Bayer Logo

NICE AF RTD - digital

Information for UK healthcare professionals only.

This round-table discussion supplement has been commissioned and funded by Bayer plc and developed in partnership with Guidelines in PracticePlease see bottom of page for full disclaimer.

View Xarelto® (rivaroxaban) adverse event reporting and prescribing information

Download the supplement

Discussion group:

Chair: Professor John Camm (JC), St George’s, University of London, London

Dr Joseph Kwan (JK), Consultant in Stroke Medicine, Imperial College, London

Dr Jim Moore (JM), President, Primary Care Cardiovascular Society

Dr Raj Mattu (RM), Consultant Cardiologist, Kettering General Hospital NHS Foundation Trust

Xarelto®▼ (rivaroxaban) is indicated for the prevention of stroke and systemic embolism in adult patients with non-valvular atrial fibrillation (NVAF), with one or more risk factors, such as congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischaemic attack (TIA).1


The updated NICE guideline on atrial fibrillation (AF)—NICE guidance (NG) 196—was published in April 2021, replacing clinical guideline (CG) 180.2,3 All four of the direct oral anticoagulants (DOACs) licensed for the prevention of stroke and systemic embolism in adult patients with NVAF with risk factors (apixaban, dabigatran, edoxaban, and rivaroxaban) are recommended as treatment options (Box 1),1,2,4–6 in line with national and international guidance.7,8 NG196 does not recommend one DOAC over the others, but instead emphasises that treatment should be tailored to the person’s clinical needs and preferences.2 Each anticoagulant has different risks and benefits that should be considered and fully discussed with the person as part of informed shared decision making.2 The guidance also includes discussing the option of transitioning to a DOAC with eligible patients established on warfarin, and emphasises the importance of shared informed decision making.2

A round-table discussion was held to explore decision making around the choice of DOAC treatment, practicalities of transitioning eligible patients from warfarin to a DOAC, and the importance of shared decision making between patient and physician.

Box 1: NG196 recommendations2,[A] discussed

1.6.3. Offer anticoagulation with a direct-acting oral anticoagulant to people with atrial fibrillation and a CHA2 DS2 -VASc score of 2 or above, taking into account the risk of bleeding. Apixaban, dabigatran, edoxaban and rivaroxaban are all recommended as options, when used in line with the criteria specified in the relevant NICE technology appraisal guidance.

1.6.6. For adults with atrial fibrillation who are already taking a vitamin K antagonist and are stable, continue with their current medication and discuss the option of switching treatment at their next routine appointment, taking into account the person’s time in therapeutic range.

[A] Apixaban, dabigatran, edoxaban, and rivaroxaban are licensed for the prevention of stroke and systemic embolism in adult patients with NVAF, with one or more risk factors, such as prior stroke or TIA; age ≥75 years; heart failure (NYHA Class ≥II); diabetes mellitus; hypertension. Please refer to the relevant summary of product characteristics before prescription.1,4–6

What factors do you consider when choosing a DOAC for your eligible patients?

JM: When choosing from the four licensed DOACs for stroke prevention in NVAF,1,4–6 my personal view is that the two key attributes from a primary care perspective are that the DOAC is as easy as possible for the patient to take and as simple as possible for the physician to prescribe. Ease of use for the patient improves adherence, so I usually recommend a DOAC with once‑daily dosing. Simplicity for the prescriber means selecting a DOAC with an uncomplicated dosing regime and limited relevant drug interactions.

Cost considerations related to individual DOACs are not typically a major consideration for primary care physicians, as GPs are independent prescribers free to choose the drug they deem most appropriate. However, in some areas, a particular drug may be recommended by clinical commissioning groups because of various factors, which can include agreed local price discounts.

It is not always fully appreciated that patients taking DOACs need ongoing monitoring and at least annual review (including assessment of renal function and other relevant factors) to ensure the correct dosage is being prescribed.2 This also provides an opportunity to review NVAF in general and factors that may increase risk of bleeding.

