How dangerous are NSAIDs in patients with AF?

One of the most commonly asked questions in the office is the treatment of arthritis pain. This comes up because of the concern over taking NSAIDs (non steroidal anti-inflammatory drugs) in patients who are on anticoagulants (such as warfarin, or dabigatran, rivaroxaban, apixaban).

My views on this matter have changed.

But first, I want to mention a study published in JACC that addressed the issue of NSAID use in pts with AF who take an anticoagulant.

This was a sub-analysis of the RELY trial, which pitted dabigatran vs warfarin. Remember, in the original RELY trial, the 150mg dose of dabigatran did better than warfarin at reducing strokes and had similar rates of bleeding

This sub-study looked back at nearly 3000 (of 18,000 total) patients who used NSAID at least once during the trial.

The results were not surprising:

• NSAID use increased the risk for major bleeding.
• GI bleeding in particular was almost double w NSAID use
• NSAID use also associated w/ a 50% increase in ischemic stroke.

In a news piece on theheart.org|Medscape Cardiology, senior author Dr M Ezekowitz said “avoid NSAIDs in patients in anticoagulated pts.” An editorialist, Dr. Sam Schulman, said telling pts to take acetaminophen is not a solution, as most patients say it does nothing for their arthritis because it’s not an anti-inflammatory agent.

These used to my views. It was simple. Avoid NSAIDs when on AF drugs because they are dangerous.

I can’t believe it took me many years to change from this decidedly doctor-centric, non-patient-centered point of view.

I am now much more opened-minded about NSAID use.

My change started on social media. I remember pontificating on NSAIDs dangers when a smart person on Twitter suggested the competing dangers of immobility in older people with pain. That made me think. Then, you can add to that concern, the new revelations on the addictive properties of opioids.

And finally, perhaps most importantly is the problems of concluding NSAID danger from this sort of observational analysis. The authors acknowledge many limitations of their paper:

• No info on dosing
• No info on reasons for taking NSAIDs; baseline data did not capture presence of inflammatory diseases.
• Crucially, we don’t know whether pts who took NSAIDs had more bleeds and ischemic strokes because they were sicker or because of the drug.

My NEW approach to patients with the problem of arthritic pain and AF is…

• Suggest acetaminophen first. I think a lot of its non-effect stems from the nocebo effect: pts think it won’t work. I’ve looked at the arthritis literature and acetaminophen’s inferiority to NSAIDs in blinded studies is not clearly demonstrated.
• Talk about dosing. NSAID dangers probably depend on dose. I try to get patients to use less of it.
• Discuss risks of both immobility and drug adverse effects

I no longer just proscribe NSAIDs because they are dangerous. Life is dangerous.

JMM