Skip to content
PharmaceuticalReviewed July 2026 · v1.0

Anticoagulants

DOACs and warfarin — among the most effective drugs in medicine, and the most misunderstood

The evidence is exceptionally strong. Anticoagulation reduces stroke in atrial fibrillation by roughly two-thirds and reliably treats and prevents venous thromboembolism, established across multiple large randomized trials. DOACs are at least as effective as warfarin with substantially less intracranial haemorrhage. The high confidence applies to the drugs working — the genuine complexity lies in WHICH drug, for WHOM, and the interactions.

How confident is BioSignal?

Our overall position, and how sure we are of it across each dimension — including where we are not sure at all.

High confidence

The evidence is exceptionally strong. Anticoagulation reduces stroke in atrial fibrillation by roughly two-thirds and reliably treats and prevents venous thromboembolism, established across multiple large randomized trials. DOACs are at least as effective as warfarin with substantially less intracranial haemorrhage. The high confidence applies to the drugs working — the genuine complexity lies in WHICH drug, for WHOM, and the interactions.

Well-supported by consistent, high-quality evidence.

Biological Role

High confidence

Precisely characterised. DOACs directly inhibit factor Xa (apixaban, rivaroxaban, edoxaban) or thrombin (dabigatran); warfarin blocks vitamin K–dependent synthesis of factors II, VII, IX and X.

Human Evidence

High confidence

Multiple large randomized trials with tens of thousands of participants, plus decades of warfarin experience. This is among the best-evidenced areas in all of therapeutics.

Clinical Benefit

High confidence

Roughly two-thirds reduction in stroke in atrial fibrillation, and effective treatment and prevention of venous thromboembolism. Few interventions in medicine produce effects of this size.

Broader Claims

Moderate confidence

Indication-dependent. The benefit is enormous where indicated and absent — with pure bleeding risk — where it is not. 'Thinning the blood' is not a general-purpose health intervention.

Safety Confidence

Caution

The mechanism IS the risk: these drugs cause bleeding, and that is not a side effect but the direct consequence of what they do. The risk is well characterised, manageable, and in indicated patients is substantially outweighed by the benefit — but it is real and requires clinical supervision.

Research Activity

High

Active work on reversal agents, factor XI inhibitors (which may separate anticoagulation from bleeding), and optimal use in cancer, kidney disease and the very elderly.

Where the evidence stands today

How mature the science is, what kinds of evidence exist, and — the part nobody else prints — what is still missing.

9/9

steps proven in humans

Evidence-rich

Proven at every applicable step — rare, and worth noticing.

  1. Guideline / regulatory support

    Proven

    First-line in international guidelines; aspirin explicitly de-recommended for AF.

  2. Clinical outcomes

    Proven

    Hard outcomes: stroke, systemic embolism, recurrent VTE, death.

  3. Large human RCTs

    Proven

    Multiple trials with tens of thousands of participants.

  4. Small human outcome trials

    Proven

    Superseded by the large trials.

  5. Human safety data

    Proven

    Extensive; bleeding risk quantified, reversal agents available.

  6. Human biomarker / pharmacology

    Proven

    INR, anti-Xa activity — well-established pharmacodynamics.

  7. Animal

    Proven

    Extensive.

  8. Cell / in vitro

    Proven

    Coagulation cascade effects fully described.

  9. Mechanistic plausibility

    Proven

    Factor Xa/thrombin inhibition and vitamin K antagonism are precisely characterised.

The bottom line

What we know, what we think, what we don't know — and what would change our mind.

What We Know

Anticoagulation prevents strokes in atrial fibrillation — reducing risk by roughly two-thirds — and treats and prevents deep vein thrombosis and pulmonary embolism. DOACs are at least as effective as warfarin with substantially less bleeding into the brain, and without routine blood monitoring. This is one of the most solidly established benefits in all of medicine.

What We Think

For most people with atrial fibrillation or venous thromboembolism, a DOAC is the right choice. But 'most' is not 'all', and the exceptions are not preferences — they are absolute. Mechanical heart valves and antiphospholipid syndrome require warfarin, and using a DOAC in those patients causes harm.

