Perspective:

This very important study should inspire us to question many assumptions about anticoagulation and interrupting anticoagulation. Most clinicians believe they know that interrupting warfarin lowers periprocedural risk of bleeding in patients receiving warfarin anticoagulation. Further, most clinicians believe that bridging with heparin will minimize the period of time not anticoagulated, and therefore, mitigate any risk of interruption, without additional bleeding risk. This study clearly shows that these assumptions are not valid. The data suggest that bridging with heparin actually increases the risk of procedural bleeding dramatically, without any demonstrable increase in thromboembolic risk, although the study was underpowered to study thromboembolic events. Too often, we do things without evidence of benefit, but with only the presumption that doing something is better than doing nothing. Along with minor dental procedures (including extractions), which have been shown to be well tolerated during continuation of warfarin anticoagulation, we must now strongly consider adding pacer and ICD implantation—not because anticoagulation patients are not at increased risk of device-pocket hematoma, but because bridging with heparin likely only increases that risk. Primum non nocere.