RM: The patient is also my starting point for choosing an anticoagulant. Most prefer a DOAC once the options are explained but some still opt for warfarin. The choice of DOAC is based on ease of monitoring and maintenance for primary care, and ease of administration for the patient. Swallowing difficulties are a consideration in some elderly patients; apixaban, edoxaban, and rivaroxaban tablets can be crushed to be administered orally or through a feeding tube.1,4,6

Patients usually take lifelong anticoagulants in the absence of symptomatic benefits and become aware of bleeding risks following an informed discussion, so adherence can be problematic.9–11 The importance of adherence can be reinforced at annual review, alongside checking bleeding and stroke risks.

While not proposed by NICE,2 I encourage GPs seeking advice to focus on one or two DOACs—a once-daily option and a twice‑daily option—so they can gain deep knowledge and familiarity with two products that they will use in most circumstances. Dose adjustments may be needed if patients develop intercurrent illness that reduces their creatinine clearance (CrCl).[B],1,4

For most people, the benefit of anticoagulation outweighs the bleeding risk.2  In patients with swallowing problems that require medications to be crushed, I would be inclined to avoid dabigatran (as the capsules cannot be opened).5 Persistent minor bleeding, such as epistaxis and a small GI bleed, or persistent decreases in haemoglobin during use of a DOAC may signal an underlying condition such as GI cancer12,13 and should be vigorously investigated, including colonoscopy. I avoid DOACs in patients with significant liver disease requiring cryotherapy, bleeding issues or coagulopathy.[C] In patients with renal impairment, DOACs have the option of dose reductions, if required.[B],1,4–6 The licensed dose should always be used unless there are special precautions or contraindications.

[B] Apixaban and rivaroxaban are contraindicated in patients with CrCl <15 ml/min; the use of edoxaban in these patients is not recommended.1,4,6  Dabigatran is contraindicated in patients with CrCl <30 ml/min.5 Caution is advised when used in patients with an increased risk of haemorrhage. Treatment should be discontinued if severe haemorrhage occurs.1,4
[C] Apixaban, edoxaban, and rivaroxaban are contraindicated in patients with hepatic disease associated with coagulopathy and clinically relevant bleeding risk.1,4,6 Dabigatran is contraindicated in patients with liver disease expected to have any impact on survival.5 Please refer to the relevant summary of product characteristics for full contraindications.

JK: Most of my patients have already had a stroke or TIA, so I use DOACs for secondary rather than primary prevention. My preference on choice of DOAC echoes that in primary care and cardiology—ease of use and ease of prescribing being common themes.

Most of my patients tend to be elderly; drug interactions and adherence are common problems due to polypharmacy. Patients admitted with a stroke often report missing doses of anticoagulants or have been prescribed (or have self-managed) lower than recommended doses due to concern around risk factors such as bleeding.

Cognitive problems commonly develop after stroke (approximately 80% (total population n=409) of patients examined 3 months after ischaemic stroke showed impairment in at least one cognitive domain in the Helsinki Stroke Aging Memory Study),14 which may affect adherence. Many patients also have limited use of their hands after stroke, which can make it difficult to extract tablets from blister packs. They often require a dosette box prefilled by the community pharmacy, which precludes the use of dabigatran (as it must be stored in the original package to protect from moisture).5 Some post-stroke patients are discharged with a percutaneous endoscopic gastrostomy (PEG) tube, which precludes use of dabigatran (as the capsules must not be opened).5

Many patients require support from social services, and a second visit for an evening dose is unrealistic. It is optimal to rationalise all medicines to once daily (with the morning meal) and most antihypertensives, statins (including atorvastatin), and even diabetes medicines have been converted to once-daily regimens.

Even in patients with haemorrhagic stroke, the availability of a reversal agent is less of a concern than when DOACs first became available. However, questions around anticoagulation after stroke remain unanswered. Evidence around restarting DOACs after ischaemic stroke (particularly the timing) is lacking, with four trials ongoing.15–18 No data are available on whether patients after haemorrhagic stroke should be on the same dose of DOAC or transitioned to a lower dose or warfarin.

