Thanks for sharing this important article. However I would hardly agree that this is a nail in the coffin of peri-operative anticoagulation. This was after all a metanalysis of observational data. Of the 34 studies in the analysis, 33 were of observational design, the authors state in the Results section "study quality was generally poor".

This observational design of the studies increases the risk of bias.  The authors discussed this possibility stating that "high thromoembolic-risk patients may have preferentially received bridging therapy, where low thromboembolic-risk patients did not".

Furthermore, from my perspective as a pulmonary critical care physician who regularly manages patients with venous thromboembolic disease, a fifth of the patients enrolled had this disease. But there is NO seperate data detailing what percentage of these patients were actually receiving full dose anticoagulation, therefore it is difficult to make any firm conclusion in this group of patients or any of the groups for that matter. 

This data does serve to heighten the awareness of bleeding risks with bridging anticoagulation as well as identify how soreley we need RCT data. Hopefully that will be coming in the not too distant future from  BRIDGE and PERIOP-2. 

 

It may well be that those patients who were "assigned" to bridging anticoagulation by their treating physicians were higher risk than those patients who were "assigned" to no therapy or limited therapy in the periprocedural period. It is literally impossible to control for every confounding factor. I think the headline that this is "another nail in the coffin" of bridging anticoagulation is sensationalistic and does not reflect the quality of the evidence being presented here. Only RCT data can really provide a mature and appropriate assessment of the risk-benefit balance of continuing vs interrupting some form of periprocedural anticoagulation.