Blood Thinners (Oral Anticoagulants) For Atrial Fibrillation: Who Should Take Them and Which One To Take

The most serious  adverse consequence of having atrial fibrillation is stroke. Since we have safe and effective ways of preventing afib-related stroke with oral anticoagulant drugs (blood thinners), a major decision for the newly diagnosed patient with atrial fibrillation is “should I take a blood thinner?”
To answer this question the afibber should engage in a lengthy discussion with his/her health-care provider which results in a shared and informed  decision. Such discussion must cover your risk of stroke, the benefits of blood thinners in preventing stroke, the bleeding risks of blood thinners and the pros and cons of the five oral anticoagulants available to prevent stroke.

Estimating Your Risk of Stroke With Afib

The best way we have of estimating a patient’s risk of stroke if they have atrial fibrillation (AF) is by the CHA2DS2-VASc scale (which I like to call the Lip scale)
Stroke Risk EstimationThis scale take the factors we know that increase the risk of stroke and assigns 1 or 2 points. The acronym comes from the first letter of the factors that are known to increase risk as listed to the left.
Most of the factors get 1 point, but prior stroke (S) and age>75 (A) get 2 points.
We then add up your points and use another chart (or app) to calculate the risk of stroke per year.
CHA2 stroke riskYour risk of stroke is very low if you have zero risk factors; it gets progressively higher as you reach the maximum number of 9.
Treatment with an oral anticoagulant (OAC),  either warfarin, or one of the four novel anticoagulant agents (NOACS), is recommended when score is >/=2 corresponding to a  risk of stroke  above 1-2% per year.
These blood thinners have consistently been shown to lower your risk of stroke or systemic embolization (when a clot from the heart goes somewhere other than the brain) by almost 70%.
The higher the risk, the more the benefit of these blood thinners in preventing stroke.
Both European and American guidelines recommend using the CHA2DS2-VASc score for initial risk stratification. The European  guideline recommends OAC therapy for males with a CHA2DS2-VASc score ≥1 and for female patients with a score ≥2., whereas the American guideline recommends use of OAC if the CHA2DS2-VASc  score is ≥2 for men and women.
I’ve been using the CHA2DS2-VASc scale for several years in my afib patients. I try to review the patient’s risk of stroke and their risk of bleeding during every office visit, and decide whether they should be on or off an OAC.

Bleeding From Blood Thinners

All OACs cause increase bleeding. They don’t discriminate between bad clots that cause strokes and good clots that stop you from bleeding.

If you’re taking one you are more likely to have nose bleeds, bleeding into the intesitnal tract or urine and you will bleed longer when cut and more profusely if in an accident. In lower stroke risk patients, the bleeding risk of OAC of 1% per year may outweigh the benefits conferred by stroke reduction.
I wrote a post entitled “Why Does The TV Tell me Xarelto Is a Bad Drug” which points out that law suits against the makers of the newer OACs are frivolous and that these NOACs are likely more safe and effective than warfarin.
In recent years, four new drugs for reducing strokes in patients with atrial fibrillation which are much less influenced by diet and medications have gained approval from the FDA. These are generally referred to as “novel anticoagulants” reflecting their newness, different effects from warfarin or aspirin, and their blood thinning properties.  The first  (brand name Pradaxa) was released to much excitement and fanfare in October, 2010.  The press release for this approval read as follows:

PRADAXA, an oral direct thrombin inhibitor2 that was discovered and developed by Boehringer Ingelheim, is the first new oral anticoagulant approved in the U.S. in more than 50 years. As demonstrated in the RE-LY® trial, PRADAXA 150mg taken twice daily has been shown to significantly reduce stroke and systemic embolism by 35 percent beyond the reduction achieved with warfarin, the current standard of care for patients with non-valvular atrial fibrillation. PRADAXA 150mg taken twice daily significantly reduced both ischemic and hemorrhagic strokes compared to warfarin

What was very clear from the study with Pradaxa  and stated very clearly in all publications and patient and doctor  information sources was that just like warfarin, patients could have severe bleeding complications, sometimes fatal. Overall serious bleeding complications were about the same (the rate of major bleeding in patients Pradaxa  in the RE-LY trial was 3.1% versus 3.4% in the warfarin group) but Pradaxa had about 50% more bleeding from the gastrointestinal tract and warfarin about 50% more bleeding into the brain.
Another big difference between the novel anticoagulants and warfarin is that we have antidotes (Vitamin K, fresh frozen plasma) that can reverse the anticoagulation state rapidly for warfarin but until recently none for the newer drugs. (There is now available an antidote for Pradaxa).  This information also was made very clear to all doctors prescribing the medications in the package insert and educational talks. Despite this, in the major trials comparing these newer agents to warfarin, the newer agents were as safe or safer than warfarin.
The most feared bleeding complication on all OACs is bleeding into the head (intracranial hemorrhage). The risk of ICH is between 0.2 to 0.4 percent per year on warfarin. Studies show with the NOACs the risk is about half of the risk on warfarin.

