Perioperative anticoagulation in device implantation: Benefitting the uninterrupted

Abstract

2012; 108: 1124–1131. One of the most common dilemmas in surgery is managing bleeding risk in patients requiring oral anticoagulation (OAC) because of risk of thrombosis or embolism, as in atrial fibrillation (AF) (1, 2). The potential additional risk of bleeding in patients who continue OAC perioperatively is balanced by the increased risk of thromboembolism if anticoagulation is discontinued (2). This risk may be as high as 2.7% in the week post pacemaker implantation in patients with AF if OAC is discontinued (3). Traditional practice has been to discontinue OAC prior to surgery and institute bridging therapy with heparin perioperatively; this is discontinued preoperatively, reinstituted post operatively and maintained until the INR is again therapeutic. Unsurprisingly, the practice of bridging has been the subject of debate and discussion (4, 5). In the December 2012 issue of Thrombosis and Haemostasis, Feng et al. (6) presented a meta-analysis which helps shed light on this dilemma in the setting of cardiac device implantation. The authors analysed six studies to examine the endpoints of pocket haematoma requiring intervention, thromboembolic events and duration of hospital stay. It is important to note, however, that only one study included was performed in a prospective randomised controlled fashion, while four were retrospective, and three of them were not matched or balanced between groups. In patients who underwent discontinuation of OAC (range 2-5 days prior to surgery), bridging therapy with unfractionated heparin (UFH) or low-molecularweight heparin (LMWH) was instituted when the international normalised ratio (INR) 2.0. Hence the time required to achieve therapeutic OAC is likely the largest factor influencing hospital stay, with a median of five days in the heparin group and two days in the OAC group. However, there was significant heterogeneity in the length of hospital stay and the shorter hospital duration in the uninterrupted OAC group should not be overstated. Many patients that are at high-risk for thromboembolism are often also taking concomitant anti-platelet therapy. The present analysis cannot be generalised to such patients, as only two studies among six provided data on concomitant usage of aspirin (range 39.6-77.1%) and clopidogrel (only 1.3-4%). Of note, the usage of aspirin was higher in the group with bridging therapy, which may represent a bias from treating physicians as well introducing an additional bias towards haematoma formation. Note also that the present study does not consider use of newer OAC (e.g. dabigatran, rivaroxaban and apixaban), which are increasingly being used in clinical practice (7, 8). Specific peri-procedural management for patients whilst on these novel OACs is advised (9), as well as management of bleeding complications should these occur (8, 10, 11). Prediction of bleeding complications during bridging therapy