What We Don't Know

The optimal approach in several difficult groups: advanced kidney disease, cancer-associated thrombosis, extremes of body weight, and the very elderly with high falls risk — where the fear of bleeding often leads to anticoagulation being withheld from precisely the people whose stroke risk is highest.

Active Research

Factor XI inhibitors, which aim to prevent pathological clotting while preserving the normal clotting needed to stop bleeding — potentially the most significant advance in this field in a generation. Also: better reversal agents, and clearer guidance in kidney disease and cancer.

What Could Change Our Mind

Nothing is likely to overturn the core benefit — the evidence is too large and too consistent. What would change practice is a factor XI inhibitor that delivers the same stroke protection with materially less bleeding, which would make the current risk–benefit conversation obsolete.

The biggest myth

Aspirin is an acceptable alternative to anticoagulation for stroke prevention in atrial fibrillation

Not establishedHigh confidence

It is not, and this is one of the most consequential misunderstandings in cardiology. Aspirin is an ANTIPLATELET, not an anticoagulant — a different mechanism entirely. It is markedly less effective than anticoagulation at preventing AF-related stroke while still causing bleeding, and guidelines have abandoned it for this indication. It was for years offered as the 'gentler option' to frail or elderly patients. It is not gentler in any way that matters. A great many people believe their daily aspirin protects them from a stroke caused by atrial fibrillation. It does not.

Ask BioSignal

Still have a question about anticoagulants?

Answers are retrieved from this record and the rest of the knowledge graph — never generated.

Ask about Anticoagulants

Why people take anticoagulants

Popularity is not evidence — but it is not stupid either. This explains the interest on its own terms.

Because 'blood thinner' is one of the most widely used and least understood phrases in medicine. People conflate aspirin with anticoagulation, believe they must give up vegetables, take St John's Wort alongside a DOAC without mentioning it, and — most dangerously — assume the newer drug must be the better one, including in the one group where it will kill them. BioSignal's role here is not to deflate a hyped supplement. It is to correct dangerous misunderstandings about a class of drugs that works superbly, and to say plainly that the evidence is excellent.

If you're here because…

Jump straight to the part of the evidence that answers your question.

Approval, safety and regulatory

What it is approved for, who should be careful, what remains unknown — and the limits of what this evidence can tell you.

Regulatory status

Approved. Multiple agents licensed for AF stroke prevention and venous thromboembolism

Availability

Prescription only

Sport (WADA)

Not a prohibited class (confirm at publication)

Known safety profile

The principal risk of an anticoagulant is bleeding, and this is not a side effect in the usual sense — it is the direct and inevitable consequence of the drug doing exactly what it is designed to do. In patients who are correctly indicated, that risk is substantially outweighed by the strokes and clots prevented. In patients who are NOT indicated, it is pure risk with no benefit, which is why 'thinning the blood' is not a general wellness intervention.

Common issues

  • Bruising and prolonged bleeding from minor cuts — common, and usually a nuisance rather than a danger
  • Gastrointestinal bleeding — some DOACs carry a modestly higher risk than warfarin
  • Heavy menstrual bleeding — common, under-discussed, and a frequent reason women stop taking these drugs without telling anyone
  • Nosebleeds
  • Intracranial haemorrhage — the rare, serious one, and substantially LESS common with DOACs than with warfarin

Use caution if

  • MECHANICAL HEART VALVES — DOACs are CONTRAINDICATED. Warfarin only. This is a demonstrated harm, not a preference
  • Antiphospholipid syndrome, particularly triple-positive — DOACs performed worse than warfarin. Warfarin only
  • Severe kidney impairment — DOAC dosing changes, and some are unsuitable entirely
  • Active peptic ulcer disease, recent gastrointestinal bleeding, or known bleeding disorders
  • Pregnancy — warfarin is teratogenic; DOACs are not recommended. This requires specialist management
  • Anyone taking St John's Wort — it lowers DOAC levels and can leave you silently unprotected
  • People on antiplatelets as well (aspirin, clopidogrel) — combined therapy multiplies bleeding risk and should be a deliberate, time-limited decision

Long-term unknowns

Long-term use is well characterised. The genuine uncertainties are not about the drugs themselves but about specific populations: advanced kidney disease, active cancer, extremes of body weight, and the very elderly — groups that are systematically under-represented in the trials and over-represented in the clinic.