At our hyperacute stroke unit, factor Xa or Xa inhibitor levels are not routinely used to check for compliance in patients with ischaemic or haemorrhagic stroke despite an Xa inhibitor for AF‑related stroke prevention, as they are informative only about recent rather than long-term use, and our stroke unit pharmacists discuss adherence with the patient and carer at home. If the patient is definitely adherent, restarting depends on the severity of the stroke. Once-daily dosing is likely to help adherence, because patients will likely have cognitive impairment and are discharged on four or five new drugs; swallowing difficulties may also be an issue.

RM: Two retrospective cohort-based studies provide useful guidance for adherence. The first, from the US, used an insurance reimbursements database to examine the impact of adherence to DOACs on ischaemic stroke ris.19 It showed that adherence to DOACs was associated with a reduction in ischaemic stroke risk compared to non-adherence.19  A retrospective cohort study using the French national healthcare data from 76,795 patients with atrial fibrillation newly treated with DOAC, showed that non-adherence to DOAC treatment was associated with an increased risk of stroke and death, compared to those who were adherent.20

What are your thoughts on the NICE recommendation to discuss transitioning eligible patients already stable on a vitamin K antagonist to a DOAC at their next appointment?

RM: The manner of the NICE recommendation indicates that NICE intends to encourage physicians to convert eligible patients from warfarin but without a direct instruction to do so.2 I have concerns about the medico-legal implications if a patient had a stroke during the transition.

JK: I am similarly concerned. My main issue is that we do not have any means to calculate time in therapeutic range (TTR) during ward rounds in the hospital—junior doctors do not have time, and an app that I used is no longer available. We have to contact the anticoagulation clinic for TTR but often do not receive a timely answer.

Would you discuss transitioning all eligible patients to a DOAC at their next appointment?

JM: In my view, people taking warfarin for stroke prevention related to non‑valvular AF whose TTR is not adequate (NICE recommends >65%2 but my personal view is that we should aim for much higher thresholds) should be prioritised; eligible patients who are well controlled should be reviewed at an appropriate time for transitioning to a DOAC. It is important to remember that DOACs are not suitable for some people, including those with prosthetic mechanical heart valves and moderate to severe rheumatic mitral (MARM) valvular AF, those with antiphospholipid antibody syndrome, as well as women who are pregnant or breastfeeding, patients with a higher target INR range for various reasons, and those with severe renal impairment (CrCl <15 ml/min).1,4–6

RM: I would be most inclined to discuss the change with patients with poor TTR or labile INRs. Head-to-head data indicate that DOACs, overall, offer benefit over warfarin21–25  and this should be discussed in detail with patients, emphasising that it may be beneficial to transition to DOACs. I would avoid transitioning to DOACs in patients with good control who are very obese and those taking CYP3A4 drugs that could interact.1,4–6

JK: I transition eligible patients who have a stroke while on warfarin to a DOAC because the risk of haemorrhage, especially intracranial haemorrhage, is lower.1 This is particularly important after ischaemic stroke, as the risk of haemorrhagic transformation may be high because the infarcted brain is more vulnerable. However, many patients find the monthly contact with their anticoagulation team reassuring and so may be reluctant to transition to a DOAC and lose that interaction. The relative costs of prescribing a DOAC versus running a warfarin clinic is also a consideration.

What is your experience of transitioning existing eligible patients on warfarin to a DOAC?

JM: Following the NHS England and NHS Improvement (NHSE&I) and Royal Pharmaceutical Society (RPS) guidance around anticoagulation during the COVID-19 pandemic,26,27 almost every eligible patient we approached chose to change. The very few who opted to remain on warfarin tended to be elderly patients who appreciated regular contact with the practice.

We have access to point-of-care testing for INR and used this to identify patients suitable for transitioning from warfarin to DOAC. We arranged a preliminary telephone discussion with those suitable for transitioning and those who agreed were invited to the practice for an extended appointment to discuss the risks and benefits and to have their TTR checked, in addition to any relevant blood tests, such as renal function. In my view, most eligible patients can be transitioned from warfarin to DOACs in primary care based on current INR and additional information, such as up-to-date weight and recent assessment of renal function. New medicines services offered by community pharmacy are incredibly useful where the pharmacist calls patients within a few weeks of starting a medicine to reiterate key messages, identify any side effects, and check and reinforce the importance of adherence.

