Using several case studies, these authors discuss guidelines on perioperative anticoagulation and share insights from their protocol approach for foot and ankle surgery patients who are on long-term warfarin therapy.
For patients on long-term anticoagulation (AC) for past venous thromboembolism (VTE), atrial fibrillation (AFib) or mechanical heart valves, physicians have historically bridged these patients with low molecular weight heparin (LMWH) during the period of warfarin interruption for elective surgery.1
The practice of bridging was initiated out of a fear of thromboembolism when warfarin therapy was interrupted. This practice is based solely on a controversial phenomenon called rebound hypercoagulability. This phenomenon predicts that interruption of chronic anticoagulation therapy results in activation of the coagulation system.2
Studies over the past five years have shown that bridging can do more harm than good and this is changing the future of how we manage perioperative anticoagulation in patients on chronic therapy. The Bridge study was a randomized, double-blind, placebo-controlled trial that found a threefold higher risk of bleeding in those bridged during the perioperative period. In addition, the study found a similar risk of thromboembolism between patients who were bridged (0.3 percent) versus those patients who were not bridged (0.4 percent).3
New guidelines recommend bridging only for carefully selected patients while others will no longer require bridging during the perioperative period. This brings up a dilemma for surgeons regarding which patients can remain on warfarin throughout the perioperative period, who should have warfarin held without low molecular weight heparin bridging and who should have traditional bridging therapy.
We have developed an internal protocol to deal with this matter. In general, traditional bridging is based on the patient’s underlying medical condition while the “no hold” or “hold no bridge” approach is more based on the procedure one is performing. Our protocol also involves intraoperative and postoperative measures to minimize the risk of hematoma formation for patients on anticoagulation medications.
A 65-year-old woman was on chronic warfarin therapy for a past history of DVT, pulmonary embolism and anticardiolipin antibody syndrome. Preoperative X-rays (see Figure 1) revealed hallux valgus, second hammertoe contracture and pathologic flatfoot deformity. Surgery involved right subtalar joint fusion, multiple midfoot fusions with use of bone graft, hallux valgus repair and hammertoe correction (see Figure 2). The decision to hold warfarin was based on extensive surgery and concern for hematoma formation postoperatively. Based on new perioperative anticoagulation guidelines, no bridging was necessary. On the morning of surgery, the patient’s INR level was 1.2, which placed her outside of her therapeutic range. The patient resumed warfarin later on the day of surgery and her INR returned to therapeutic range by postoperative day 5.
Of note, this patient was probably outside of the therapeutic range for about eight days. Traditional bridging would have eliminated this concern however perioperative low molecular weight heparin increases the risk of intraoperative and postoperative bleeding, which can be a major impediment to wound healing. The low molecular weight heparin is also expensive and is not without inherent side effects. There were no issues related to perioperative venous thromboembolism or surgical site bleeding.
A 51-year-old woman was on chronic warfarin therapy for a history of Factor V Leiden. She presented with significant hallux valgus deformity (see Figure 3). She had a Lapidus fusion (see Figure 4) without interruption in her warfarin therapy.
The “no hold” approach requires perioperative monitoring of the INR level. One should check this a few days preoperatively to ensure that the patient is at the lower end of his or her therapeutic INR level, which can range from 2.0-3.0 or 2.5-3.5 depending on the medical condition requiring warfarin therapy. Physicians can hold warfarin for one or two days preoperative if the INR is above 3.0. Podiatric surgeons can perform most isolated foot and ankle surgery procedures safely with a patient’s INR level below 3.0, provided they take precautions regarding bleeding. For this patient, we checked the morning of surgery and found the patient had a 2.2 INR.
We took intraoperative measures to minimize bleeding. In addition to using a tourniquet, we emphasized more judicious use of electrocautery and incorporated suture ties. The tourniquet allows better visibility of vessels. Also, using a less aggressive Ace bandage for exsanguination instead of an Esmarch wrap is helpful to leave some blood in the extremity. We spot cauterized all visible vessels or used 3.0 absorbable suture ties. Then we deflated the tourniquet in order to assess bleeding prior to closure. We employed deep figure-of-8 sutures to close the deep periosteal incision over the dorsal aspect of the medial cuneiform where the transverse tarsal vessels are present in order to decrease bleeding of this tissue layer. Strict avoidance of vessel injury in the intermetatarsal space is also helpful both distally and proximally in Lapidus fusion.