Limits of this evidence

The evidence for benefit is exceptionally strong; the limitations are about EDGES rather than the centre. Trial populations under-represent the very elderly, the frail, those with advanced kidney disease, and those at extremes of weight — precisely the patients clinicians most often agonise over. And the real-world interaction picture, particularly with over-the-counter supplements, is poorly captured by trials, because participants rarely report what they buy in a health-food shop.

Full regulatory and sport detail
Regulatory approval
Approved. Apixaban, rivaroxaban, edoxaban, dabigatran and warfarin are licensed for stroke prevention in non-valvular atrial fibrillation and for the treatment and prevention of venous thromboembolism.
Approved indication
Stroke prevention in non-valvular atrial fibrillation; treatment and prevention of deep vein thrombosis and pulmonary embolism. DOACs are CONTRAINDICATED in mechanical heart valves.
Sport (WADA)
Not a WADA-prohibited class — confirm against the current list.
Publication note
Re-confirm licensed indications, dose-reduction criteria and contraindications against current prescribing information at publication. Anticoagulant dosing is not a topic on which to rely on a summary.

Demand, separated from evidence

Every claim people make about this, counted against what the evidence actually showed.

BioSignal evaluated 8 popular claims about anticoagulants.

Here is where each one landed — including the claims of harm, where “not established” is reassuring rather than damning.

Supported by evidence
2
Mixed evidence
1
Not established
5

Popularity is not evidence. This is simply a count of every claim BioSignal evaluated on this page, sorted by what the evidence actually showed — the full reasoning behind each verdict is in the Evidence Review below.

Supported by evidence

  • Anticoagulants prevent strokes in atrial fibrillation
  • DOACs are safer than warfarin

Mixed evidence

  • Anticoagulants are too dangerous for elderly people who might fall

Not established

  • Aspirin is an acceptable alternative to anticoagulation for stroke prevention in atrial fibrillation
  • A DOAC can be used instead of warfarin for a mechanical heart valve
  • People on warfarin must avoid green vegetables
  • Herbal supplements are safe to take alongside anticoagulants
  • You can stop an anticoagulant if you feel fine or your atrial fibrillation seems settled

The evidence review

Every claim with the reasoning behind its verdict, the doses actually studied, where scientists agree and disagree, and the questions still open.

What people claim

Every popular claim, with BioSignal’s verdict and how confident we are in it. The verdicts are always visible; open any claim to read the evidence behind it.

Anticoagulants prevent strokes in atrial fibrillationSupportedHigh confidence

Overwhelmingly. Anticoagulation reduces the risk of stroke in atrial fibrillation by roughly two-thirds, established across large randomized trials for both warfarin and the DOACs. This is one of the largest and most reliable treatment effects in medicine.

DOACs are safer than warfarinSupportedHigh confidence

For most indications, yes — with an important nuance. DOACs produce substantially LESS intracranial haemorrhage than warfarin, which is the bleeding that kills and disables. Some DOACs carry a somewhat HIGHER risk of gastrointestinal bleeding, which is far more survivable. They also need no routine monitoring and have fewer dietary interactions. 'Safer' is accurate overall, but it is a trade, not a free lunch.

A DOAC can be used instead of warfarin for a mechanical heart valveNot establishedHigh confidence

NO — and this is the most dangerous misconception on this page. DOACs are not merely less effective in mechanical valve patients; they are HARMFUL. The RE-ALIGN trial of dabigatran in this group was stopped early for BOTH excess clotting and excess bleeding. Mechanical heart valves require warfarin. A patient who switches — or is switched — to a DOAC can throw a clot and die.

People on warfarin must avoid green vegetablesNot establishedHigh confidence

A persistent and harmful myth. The goal is CONSISTENCY of vitamin K intake, not avoidance of it. Warfarin dosing is calibrated to your habitual diet — so eating a steady amount of leafy greens is entirely compatible with stable INR, while wild swings (a sudden salad phase, or abruptly stopping) destabilise it. People have avoided vegetables for years, unnecessarily and to the detriment of their health, on the strength of this misunderstanding.