Should You Take a NOAC or Warfarin?

Once the decision has been made to start a blood thinner, the next question is whether to take warfarin or a NOAC. Warfarin (brand name Coumadin) has been utilized since the 1950s  and prior to 2010 was  the only drug available for doctors to reduce clot formation in the heart and susbsequent strokes.. Warfarin is only effective and safe within a narrow window and its effects are strongly influenced by Vitamin K in the diet and most medications. Thus, frequent blood testing and adjustment in dosage is needed, and close monitoring of diet and changes in medications. Even with this close monitoring, serious and sometimes fatal bleeding occurs frequently with warfarin.
Here is a patient information sheet on warfarin which gives you an idea of issues you will need to be aware of when taking the drug. (WArfar patient handout)
If you do a Google search on warfarin you will quickly discover that it is used as a rat poison. Scientists isolate the chemical from  sweet clover that was causing cows to bleed and then developed a more potent form that they named warfarin in the 1940s. After developing blood tests that allowed the drug to be used safely  to dissolve clots it was approved for human use in the 1950s.
Warfarin or more potent variations on its chemical structure have been utilized as rat poison since the 1940s.
The rats are consuming much larger quantities of the blood thinner and are clearly not being monitored for blood thinness.
Some despicable sites peddling alternative or natural products such as this “Healthy Habits” site engage in fear-mongering over the warfarin/rat poison connection in order to promote totally unproven products. Healthy Habits indirectly suggests  that Nattokinase : is a safer, more effective natural alternative to warfarin” This remarkable enzyme has the ability to dissolve blood harmful clots involved in heart disease and strokes without upsetting normal healthy clotting.”
Such misinformation is dangerous and could lead to patients stopping a life-saving medication and suffering a stroke.
By the way, in this Xarelto (another NOAC competitor) ad, Screen Shot 2016-06-29 at 2.21.20 PMKevin Nealon says he chose Xarelto over warfarin because he wanted to eat salads. This is a common misconception and the makers of Xarelto should be ashamed for promulgating it.
I tell my patients it is fine to eat green, leafy vegetables while taking warfarin. The Vitamin K in the vegetables does influence the effectiveness of warfarin thinning blood but this is why we check the blood test to determine the appropriate dosage of warfarin for you and your personal dietary Vitamin K consumption , be it high or low.

Novel Oral Anticoagulant Drugs

The newer OACs, in contrast to warfarin do not require blood tests for monitoring of their efficacy because their levels are not significantly influenced by changes in diet or most medications.
In head to head studies versus warfarin four of these NOACs have demonstrated at least similar efficacy in preventing stroke and at most similar bleeding risk.
Due to their perceived advantages most new prescriptions for OACs are for NOACs. In contrast to 2014 American afib guidelines which don’t state a preference, the most recent European afib guidelines recommend choosing a NOAC over warfarin when initiating anticoagulant therapy in patients who are eligible for NOACs. (Ineligible patients include those with mitral stenosis, mechanical heart valves and end-stage renal disease.)
The ESC guidelines published in 2016, , make choosing a NOAC over warfarin a IA recommendation. This means there is a consensus that the treatment should be recommended (Class I recommendation) and that there is strong evidence from randomized controlled trials to support it (Level A.)
I have decided to primarily use Eliquis (apixaban) as my NOAC of choice based on my comparison of the different NOAC studies. If a patient’s insurance covers another NOAC better , making it cheaper then  I am happy to switch.
Because these four NOACs are new and brand name they are significantly more expensive than warfarin. Cost varies substantially based on type of insurance coverage and we can only determine how much a patient will pay for any given NOAC based on writing a prescription and having a pharmacy check out the cost.
I have found some patients paying nothing for their NOAC whereas some are paying several hundred dollars monthly. The more NOACs cost, the more likely the patient and I are to choose warfarin.
While waiting to determine if cost is going to be prohibitive I will typically provide the patient with samples of the NOAC chosen. The pharmaceutical companies making these NOACs are clearly making substantial profits off them and they are happy to provide lots of samples to doctors to influence the doctors to utilize their product.
NOACs are being extensively promoted both to physicians and directly to patients. Physicians have to be especially careful to make sure they are presenting a true summary of the relative risks and benefits of warfarin versus NOACs in light of these constant attempts to influence them.
Despite now having four NOACs with similar benefits and ease of use compared to warfarin, the cost of these agents doesn’t seem to have declined significantly from when the first NOAC came on the market. Personally, I would love to see Medicare step in and negotiate significantly lower costs for American senior citizens.
An abstract at the ACC meeting in March of 2017 suggested  a reduction in medical costs with NOACs despite their high costs. This was related to a lower rate of major bleeding complications: Xarelto cost $542 per patient compared with warfarin’s $500, or $42 more. Pradaxa, cost $367 to warfarin’s $452, saving $85.  Eliquis cost  $286 charge against warfarin’s $537 resulting in $251 in savings. Data were from from a study of U.S. Medicare patient records.

