North of Tyne, Gateshead and North Cumbria APC Guidelines for prescribing in primary care: Non-valvular Atrial Fibrillation

Version 4 (Approved October 2019, Review October 2022) Page 1

Guidelines for prescribing in primary care:

Non-valvular atrial fibrillation

Review date: October 2022

This guideline has been prepared and approved for used within Newcastle Gateshead CCG,

North Tyneside CCG and Northumberland CCG in consultation with Secondary Care Trusts.

Author Guideline consultation group

Gary Armstrong – CBC Health Ltd.

(on behalf of NGCCG)

Dr Paula Batsford (GP – Northumberland CCG)

Dr John Bourke (Consultant Cardiologist – NuTH)

Dr Chris Jewitt (GP – Newcastle Gateshead CCG)

Dr Stephen Kirk (Clinical Director – Delivery – Newcastle

Gateshead CCG / GP – Newcastle Gateshead CCG)

Dr James Lunn (GP – North Tyneside CCG)

Dr Christopher Scott (Consultant Cardiologist – QEH

Gateshead)

Approved by:

Committee Date

This guideline is not exhaustive and does not override the individual responsibility of health

professionals to make decisions appropriate to the circumstances of the individual patient, in

consultation with the patient and/or guardian or carer.

Full details of contra-indications and cautions for individual drugs are available in the BNF or in the Summary of

Product Characteristics (available in the Electronic Medicines Compendium) www.emc.medicines.org.uk

1 Renal Disease - Dialysis, transplant, Cr >2.6 mg/dL or >200 µmol/L

Liver Disease - Cirrhosis, Bilirubin >2x Normal, AST/ALT/AP >3x Normal 2 Unstable/high INRs, Time in Therapeutic Range < 60%

Prescribing in Non-Valvular AF in Primary Care

Estimates suggest the prevalence of atrial fibrillation is increasing and left untreated atrial

fibrillation is a significant risk factor for stroke and other morbidities.

AF – Stroke risk assessment: NICE CG180 – Atrial Fibrillation: the management of atrial fibrillation was published in June

2014 and made some significant changes to the diagnosis and treatment of AF.

The overall stroke risk in patients with AF is around 5%. This, however, can vary

substantially between patients ranging from a 15-year risk of around 1.3% in younger

patients with lone AF up to an annual figure of almost 20% or higher if the individual suffers

from valve disease.

In patients with non-valvular atrial fibrillation NICE now recommends risk is assessed using

the CHA2DS2-Vasc and HAS-BLED tools. CHA2DS2-Vasc is not intended for use in patients

with significant valve disease. The HAS–BLED score is a guide to bleeding risk. Clinicians may

want to consider consulting secondary care for advice in those with higher risk scores.

CHA2DS2-Vasc

Risk Factor Score

Congestive heart failure / LVSD 1

Hypertension 1

Age ≥ 75 2

Diabetes Mellitus 1

Stroke / TIA / Thromboembolism 2

Vascular Disease 1

Age 65-74 1

Sex category – Female 1

Maximum Score 9

HAS-BLED

Risk Factor Score

Hypertension (Uncontrolled) 1

Abnormal Renal / Hepatic Function1

- (1 each)

1 or 2

Stroke 1

Bleeding 1

Labile INRs2

1

Elderly (>65yrs) 1

Drugs (e.g. NSAIDs) and/or

Alcohol (≥8 drinks per week) - (1

each)

1 or 2

Maximum Score 9

Stroke risk:

The CHA2DS2-Vasc score is used to estimate stroke risk in patients with non-valvular AF. These figures below

are the approximate number per 1000 patients each year whom have AF and who are predicted to still get a

stroke. (NICE QS93 – Atrial Fibrillation Quality Statement 1: Adults with non-valvular atrial fibrillation and a

CHA2DS2-Vasc stroke risk score of 2 or above are offered anticoagulation)

Score No

medication Warfarin DOAC

CHA2DS2-Vasc =0

CHA2DS2-Vasc =1 13 5 4

CHA2DS2-Vasc =2 22 8 6

CHA2DS2-Vasc =3 32 12 9

CHA2DS2-Vasc =4 40 14 10

CHA2DS2-Vasc =5 67 24 18

CHA2DS2-Vasc =6 98 35 25

HAS-BLED Score:

These figures are the approximate number per 1000 patients each year who are predicted to have major

bleeds (GI or intra-cranial) whilst on anticoagulation. A higher HAS-BLED score should not automatically

exclude a patient from using an anticoagulant and the risk-benefit balance should be carefully considered for

all patients.