Herbal supplements are safe to take alongside anticoagulantsNot establishedModerate confidence

Several are not, and one is genuinely dangerous. ST JOHN'S WORT induces CYP3A4 and P-glycoprotein, lowering DOAC blood levels — meaning a person taking it may be silently UNPROTECTED from stroke while believing they are covered. Fish oil, vitamin E, garlic, ginkgo and high-dose curcumin all have at least theoretical antiplatelet effects that may add to bleeding risk. Tell your clinician and your pharmacist everything you take, including things bought in a health-food shop.

You can stop an anticoagulant if you feel fine or your atrial fibrillation seems settledNot establishedHigh confidence

No. Anticoagulation in atrial fibrillation is based on stroke RISK, not on symptoms — and AF is frequently asymptomatic. Feeling well does not mean the risk has gone, and stopping without advice removes protection from a stroke that may be disabling or fatal. Restoration of normal rhythm, including after ablation, does not automatically mean anticoagulation can stop.

Anticoagulants are too dangerous for elderly people who might fallMixedModerate confidence

This is a genuine clinical judgement, and it is very frequently made in the wrong direction. Falls risk is routinely used to withhold anticoagulation from elderly patients — but analyses suggest a person would need to fall an implausibly large number of times per year for the bleeding risk to outweigh the stroke protection. The people denied anticoagulation for falls risk are often precisely those at the highest risk of a devastating stroke. This is a conversation to have with a clinician, not a reason to assume the drug is unsuitable.

Doses used in human studies

What was actually given to participants in the research. These are descriptions of studies, not recommendations.

Adults with non-valvular atrial fibrillation

Intervention
Apixaban (Eliquis)
Dose
5 mg twice daily; 2.5 mg twice daily if dose-reduction criteria are met
Duration
Long-term
Outcome
Stroke and systemic embolism
Notes
Dose reduction depends on age, weight and kidney function. Not a self-adjusted decision.

Adults with non-valvular atrial fibrillation

Intervention
Rivaroxaban (Xarelto)
Dose
20 mg once daily; 15 mg if renal impairment
Duration
Long-term
Outcome
Stroke and systemic embolism
Notes
Must be taken WITH FOOD — absorption is substantially reduced on an empty stomach.

Adults with venous thromboembolism (DVT/PE)

Intervention
DOAC, typically with a higher loading phase
Dose
Regimen-specific (e.g. apixaban 10 mg twice daily for 7 days, then 5 mg twice daily)
Duration
At least 3 months; longer or indefinite if the clot was unprovoked
Outcome
Recurrent VTE
Notes
Duration depends on whether the clot had a reversible cause. A clinical decision.

Adults with a MECHANICAL HEART VALVE

Intervention
WARFARIN — DOACs are contraindicated
Dose
Titrated to a target INR (range depends on valve type and position)
Duration
Lifelong
Outcome
Valve thrombosis and thromboembolism
Notes
DOACs are HARMFUL in this group. The dabigatran trial was stopped early for excess clotting AND bleeding.

Adults with antiphospholipid syndrome (particularly triple-positive)

Intervention
WARFARIN
Dose
Titrated to target INR
Duration
Long-term
Outcome
Recurrent thrombosis
Notes
DOACs performed WORSE than warfarin in trials in this population. Another warfarin-only indication.

Where scientists agree — and don’t

Agreed

  • Anticoagulation reduces stroke in atrial fibrillation by roughly two-thirds.
  • DOACs cause substantially less intracranial haemorrhage than warfarin.
  • Mechanical heart valves require warfarin; DOACs are contraindicated and harmful.
  • Aspirin is not adequate stroke prevention in atrial fibrillation.
  • Anticoagulation is effective for treating and preventing venous thromboembolism.

Debated

  • Optimal anticoagulation in advanced chronic kidney disease and dialysis.
  • How aggressively to anticoagulate the very elderly with high falls risk — where the fear of bleeding routinely wins over the far larger stroke risk.
  • Duration of anticoagulation after an unprovoked venous thromboembolism.
  • Management at the extremes of body weight, where DOAC dosing evidence is thinner.