RM: I discuss the options so patients can make an informed decision. Given this is an elective process rather than a clinical decision and there is no need for urgency, we should be particularly cautious. I recommend basing the decision on test results no older than 1 week to minimise risk with the transition, as INR and even kidney function that were normal 3 months ago may be higher or lower due to dehydration as a result of intercurrent illness, diarrhoea and vomiting, or hot weather. It is also important to take into account any comorbidities.

JK: Secondary care tends to convert most eligible patients to a DOAC for secondary prevention, unless there are contraindications. The degree of urgency will depend on the INR history: if INR control has been poor in the previous few weeks, transitioning should be immediate, whereas the decision to change can be more measured in those with stable INR.28

Renal function can decline very rapidly in older patients (for example, if they develop a urinary tract infection or become dehydrated). Once patients are transitioned to a DOAC, renal function should be monitored annually for CrCl >60 ml/min, every 6 months for CrCl 30–60 ml/min and every 3 months for CrCl 15–30 ml/min.29

Does choice of anticoagulant differ when considering patients transitioning from warfarin compared to newly diagnosed, anticoagulant-naïve patients?

JM: In primary care, choice of anticoagulant when transitioning patients would be based on the same factors as for anticoagulant-naïve patients. However, there are additional considerations for starting a DOAC after stopping warfarin, as outlined in the SPC for each drug (Table 1)1,4–6 or, if preferred, the more generic guidance published by the European Heart Rhythm Association (EHRA).29 Patients must also take the correct number of doses if transitioning from once-daily warfarin to a twice-daily DOAC.

Table 1: Guidance when transitioning from warfarin to a DOAC1,4–6  


When converting patients from VKA therapy to apixaban, warfarin or other VKA therapy should be discontinued and apixaban started when INR is <2.04


The VKA should be stopped; dabigatran can be given as soon as INR is <2.05


Discontinue the VKA and start edoxaban when INR is ≤2.56  


VKA treatment should be stopped and rivaroxaban should be initiated when INR is ≤3.0 (for prevention of stroke and SE)1  

INR=international normalised ratio; VKA=vitamin K antagonist; SE=systemic embolism.

JK: As warfarin is taken once daily, some patients transitioning to a DOAC may opt for a once-daily regimen.

RM: I would use this as an opportunity to review the patient’s profile (as if they were a new patient), but the fact they were already taking warfarin would not change my choice of DOAC.

How important do you feel it is to involve the patient in decision making around treatment options?

RM: Shared decision making is key. Patients are increasingly well informed. As an important barrier to protecting patients with NVAF is adherence, I do everything I can to maximise this. I engage the patient from the start so they feel ownership over their treatment. I provide lay descriptions of the CHA₂DS₂-VASc, HAS-BLED, and ORBIT scores, and describe the pros and cons of the treatment options, emphasising they may be taken once or twice daily. Some patients opt for twice daily as they feel they will not lose all of their protection if they miss a dose for any reason—this may include elderly patients with cognitive impairment who may forget to take single doses and young busy patients who may not always have reliable routines or return home before the evening dose. Obviously there is a risk of reduced protection from stroke if the full dose is not taken correctly.

JK: Patients and families are often scared after stroke. I explain that those with mild strokes are fortunate to be able to swallow, walk, and dress themselves and that this medicine will protect them from future, potentially more serious, events. I emphasise modifiable risk factors, which I highlight using the HAS-BLED score, particularly the risk of excessive alcohol consumption.

JM: Patients should be central to any discussions around treatment and ease of use of any therapy is central to that dialogue. A fully informed discussion around the benefits and risks of anticoagulation requires an extended appointment, and how information is shared with the patient can influence their decision making. It is vital to share the relevant information honestly and openly and in a way the patient will understand. I do not share the actual stroke or bleeding risk scores with patients, as I think this can be confusing. I explain their risk of stroke over a year and how this is significantly reduced with anticoagulation. I use the tabulated risk factors in the bleeding risk score as an aide-memoire for areas to consider that can increase their risk of bleeding and that I or the patient may be able to change, thereby reducing risk. Importantly, recently published NICE guidance now recommends that DOACs be considered as the first-line therapy for stroke prevention in people with NVAF,1 therefore I no longer routinely discuss the option of using VKAs (warfarin) for stroke prevention in NVAF unless there are specific reasons why a DOAC should not be considered.