Score Per 1000 pt per year

0 10

1 20

2 30

3 40

4 88

Anticoagulation prescribing in Non-valvular AF

Antiplatelets Do not offer antiplatelets as sole treatment for the prevention of stroke in people with atrial fibrillation.

Where anticoagulation is not indicated antiplatelets should be reviewed and stopped where there is no

other indication which would warrant anti-platelet use. (NICE QS93 – Atrial Fibrillation Quality Statement 2:

Adults with atrial fibrillation are not prescribed aspirin as monotherapy for stroke prevention)

In cases where an individual has a stent or is post ACS and would normally be treated with dual antiplatelet

therapy please discuss on-going treatment of these patients on an individual basis with their consultant

cardiologist.

Choice of anticoagulant There are currently five available anticoagulants, warfarin, a vitamin K antagonist and four direct oral

anticoagulants (DOACs) apixaban, dabigatran, edoxaban and rivaroxaban. Direct oral anticoagulants has

been adopted as the preferred terminology for the group of drugs previously known as novel oral

anticoagulants (NOACs). The direct oral anticoagulants should be targeted to patients who are likely to

derive greatest benefit.

Key groups in whom DOACs should especially be considered includes:

Those who cannot take vitamin K antagonists or have declined to take warfarin

Those who cannot be stabilised on vitamin K antagonists with poor time in therapeutic range (e.g.

less than 65% despite adequate adherence). TTR should be calculated over a maintenance time of at

least 6 months, excluding measurements taken within the first 6 weeks of treatment, or

measurements taken where values out of range may be attributed to a known cause e.g. acute

antibacterial treatment. (NICE QS93 – Atrial Fibrillation Quality Statement 4: Adults with atrial

fibrillation taking a vitamin K antagonist who have poor anticoagulation control have their

anticoagulation reassessed)

Those taking anti-platelets for stroke prevention where an assessment has been made and warfarin

may not be suitable due to reasons that would not specifically exclude them from using

anticoagulation

The Direct Oral Anticoagulants should only be used for non-valvular AF within product license and in line

with the relevant NICE technology appraisals:

NICE TA275 - Apixaban for preventing stroke and systemic embolism in people with nonvalvular

atrial fibrillation

NICE TA249 - Dabigatran etexilate for the prevention of stroke and systemic embolism in atrial

fibrillation

NICE TA355 - Edoxaban for preventing stroke and systemic embolism in people with non-valvular

atrial fibrillation

NICE TA256 - Rivaroxaban for the prevention of stroke and systemic embolism in people with atrial

fibrillation

The decision regarding which treatment is to be used should be made after an informed discussion between

the clinician and the patient about the risks and benefits of each of the treatments compared with each

other and against no treatment at all. (NICE QS93 – Atrial Fibrillation Quality Statement 3: Adults with atrial

fibrillation who are prescribed anticoagulation discuss the option with their healthcare professional at least

once a year)

Use of serum creatinine and eGFR estimated with the eGFRcreat CKD-EPI equation may result in a

misrepresentation of renal function. This is of particular relevance in elderly patients with low body weight

(and conversely younger patients with high muscle mass). Therefore, any patients considered for a newer

anticoagulant with reduced renal function should have their creatinine clearance more accurately estimated

using the Cockcroft-Gault equation:

𝐶𝑟𝐶𝑙 = (140 − 𝑎𝑔𝑒) 𝑥 𝑤𝑒𝑖𝑔ℎ𝑡 (𝑘𝑔) 𝑥 1.23 𝑓𝑜𝑟 𝑚𝑒𝑛 𝒐𝒓 1.04 𝑓𝑜𝑟 𝑤𝑜𝑚𝑒𝑛

𝑆𝑒𝑟𝑢𝑚 𝑐𝑟𝑒𝑎𝑡𝑖𝑛𝑖𝑛𝑒 (𝜇𝑚𝑜𝑙 𝐿⁄ )

Dosages in renal impairment

Creatinine

Clearance

Apixaban Dabigatran Edoxaban Rivaroxaban

≥50ml/min 5mg BD (or 2.5mg BD if 2 or more of the following are

present - >80y, <60kg or serum cr >133mmol/L)

150mg BD 60mg OD 20mg OD

30-49ml/min 110mg BD 30mg OD 15mg OD

15-29ml/min 2.5mg BD Avoid 30mg OD 15mg OD (with

caution)

<15ml/min Avoid

In order for the Cockcroft Gault equation to be accurate we require a recent weight. If these have not been

updated recently or the patient’s weight has changed significantly recently these will need to be re-

measured. In some cases obtaining an up to date weight may be difficult. Below is the preferred order in

which weights may be obtained:

1. Measured in practice using calibrated scales

2. Measured at home visit with calibrated scales

3. Measured at home visit with uncalibrated scales

4. Patient reported weight

The clinical systems have approved creatinine clearance calculators. See Appendix 1 for details. For patients

with an amputation the EMIS and SystmOne calculators may not be accurate and creatinine clearance

should be calculated using a different method. One such calculator is available at

https://clincalc.com/kinetics/crcl.aspx alternatively seek specialist advice on appropriate dosing.

Individuals stable on warfarin should not routinely be considered for changing to a newer oral anticoagulant.

Care should be taken to ensure a safe transition between preparations and advice sought from the patient’s

anticoagulation service where appropriate. Further information is also available on the SPC for each of the

newer anticoagulants.

For patients using antiplatelets it is not necessary to have a break between stopping the antiplatelet and

starting the DOAC and it is safe to start the new medication the day following the last antiplatelet dose.

DOACs and Compliance aids / swallowing difficulties Apixaban and rivaroxaban have been assessed and whilst no stability data is available, the manufacturer

does not, or cannot recommend use in CAs but there are no theoretical concerns with the product and they

may be used in compliance aids. Dabigatran is not suitable for use in a compliance aid. The stability of

edoxaban in compliance aids has not been assessed and use in this way should be avoided.

Both apixaban and rivaroxaban can be crushed and administered via feeding tube as per the SPC.

DOACs and Food To ensure appropriate bioavailability rivaroxaban must be given with food. The bioavailability of apixaban,

dabigatran and edoxaban are not affected by food.

DOAC Antidotes Currently there is only a reversal agent available for dabigatran. Whilst in development, there are currently

no specific antidotes for the other DOACs and the following points should be taken into consideration when

prescribing these drugs:

The half-life of DOACs in patients with normal renal function is between 9-14 hours.

It takes 4-6 hours to effectively lower INR using vitamin K in patients taking warfarin. Prothrombin

complex concentrate (PCC) will reverse warfarin effect immediately.

Aspirin has no antidote and a similar /higher bleeding risk to a DOAC and a duration of effect of 5-7

days.

Switching between anticoagulants

Warfarin to DOAC:

Apixaban - Stop warfarin and start DOAC once the INR is <2.0

Dabigatran - Stop warfarin and start DOAC once the INR is <2.0

Edoxaban - Stop warfarin and start DOAC once the INR is <2.5

Rivaroxaban - Stop warfarin and start DOAC once the INR is <3.0

DOAC to warfarin:

Start warfarin.

After 2 days of co-administration of warfarin and DOAC obtain INR.

Continue to monitor INR regularly until INR >2 (INR should ideally be taken before the dose of DOAC

is taken that day)

Discontinue DOAC when the INR is >2.0

Ongoing monitoring

Warfarin:

Regular INR checks must be carried out. Frequency is dependent upon patient stability.