Unknown

  • Whether factor XI inhibitors can deliver the same protection with meaningfully less bleeding.
  • Optimal management of anticoagulation in active cancer, where both clotting and bleeding risks are elevated.
  • The true magnitude of risk from concurrent herbal supplement use, which is poorly captured in trials.

What remains unknown

  • Can factor XI inhibitors separate the prevention of pathological clotting from the impairment of normal haemostasis?
  • What is the right anticoagulation strategy in advanced kidney disease, where trial evidence is thin and both risks are high?
  • How much of the under-treatment of elderly AF patients is driven by an overestimate of falls-related bleeding risk?
  • How commonly do over-the-counter supplements meaningfully alter DOAC levels in real-world use?

Questions people actually ask

What is the difference between a 'blood thinner', an anticoagulant, and an antiplatelet?

'Blood thinner' is a lay term that unhelpfully lumps together two entirely different drug classes, and the confusion causes real harm. ANTICOAGULANTS (warfarin, apixaban, rivaroxaban, edoxaban, dabigatran) interfere with the clotting cascade and are what you need to prevent stroke in atrial fibrillation or to treat a DVT. ANTIPLATELETS (aspirin, clopidogrel) stop platelets sticking together and are used mainly after a heart attack or stent. They are not interchangeable. Aspirin does not do the job of an anticoagulant, and a great many people with atrial fibrillation believe otherwise. Also: nothing here actually 'thins' the blood. Its viscosity is unchanged; what changes is how readily it clots.

I have atrial fibrillation and my doctor only put me on aspirin. Is that enough?

Almost certainly not, and this is worth raising directly. Aspirin is markedly less effective than anticoagulation at preventing AF-related stroke, and it still causes bleeding — so it delivers a fraction of the benefit at much of the risk. It was widely used for years as a 'gentler' option for older or frailer patients, and guidelines have since abandoned it for this purpose. If you have atrial fibrillation and are taking only aspirin for stroke prevention, ask your clinician specifically whether an anticoagulant is indicated. Do not stop anything on your own.

I have a mechanical heart valve. Can I switch to one of the newer tablets?

No — and this is the single most important thing on this page. DOACs are CONTRAINDICATED with mechanical heart valves. This is not a matter of them being slightly less effective: the trial of dabigatran in mechanical valve patients was STOPPED EARLY because those on the DOAC had both more clots and more bleeding. Mechanical valves require warfarin, with INR monitoring, for life. The DOACs are more convenient, require no blood tests, and sound more modern — and in this specific group they can kill you. If any clinician proposes switching you, that warrants a second opinion from cardiology.

Do I have to avoid green vegetables on warfarin?

No, and this myth has done real damage. Warfarin works by blocking vitamin K, so vitamin K in food affects it — but the answer is CONSISTENCY, not avoidance. Your warfarin dose is calibrated to your usual diet. If you habitually eat leafy greens, keep doing so at a steady rate and your INR will be stable. What destabilises it is large swings: a sudden health kick, or abruptly cutting greens out. People have avoided vegetables for years, to the detriment of their health, believing they had to. You do not have to. Just be consistent, and tell your anticoagulation clinic if your diet changes substantially.

Can I take supplements while on an anticoagulant?

Ask first, and be specific — because one of them is genuinely dangerous. ST JOHN'S WORT induces the enzymes and transporters that clear DOACs from the body, lowering the drug level in your blood. The consequence is that you may be UNPROTECTED FROM STROKE while believing you are covered, and there is no symptom to warn you. Fish oil, vitamin E, garlic, ginkgo and high-dose curcumin all have antiplatelet effects that may add to bleeding risk. Grapefruit juice affects some DOACs. None of this means every supplement is dangerous — it means your clinician and pharmacist need the full list, including anything bought in a health-food shop, which people routinely fail to mention because it feels like food rather than medicine.

What should I do if I have a bleed, or a bad fall?

