Bleeding Events, Hospitalizations Among Patients with AF

Atrial fibrillation patients on novel oral anticoagulants (NOACs) don’t need to be managed as closely as patients on warfarin, so it’s widely assumed that bleeding events associated with the newer anticoagulants are more dangerous. However, that might not be the case, according to Blake Charlton, MD a clinical fellow at the University of California, San Francisco School of Medicine.

His review of insurance claims for more than 3,000 atrial fibrillation patients who were hospitalized for bleeding events after taking warfarin, dabigatran, or rivaroxaban revealed no significant differences in rates of 30- and 90-day all-cause mortality. The study also revealed hospitalizations for patients on NOACs were approximately two days shorter than for patients who were on warfarin.

Dr. Charlton took a few minutes to discuss the interesting findings, which highlighted exciting opportunities to save health care dollars and, more importantly, patient lives.

Why did you conduct this research?

The extended hospital stays associated with warfarin were perhaps the most interesting aspect of this study, but the most important finding is the lack of a difference in mortality rates between the two patient groups. Whenever a patient is started on full anticoagulation, you're always most concerned about creating a bleed. When I began this analysis, we were very used to treating patients with warfarin-related bleeds but were just getting used to using NOACs.

A large difference in mortality rates with the NOACs would have moved the needle for me with respect to whether or not I advocated starting patients on one of the newer agents. That was the original intent of the study — to make sure there wasn’t a clear signal of increased mortality risk with the NOACs. Importantly, there wasn’t.

Were you surprised by the findings?

Several years ago, reversal agents weren’t available for the NOACs. Now there’s a reversal agent for dabigatran and soon I think we’ll have reversal agents for the Xa inhibitors. In my mind, back before reversing NOACS was a possibility, it made perfect sense to use warfarin, which has several reversal agents available, to treat patients who are bleeding. That's why we looked at secondary outcomes such as which patients stayed in the hospital longer and which ones were going to the ICU more often. When we analyzed those outcomes, we found that the reversibility of warfarin did not confer a benefit. That was very surprising to me.

Why didn’t warfarin’s reversibility impact lengths of stay?

I tried to come up with reasons, including a sensitivity analysis that excluded patients who were hospitalized with intracranial hemorrhage, which has a higher occurrence in patients on warfarin. Restarting warfarin is labor intensive, so I assumed patients were remaining in the hospital longer while waiting for their INR results. However, I conducted another sensitivity analysis of patients who did not continue anticoagulation therapy and the difference in lengths of hospital stay persisted. Does that mean NOAC-related bleeding is less complicated and less onerous on the healthcare system? That’s uncertain.

There were some confounding issues we couldn't address in the study, but the difference in lengths of stay was of a significant magnitude—it wasn't a half day or a quarter of a day, it was several days—so it's hard to imagine that was completely due to residual confounding. The findings highlighted a fruitful area for more granular research focused on determining which anticoagulant provides the most benefit in terms of preventing patients from having to endure longer hospital stays.

What’s the best way to decide which anticoagulant is optimal for patients?

There are many nuances involved and experts spend all day, every day thinking about that. It’s a rapidly evolving patient care area and the decision ultimately comes down to sound clinical judgment and patient preference. NOACs are showing clinical benefits in very encouraging ways, but it’s important to remember that trials showing their non-inferiority to warfarin were designed to be successful in the most promising patients.

Keep in mind that the newer agents have proven to be harmful in patients with mechanical cardiac valves and no data that I’m aware of show the agents can be safely used in patients with valvular atrial fibrillation. I’m also hesitant to use them in patients with extremely bad renal function.

Do NOACs have the potential to reduce healthcare spending?

Maybe. This could be an area where we could realize a significant amount of healthcare savings because of the untoward downstream effects of prolonged hospitalizations, particularly in frail patients. My analysis did not look at incidences of bleeding associated with the different agents, but trials run by the drug companies that make NOACs and some surveillance data are favorable to the NOACs. You also have to consider the cost benefit of reducing really expensive hospitalizations and the overall risks patients face—allergic reactions, falls, and delirium—during prolonged hospitalizations against the increased costs of the NOACs. It’s going to be a complicated calculation to determine if NOACs actually result in cost savings.

—Dan Cook

For more Pharmacy Learning Network articles, visit the homepage

To learn about Pharmacy Learning Network Live meetings, click here