To compare blood loss, delay of surgery, and short-term adverse events in (1) patients admitted on warfarin versus nonanticoagulated controls and (2) warfarin patients with day of surgery (DOS) international normalized ratio (INR) of 1.5 or greater versus below 1.5.Design:
Academic Level I trauma center.Patients/Participants:
One hundred twenty four patients treated surgically for hip fractures including patients presenting on warfarin (n = 62) and matched controls (n = 62).Intervention:
Cephalomedullary nailing (CMN), hemiarthroplasty, or total hip arthroplasty.Main Outcome Measures:
The primary outcome was transfusion rate. Secondary outcomes included calculated blood loss, 30-day complication rate, and hours from emergency department presentation to surgery.Results:
There was no significant difference in blood transfusion rates between the warfarin and control groups (P = 0.86). Blood transfusion was required in 58.1% of patients in the warfarin group (48.3% of arthroplasties and 65.5% of CMNs) compared with 56.6% of controls (41.9% of arthroplasties and 73.3% of CMNs). There were also no significant differences in calculated blood loss or in complication rates. Patients on warfarin had significantly longer time to surgery (P < 0.01). Subanalysis of the warfarin group showed that patients with DOS INR at or above 1.5 had similar transfusion rates, blood loss, and complications compared with patients with INR below 1.5. Treatment with CMN was the only covariate that was found to be a significant independent predictor of transfusion on multivariable analysis (P = 0.048).Conclusions:
Patients with hip fractures admitted on warfarin seem to be at similar risk of transfusion or adverse events compared with nonanticoagulated patients. Awaiting normalization of INR delayed surgery without reducing bleeding or preventing complications. Within reason, surgeons may consider proceeding with surgery in patients with INR above 1.5 if patients are otherwise medically optimized. The upper limit above which surgery causes increased blood loss is currently unknown.Level of Evidence:
Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.