Seek medical attention, and tell them immediately that you are anticoagulated — this is the single most useful thing you can say in that room. A head injury while anticoagulated needs assessment even if you feel fine, because bleeding inside the skull can develop over hours. Reversal agents exist and work: idarucizumab reverses dabigatran, andexanet alfa reverses the factor Xa inhibitors, and vitamin K with prothrombin complex concentrate reverses warfarin. Carry an anticoagulant alert card. And do not stop your medication after a minor bleed without advice — the stroke risk does not pause while you decide.

I'm elderly and I fall sometimes. Isn't anticoagulation too risky for me?

This is a real question and it is very often answered wrongly. Falls risk is one of the commonest reasons anticoagulation is withheld from elderly patients with atrial fibrillation — and analyses suggest a person would need to fall an implausible number of times a year for the bleeding risk to outweigh the stroke protection. The uncomfortable truth is that the people denied anticoagulation because they might fall are frequently the same people at the highest risk of a devastating, disabling stroke. It is a conversation to have properly with a clinician, weighing your actual risks — not a reason to assume the answer is no.

Practical takeaways

  • In atrial fibrillation, anticoagulation cuts stroke risk by about two-thirds — one of the biggest treatment effects in medicine.
  • ASPIRIN IS NOT AN ANTICOAGULANT and is not adequate stroke prevention in AF. If you have AF and take only aspirin, ask why.
  • MECHANICAL HEART VALVE = WARFARIN ONLY. DOACs are harmful in this group and the trial was stopped early for deaths and clots.
  • On warfarin, the rule is CONSISTENT vitamin K intake, not avoidance. Eat your greens — just eat a steady amount of them.
  • Tell your clinician about every supplement. St John's Wort can lower DOAC levels and leave you unprotected from a stroke without any warning sign.

How it works

The mechanism comes last on purpose. A compelling explanation of how something might work is the easiest part of the story to tell, and the part most likely to outlive the evidence for it.

Precisely characterised. DOACs directly inhibit factor Xa (apixaban, rivaroxaban, edoxaban) or thrombin (dabigatran); warfarin blocks vitamin K–dependent synthesis of factors II, VII, IX and X.

How we found out

  1. Warfarin

    Developed from a compound that caused fatal haemorrhage in cattle eating spoiled sweet clover, warfarin became the anticoagulant of record for over half a century — highly effective, and demanding, with a narrow therapeutic window, constant INR monitoring, and a long list of dietary and drug interactions.

  2. The Aspirin Era, and Its End

    For years aspirin was offered as a gentler alternative for AF patients thought too frail for warfarin. Evidence eventually showed it delivers far less stroke protection while still causing bleeding, and guidelines abandoned it for this indication. Many patients were never told.

  3. The DOACs

    Large randomized trials established that direct oral anticoagulants match or beat warfarin for stroke prevention, with substantially less bleeding into the brain and no routine monitoring. They rapidly became first-line for most indications.

  4. The Limits of the DOACs

    The dabigatran mechanical-valve trial was stopped early for excess thrombosis AND bleeding, and subsequent trials showed DOACs performing worse than warfarin in antiphospholipid syndrome. The 'newer is better' story turned out to have hard, specific exceptions — and they are exceptions in which people die.

  5. Factor XI

    Current research targets factor XI, on the hypothesis that pathological clotting can be prevented while the normal clotting that stops bleeding is preserved — potentially the most significant advance in this field in decades.

References

Verified sources. BioSignal does not print a citation it has not checked.

References for this record are being verified and will be published with the next review. BioSignal does not print citations it has not checked.

Version history

  • Version 1.0

    Initial record. Calibration: Established maturity, HIGH confidence — a deliberate example of BioSignal's anti-hype principle running in the positive direction. The evidence here is exceptionally strong and the record says so plainly. Safety graded CAUTION not because the drugs are dubious but because bleeding is the direct consequence of the mechanism and requires clinical supervision. Three facts are carried as the record's core purpose: mechanical valves require warfarin (DOACs are harmful), aspirin is not adequate AF stroke prevention, and St John's Wort can silently disable a DOAC. Review cadence: Annually, and sooner if factor XI inhibitors reach practice.