Aspirin May Not  Prevent Stroke In Afib

Many patients consider aspirin to be  a “blood thinner” that has some benefit in preventing clots and strokes in patients with afib. However,  aspirin is not considered a blood thinner or anticoagulant and is more properly  termed an anti-platelet agent.
I used to consider aspirin at doses of 120 to 200 mg daily provided some protection against stroke in afib and put afib patients on aspirin who were low risk for stroke or would not or could not take OACs.
More and more, however, experts are reaching the conclusion that the substantial bleeding complications from aspirin usage outweigh its very slight benefit in stroke prevention.
The most recent ESC guidelines, in fact, list aspirin therapy for stroke prevention in atrial fibrillation as IIIA. That means that overall it is felt to be harmful (III) with a high level of evidence (A.)
Bleeding risks for aspirin are similar to warfarin and Eliquis. Thus, patients should not consider aspirin as a safer alternative to prevent stroke in afib.
Finally, do not take any “natural” supplement that has been promoted as a blood thinner. These are neither safe nor effective. Remember that it took years of scientific investigation and careful testing in animals then humans before warfarin (the agent in sweet clover that caused cows to bleed ) was transformed into a safe and effective anticoagulant.
Antiemboligenically yours


19 thoughts on “Blood Thinners (Oral Anticoagulants) For Atrial Fibrillation: Who Should Take Them and Which One To Take”

    • Diane,
      I’ve had a pill in the pocket anticoagulation article in the works for some time but never seem to have the time to get it into publishable form.
      In my practice I factor in AF burden (especially if it is zero) into the conversations on NOAC or not.
      Dr. P

  1. I copied and pasted another question but what really seems difficult to get an answer to if I go into Afib (my score is 1 Female) does the time in Afib determine that chance for stroke? Is 15 minutes the same risk as 5 hours as long as it is contained? Thanks Here was the original question. BTW AS PIP I take Propanolol

    While I understand the ischemic and bleeding stroke risk calculations, I wonder if there is any advice based on atrial fibrillation burden. If the episodes are few and far between, monitored carefully by the Kardia device and Apple Watch, and suppressed quickly by the pill in the pocket approach of triple dose of flecainide, does that change the weighting of the ischemic/bleeding stroke tradeoff? Has any research been done on this issue?

    • I have been preparing a post on this very topic. Hopefully I’ll have enough time and motivation to get it out in the next few weeks.

  2. I had ablations at Cleveland clinic 19 years ago for lone af that were successful. No problems since. New cardiologist wants me to take Pradaxa as I am female and 69. With no afib why? Because I have a score of 2? I take Advil for migraines several times a month, no other prescriptions and am healthy. I don’t understand this recommendation.

    • Vicky,
      That is a great question. I see this type of situation handled in multiple ways.
      I hope to write a post discussing it soon.
      You are a woman with a CHADS2 of 2 (1 for being a woman). With this score and a history afib of any kind guidelines recommend either no treatment or a NOAC.
      A third option is to monitor for recurrence of afib and start NOAC if any is detected. Monitoring could be by a device, by periodic long term monitor or by daily monitoring with a reliable device like AliveCor or Apple Watch.

  3. Dear Dr Anthony I have been a marathon athlete since 1975 and during 2008 while having my annual medical exam the doctor discovered that my heartbeat was irregular (this was by accident as it was his intern that picked it up). My subsequent visit to a cardiologist revealed that I had AFib. I was 64 years old at the time and still running marathons. Owing to a knee injury I have stopped running and have opted for cycling and walking. My question to you is how does AFib develop, what causes it to develop, has long distance running have anything to do with it? I cycle 10 miles a day over a very hilly course and also walk 2-5 miles a day.
    Two years ago when I was 73 my cardiologist said when I turn 75 I should visit him again and then he would prescribe a blood thinner, in this case it was Eliquis. What is so magical about the age 75 and taking a blood thinner. According to your Lip scale I score 2 points (age 75) and fall into that category. The other option is having the ‘Watchman’inserted into the LAA of my heart, what are your thoughts on this procedure compared to taking a blood thinner.

