Misdosing NOACs Common in Everyday Practice, With Implications for Stroke and Bleeding

Clinicians appear to be misusing the reduced “renal” NOAC doses, with worrying consequences.

Misdosing NOACs Common in Everyday Practice, With Implications for Stroke and Bleeding

NEW ORLEANS, LA—A fresh look at how US physicians are dosing patients with non-vitamin K antagonist oral anticoagulants (NOACs) shows that some patients frequently get doses both higher and lower than what the labelling recommends. The issue is particularly problematic in patients with severe kidney disease and the elderly, in whom physicians may be using other subjective considerations to alter NOAC dose, potentially putting their patients at risk.

Xiaoxi Yao, MD (Mayo Clinic, Rochester, MN), presented the analysis last week here at the American Heart Association Scientific Sessions 2016.

“For both cardiologists and nephrologists who treat a lot of people with chronic kidney disease (CKD), this dosing question actually comes up a lot, because we treat a lot of complex patients with mild-to-severe CKD,” Yao said. “Sometimes patients are very old and fragile but have relatively normal GFR or renal function. Physicians may be very concerned about their bleeding risk, so they may just want try to be safe and prescribe a low dose. And at the other end [of the spectrum], sometimes we see relatively healthy patients with poor renal function, and clinicians may be very concerned about their stroke risk. So they are both overdosing and underdosing.”

Yao’s analysis drew on the Optum Labs claims database that includes pharmacy information, blood-based lab results, and sociodemographic characteristics for over a million privately insured Americans. In all, the study included 3,713 patients with A-fib started on apixaban (Eliquis; Bristol-Myers Squibb), 4,724 on dabigatran (Pradaxa; Boehringer Ingelheim), and 6,428 on rivaroxaban (Xarelto; Janssen Pharmaceuticals), all of whom also had linked data on kidney function.

To explore over- and underdosing, Yao and colleagues followed the labelled recommendations for who should be receiving a reduced dose based on their renal function: creatinine ≥ 1.5 mg/dL for apixaban, an estimated glomerular filtration rate (eGFR) < 30 mL/min/1.73 m2 for dabigatran, or an eGFR < 50 mL/min/1.73 m2 for rivaroxaban.

They found that 1,473 patients (10%) of patients met the renal criteria for receiving the reduced dose, yet 43% of these patients received the standard NOAC dose. This overdosing was associated with a higher risk of major bleeding, but not a reduced risk of stroke. “In other words, overdosing was related to worse safety, without a benefit in effectiveness,” Yao said.

Among the remaining 13,392 patients with no indication for a dose reduction, 13.3% received the lower dose. This reduced dose was associated with a higher risk of stroke but no difference in the risk of major bleeding among the apixaban-treated patients. Here, said Yao, underdosing was related to worse effectiveness, without a benefit in terms of safety. No significant differences were seen for either stroke or bleeding with under- or overdosing in the dabigatran or rivaroxaban groups.

“I think one thing clinicians always need to do is check the renal function of the patient and make sure that in those who have low GFR, their dose is adjusted accordingly,” Yao told TCTMD. “That would address the overdosing problem. But for the underdosing, we only found a significant underdosing effect for apixaban so I think the message is, apixaban is actually quite a safe drug. For those patients who don’t meet the dose-reduction criteria, I think the high dose may be safe and effective and if you use a low dose, you may actually increase the rate of stroke.”

Commenting on the study for TCTMD, Freek Verheugt, MD (Hartcentrum OLVG, Amsterdam, the Netherlands), agreed that with apixaban, “the reduced dose is counterproductive: more strokes, but doesn’t help you with bleeding. So what you may say is that apixaban is the only correctly dosed NOAC, because if you give the standard dose, you have very few strokes, and the underdosing is not helpful for bleeding.”


As for why dabigatran and rivaroxaban didn’t show the same significant differences as apixaban, Yao suggested that different factors may be at play for the different agents. In the case of dabigatran, rates of stroke or systemic embolism, as well as bleeding, followed a pattern similar to that seen in apixaban-treated patients, and the lack of statistical significance may simply be a question of low patient numbers. In the United States, Yao noted, physicians do not have access to the lower 110-mg dose that was tested in the pivotal trial—the FDA controversially approved an untested dose of 75 mg for patients with a renal indication. “Relatively few patients receive this dose, [possibly] because clinicians feel that the 75-mg dose was not properly tested.”


That’s a theory that Verheugt strongly supports. He noted that underdosing in the United States “is severe underdosing, because of the untested 75-mg dose,” making the findings of Yao’s analysis “a bit skewed” for dabigatran.


In the rivaroxaban-treated patients, Yao and colleagues had a different theory. Whereas the renal dose for both apixaban and dabigatran in the United States is exactly half the standard dose, the renal dose for rivaroxaban is 75% of the standard dose.

“For rivaroxaban, we saw actually the same risk of both stroke and bleeding, so one hypothesis we have is, because the low dose of rivaroxaban that we have is 75% of the full dose . . . maybe for rivaroxaban, it’s not that big of a dose reduction,” and therefore less likely to lead to as many problems when the drug is over- or underdosed.

Other work by Mayo researchers, however, has shown that rivaroxaban at its standard dose, while equivalent to other NOACs in terms of efficacy, may be associated with more bleeding.

This leaves open the possibility that there might be less bleeding with rivaroxaban if it, like the other NOACs, was available in a dose that was 50% the standard strength. “We really don’t know because in the real world, we actually see rivaroxaban associated with a high bleeding risk so maybe 50% of the standard dose would be fine.” Indeed, Yao said that in Asia, 80% to 90% of patients are actually taking the reduced rivaroxaban dose, regardless of their renal function.

Verheugt agreed, noting “for rivaroxaban, underdosing is not harmful and not helpful either and probably the ‘underdosing’ is actually the right dosing."

The advent of NOAC antidotes may help with at least part of the misuse of NOAC doses, Verheugt noted, since much of the underdosing stems from concerns about bleeding risk in elderly, frail patients. For now, however, dabigatran is the only agent with an approved antidote and at least in the United States, Verheugt believes severe underdosing, leading to increased stroke risk, is a bigger issue than underdosing.

Shelley Wood is the Editor-in-Chief of TCTMD and the Editorial Director at CRF. She did her undergraduate degree at McGill…

Read Full Bio
  • Yao X. Deviation from dosing labels of NOACs may be associated with reduced effectiveness or safety. Presented at: American Heart Association Scientific Sessions 2016. November 15, 2016. New Orleans, LA.

  • Yao reports no relevant conflicts of interest.