Jemma Lough, independent medical writer, helped draft this document.

Conflicts of interest

The group members received an honorarium from Bayer plc for this round-table discussion. John Camm has received personal fees and institutional grants from Bayer, Boehringer Ingelheim, and Daiichi Sankyo, and personal fees from Bayer, Daiichi Sankyo, Boehringer Ingelheim, Pfizer, Boston Scientific, Medtronic, and Abbott.


  1. Bayer plc. Xarelto (rivaroxaban) 20mg film-coated tablets—summary of product (accessed August 2021).
  2. NICE. Atrial fibrillation: diagnosis and management. NICE Guideline 196. NICE, 2021. Available at:
  3. NICE. Atrial fibrillation: management. Clinical Guideline 180. NICE, 2014. Available at:
  4. Bristol-Myers Squibb-Pfizer. Eliquis (apixaban) 2.5 mg film-coated tablets—summary of product (accessed August 2021).
  5. Boehringer Ingelheim Limited. Pradaxa (dabigatran) 150 mg hard capsules—summary of product (accessed August 2021).
  6. Daiichi Sankyo UK Limited. Lixiana (edoxaban) 60mg film-coated tablets—summary of product (accessed August 2021).
  7. Hindricks G et al. Eur Heart J 2020; 42: 373–498.
  8. January C et al. J Am Coll Cardiol 2019; 74: 104–132.
  9. Shore S et al. Am Heart J 2014; 167: 810–817.
  10. Beyer-Westendorf J et al. EP Europace 2015; 4: 530–538.
  11. Nelson W et al. Curr Med Res Opin 2014; 30: 2461–2469.
  12. Yuqing D et al. J Am Heart Assoc 2019; 8: e012540.
  13. Hendrie P, Garcia D. J Natl Compr Canc Netw 2013; 11: 1446–1449.
  14. Jokinen H et al. Eur J Neurol 2015; 22: 1288–1294.
  15. NIH. NCT03148457. 2021. Available from: (accessed September 2021)
  16. NIH. NCT03021928. 2020. Available from: (accessed September 2021)
  17. NIH. NCT02961348. 2021. Available from: (accessed September 2021)
  18. Åsberg S et al. Trials 2017; 18: 581.
  19. Deshpande C et al. Curr Med Res Opin 2018; 34: 1285–1292.
  20. Gabet A et al. Clin Epidemiol 2021; 13: 131–140.
  21. Patel M et al. N Engl J Med 2011; 365: 883–891.
  22. Granger C et al. N Engl J Med 2011; 365: 981–992.
  23. Hijazi Z et al. Circulation 2014; 129: 961–970.
  24. Giugliano R et al. N Engl J Med 2013; 369: 2093–2104.
  25. Ruff C et al. Lancet 2014; 383: 955–962.
  26. NHS England and NHS Improvement. Clinical guide for the management of anticoagulant services during the coronavirus pandemic. November 2020 (updated February 2021). Available at:
  27. Royal Pharmaceutical Society. Guidance for the safe switching of warfarin to direct oral anticoagulants (DOACs) for patients with non-valvular AF and venous thromboembolism (DVT / PE) during the coronavirus pandemic. London: RPS, 2020. Available at:
  28. Seiffge D et al. Ann Neurol 2020; 87: 677–687.
  29. Steffel J et al. EP Europace 2021; doi:10.1093/europace/euab065.

This round-table discussion supplement has been commissioned and funded by Bayer plc and developed in partnership with Guidelines in Practice. Bayer plc suggested the topic and participants, and carried out full medical approval on all materials to ensure compliance with the ABPI code of practice. The sponsorship fee included honoraria for the participants. The views and opinions of the participants are not necessarily those of Guidelines in Practice, its publisher, advisers, or advertisers. No part of this publication may be reproduced in any form without the permission of the publisher.


Date of preparation: November 2021