Typically, our first clinic visit after forefoot surgery is 14 days postoperatively. However, for patients with diabetes and patients on anticoagulation therapy, we often see them at 1 week post-op to assess wound status. No hold patients should also assume that their wound may not heal as fast and they may struggle with early postoperative swelling for a longer period of time. Aggressive return to work plans are often counterproductive in patients on chronic warfarin therapy.
Of note, holding warfarin and bridging with Lovenox probably leads to more bleeding than simply leaving the patient on warfarin and carefully managing the INR level. Bridging therapy results in Lovenox and warfarin during the initial three to five days postop where bleeding is most likely to occur. This patient healed uneventfully and had no wound complications (see Figure 5).
Case Study 3: An Example Of A “Hold and Bridge” Approach For A Patient Having A Second Toe Amputation
A 65-year-old patient with diabetes was on chronic warfarin therapy for a heart valve replacement. We scheduled an elective amputation and allowed for warfarin to be held for five days preoperatively. The patient was bridged based on new guidelines related to heart valve replacement. Bridging with low molecular weight heparin raised concern for postoperative bleeding and we took intraoperative measures to minimize hematoma formation. There were no complications and the patient went on to heal uneventfully.
Primary care providers are aware of these new perioperative anticoagulation guidelines but they often leave anticoagulation up to the surgeon. It is therefore in our best interest to communicate our desires for no hold, hold no bridge or hold and bridge based on our expectations of postoperative bleeding. Primary care providers are in charge of the hold and bridge patients based on the medical condition, but we can express our comfort level with leaving them on warfarin through the perioperative period.
There are many patients with heart valve replacement, a history of VTE, active cancer, clotting disorders, etc. who can have minor foot and ankle procedures as long as we keep their INR within the therapeutic range. We commonly communicate by sending a message through the EMR to the primary care physician in regard to our preference for the patient’s anticoagulation. Podiatrists should flag patients on anticoagulation therapy for additional intraoperative and postoperative efforts to minimize bleeding.
Table 1 at right divides patients into different categories according to the no hold, hold no bridge, and hold and bridge care plans. No hold is generally appropriate for isolated procedures and forefoot procedures, procedures that allow surgeons to control intraoperative bleeding in a relatively easy manner. Most forefoot surgeries have minimal dead space and postoperative bleeding is expressed into the bandage, which can be changed a few days later. This approach also works well for patients with compromised circulation such as a patient with a transmetatarsal amputation for gangrene and minor arthroscopy procedures with minimal bleeding. The hold no bridge approach works well for ankle arthroscopy with increased periarticular debridement, amputation in patients with normal blood flow, reconstructive procedures involving major bone work, total ankle replacement and Charcot reconstruction.
Dr. Boffeli is a board-certified foot and ankle surgeon practicing at HealthPartners Specialty Center in St. Paul, Minn. He is a Fellow of the American College of Foot and Ankle Surgeons, and the Director of the Foot and Ankle Surgical Program at Regions Hospital/HealthPartners Institute for Education and Research.
Dr. Schaefbauer is a third year Chief Resident at HealthPartners Specialty Center in St. Paul, Minn.
- Lange CM, Fichtlscherer S, Miesbach W, et al. The periprocedural management of anticoagulation and platelet aggregation inhibitors in endoscopic interventions. Dtsch Arztebl Int. 2016; 113(8):129–35.
- Rose AJ, Allen AL, Minichello T. A call to reduce the use of bridging anticoagulation. Circ Cardio Qual Outcomes. 2016; 9(1):64-67.
- Douketis JD, Spyropoulos AC, Kaatz S, et al. Perioperative bridging anticoagulation in patients with atrial fibrillation. N Engl J Med. 2015; 373(9):823-33.
- Clark NP, Witt DM, Davies LE, et al. Bleeding, recurrent venous thromboembolism, and mortality risks during warfarin interruption for invasive procedures. JAMA Intern Med. 2015; 175(7):1163-1168.