warfarin therapy (p=0.001). Body mass index greater than 30 kg/m2, a history of previous venous thromboembolism, coronary artery disease, ASA score, and type of operation were all found to not be associated with increased risk of in-hospital PE.
Conclusion
Acute symptomatic in-hospital pulmonary embolism was correlated with chronic pre-operative warfarin anticoagulation. Compared to historical controls, time to PE was shorter in patients on chronic warfarin. Rebound hypercoagulability after discontinuation of chronic warfarin in preparation for total joint arthroplasty represents a greater hazard than excessive bleeding; we advocate bridging anticoagulation with LMWH in these patients.