More than 3 million people in the United States have atrial fibrillation, most of whom are being managed with anticoagulation therapy for life. The goal of the present study was to examine the effect of chronic anticoagulation therapy on patients with atrial fibrillation who undergo total joint arthroplasty.
We retrospectively reviewed all patients undergoing aseptic primary or revision total joint arthroplasty at our facility from March 2007 to August 2011. One hundred and sixty-one patients with atrial fibrillation (Group A) were compared with 161 matched controls (Group B). A total of 112 hips and 210 knees underwent 239 primary arthroplasties and eighty-three revisions. The groups were compared with use of conditional logistic regression (with matching on the basis of the involved joint [hip or knee], type of procedure [revision or primary], age, and sex) with regard to the length of hospital stay, postoperative hemoglobin levels, transfusion requirements, and readmissions.
The preoperative length of stay (1.7 versus 0.2 days; p < 0.0001), postoperative length of stay (4.6 versus 3.2 days; p = 0.0002), and total length of stay (6.3 versus 3.4 days; p < 0.0001) were significantly longer for patients with atrial fibrillation (Group A). Hemoglobin levels were lower (but not significantly so) for Group A at baseline (13.1 versus 13.8 mg/dL), on Postoperative Day 2 (10.1 versus 10.6 mg/dL), on Postoperative Day 3 (9.8 versus 10.2 mg/dL), on Postoperative Day 4 (9.6 versus 10.1 mg/dL), on Postoperative Day 5 (9.7 versus 9.9 mg/dL), and at discharge (9.9 versus 10.3 mg/dL). Group A had a significantly higher prevalence of blood transfusion (15.5% versus 3.7%; p = 0.0005) and periprosthetic joint infection (5.6% versus 0.62%; p = 0.0196). A diagnosis of atrial fibrillation (odds ratio, 4.09; 95% confidence interval, 2.05 to 8.18; p < 0.0001) significantly increased the odds of total joint arthroplasty complication and the need for hospital readmission.
Patients with preoperative atrial fibrillation undergoing total joint arthroplasty had an increased length of hospital stay, increased transfusion requirements, and an increased risk of periprosthetic joint infection and unplanned hospital readmission.
Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.
1The Rothman Institute of Orthopedics at Thomas Jefferson University Hospital, 925 Chestnut Street, 2nd Floor, Philadelphia, PA 19107. E-mail address for V.K. Aggarwal: firstname.lastname@example.org. E-mail address for A. Ong: Alvin.Ong@rothmaninstitute.com