Anticoagulated patients facing high-risk endoscopic procedures might benefit from a switch from warfarin to a direct oral anticoagulant (DOAC), new research suggests.

A Japanese study of almost 17,000 patients found a significantly higher risk of gastrointestinal bleeding within 30 days of the procedure in patients on warfarin (12%) compared to DOACs (9.9%).

The study also found heparin bridging was associated with an increased risk of bleeding and thromboembolism in both warfarin and DOAC users.

The retrospective study identified endoscopic submucosal dissection and endoscopic mucosal resection as two of the most high-risk procedures for bleeding.

Commenting on the study, Associate Professor William Tam said the study raised a number of important questions given the increasing use of DOACs.

 “Up to 50% of people with AF are currently using these medications. We will encounter them commonly in clinical practice and should therefore be aware of the risks and benefits of these types of drugs.”

“The risks of post-procedural bleeding was in fact higher in warfarin than in these new drugs.”

Associate Professor Tam, chair of the Australian Gastrointestinal Endoscopy Association, said the opposite had been shown in post-hoc analyses of cardiac trials.

While more data and prospective studies were required, he said the idea of switching drugs in some patients was a novel concept.

“If you are going to do a high-risk procedure, and the person is on warfarin you might consider switching them to one of these new drugs. Currently we would only switch if the patient was unstable or unhappy with warfarin.”

Associate Professor Tam said the study provided convincing evidence against the use of heparin bridging.

“We commonly use heparin bridging to avoid clots but this retrospective study showed not only did it not prevent clots but it in fact increased the risk of bleeding.”

“It actually questions whether we should be doing this. The answer is probably no.”

A sub-analysis of individual agents showed patients on warfarin had significantly more gastrointestinal bleeding than patients on rivaroxaban and significantly more thromboembolism events than patients on rivaroxaban or dabigatran.