    • Lawrence,
      I’ve written about the causes of afib ( In addition to the standard causes, those who engage in endurance sports appear to be a higher risk for afib. John Mandrola has written a book about this called the Haywire Heart you might want to check out.
      Great question on the magic number of 75 years. We know that the risk progressively rises as we age and around age 75 that risk is considered high enough (all other factors being equal) compared to <75 years that you are better off taking an anticoagulant.
      As to whether or not you should have the Watchman, that is a complicated question and I've been meaning to write about the Watchman for some time. Perhaps I'll be motivated by your case to write one soon…

  4. While I understand the ischemic and bleeding stroke risk calculations, I wonder if there is any advice based on atrial fibrillation burden. If the episodes are few and far between, monitored carefully by the Kardia device and Apple Watch, and suppressed quickly by the pill in the pocket approach of triple dose of flecainide, does that change the weighting of the ischemic/bleeding stroke tradeoff? Has any research been done on this issue?

  5. Hi,
    I really appreciate the true scientific approach you take in your articles. When I say true scientific approach, it reflects your openness to discuss cardiac matters whether discussion could considered for/against the doctor/ patient.
    As a TVCAD patient myself (6 DES), I have suffered too much from cardiologist who attribute effects of drugs or therapy or something they are unable to explain to psychological state. Although psychology has a great impact but ignoring your patient descriptions and considering oneself almost superior in understanding is truly unscientific.
    A truly scientific approach lets the data decide the outcome rather than opinions. Thank you for your great articles
    I was was wondering if I want to get consult online what would the right way to do so…I am not in United States
    I’ll appreciate if it’s possible to be reviewed by you…I can pay online also
    Many thanks for your excellent approach

    • Mazhar,
      Thanks for your kind words. I am actively working with my medical group to come up with a process to be able to remotely consult for patients like you. I’ll let you know when/if that becomes a reality.

  6. Thank you. My husband just went off Warfarin, to a new drug – Riveroxaban. No salads though either! He had to ask – in the UK the National Health Service simply prescribe Warfarin – its cheaper…..

  7. I have three female relatives who lived well into their 90s, and one to 102, who took no medications until sometime after being put under a new doctor were, unknown to family who would have tried to stop it, put on a statin. They all three had hemorrhagic stroke within a year.
    Statins are documented to cause hemorrhagic stroke. Of course, we cannot prove that statin caused that stroke in that family member.
    I continue to post hoping you might see you’re not in the least a “skeptical” cardiologist.
    We are indigenous and small. We all, female and male, have cholesterol levels over 10. We are all thriving w/o those drugs.

    • Janey,
      Thanks for your comments. They prompted me to look into the association of statins and hemorrhagic stroke and although one study (SPARCLE) did raise concerns of an association most data don’t support that.
      One of the reviews I looked at is typical of expert guidance and concluded
      “Concern was raised by previously published post hoc analyses and observational studies that noted an increased risk of hemorrhagic stroke in patients receiving a statin. Subsequent studies have demonstrated conflicting results regarding the role of statin therapy on hemorrhagic stroke risk and patient outcomes. New evidence suggests that statins taken prior to or continued during admission for intracerebral hemorrhage (ICH) may be associated with positive outcomes. Evidence also suggests deleterious outcomes resulting from the abrupt discontinuation of statins upon hospital admission for multiple disease states including ICH.”
      Another recent review concluded
      “Based on the current data, we think statin use in the prevention and treatment of atherosclerotic cardiovascular disease does not increase the risk of ICH in most conditions. The pleiotropic effects of statins such as inhibition of inflammation and protection of BBB may improve the prognosis of acute hemorrhagic stroke patients. But for elderly patients who have history of ICH and poorly-controlled hypertension, statins especially used in high dose, may increase the risk of ICH. ”
      I would agree that we should be very cautious in the use of statin in elderly patients. The evidence supporting their benefit in those over age 75 is weak and the potential for side effects is highest in this population, especially those on multiple other drugs and with kidney problems. I try to minimize doses and only utilize in those at high risk.

  8. As a skeptical patient, I actually spent a day investigating the source for the “you’re 5 times more likely to have a stroke with A-Fib” claim. It’s an interesting investigation. I ultimately opted not to take a thinner, even though I’m now a 1 instead of a 0. But, as previously noted, I’m not a great patient either.
    Is your problem with “natural” supplements due to lack of regulation, lack of rigorous scientific investigation, or both? The reason I ask is that I don’t see a great track record for either, from a purely practical standpoint.
    But, as you’ve come to know, I’m very skeptical…
    Great article summarizing the scoring etc. I wish I had found something similar 11 months ago.


Please leave your comments. The skeptical cardiologist loves feedback. He reads all and replies to all that warrant a reply.