DOACs:

Dabigatran Apixaban, Edoxaban or Rivaroxaban

U&Es, LFTs, FBC at least once a year

Repeat U&Es every 6 months if CrCl 30–60 mL/min,

patient > 75 years or fragile

Repeat U&Es every 6 months if CrCl 30–60 mL/min

Repeat U&Es every 3 months if CrCl 15–30 mL/min

More frequent U&Es /LFTs advised where intercurrent illness may impact on renal or hepatic function

Contraindications Many patients do not receive anticoagulation due to perceived contraindications. However, absolute

contraindications are relatively rare and in studies were only found to make up about 7% and the remainder

of patients had relative contraindications which do not specifically exclude them from using anticoagulants

and many may be able to be treated.

A risk of falls is not a contraindication to initiating an oral anticoagulant. For example; a patient with an

annual stroke risk of 5% would need to fall almost 300 times for the risk of falling to outweigh the stroke

reduction benefit of an oral anticoagulant.

Absolute contraindications3 Relative contraindications3

History of inter-cranial haemorrhage History of gastro-intestinal haemorrhage

Existing or recent peptic ulcer disease Unexplained anaemia

Oesophageal varices Bleeding diasthesis

Previous hypersensitivity / adverse reaction

to warfarin

Alcohol abuse

Advanced malignancy / terminal illness Renal impairment (Creatinine clearance

<15ml/min/1.73 is considered an absolute contraindication

for all DOACs)

BP >180/110 - (reconsider once BP controlled) Hepatic impairment (Child-Pugh rating C could be

considered absolute contraindication for all DOACs)

Endocarditis Adverse drug interaction

Pregnancy Non-compliance

Platelet count below 50 x 109/L Platelet count between 50-150 x 109/L 3 – These lists are non-exhaustive and for up to date information the latest SPC should be checked – available via http://www.medicines.org.uk

Child-Pugh Score

The Child- Pugh classification is a means of assessing the severity of liver cirrhosis.

If there is primary biliary cirrhosis or sclerosing cholangitis then bilirubin is classified as <68=1; 68-170=2; >170=3.

The individual scores are summed and then grouped as:

• <7 = A

• 7-9 = B

• >9 = C

Score 1 2 3

bilirubin (micromol/l) <34 34-50 >50

albumin (g/l) >35 28-35 <28

INR <1.7 1.7-2.3 >2.3

encephalopathy none mild marked

ascites none mild marked

Patient information and support A variety of patient information leaflets are available to help individuals decide whether or not to start an

anticoagulant. The Atrial Fibrillation Association has several booklets which are available to download –

www.afa.org.uk

All patients prescribed a DOAC should be provided with an alert card. The Cardiac Advisory Group of the

Northern England Strategic Clinical Networks have developed an alert card which is suitable for use with any

of the DOACs currently available. Supplies of this card may be requested by emailing –

england.nescn@nhs.net

A document containing answers to common questions on the use of oral anti-coagulants in non-valvular AF

is available via the Specialist Pharmacy Service (formerly UKMI) – https://www.sps.nhs.uk/articles/common-

questions-and-answers-on-the-practical-use-of-oral-anticoagulants-in-non-valvular-atrial-fibrillation/

Rate and Rhythm Control Management in Non-valvular AF in Primary Care

Rate Control Rate control should be first line treatment in patients with AF except in those patients:

Where their AF is reversible Who have heart failure caused by AF

Who have new onset AF

Who have atrial flutter which is suitable for ablation strategy to restore sinus rhythm

Where rhythm control would be more suitable

Drug choice depends on patient symptoms, heart rate, co-morbidities and preferences.

Treatment Options

1. Standard beta-blocker (not sotalol) or rate-limiting calcium channel blocker

2. Digoxin monotherapy for people with non-paroxysmal atrial fibrillation only if they are sedentary (do

no or very little physical exercise)

3. If monotherapy does not control symptoms, and if symptoms are thought to be due to poor

ventricular rate control, consider combination therapy with any 2 of the following:

a. a beta-blocker

b. diltiazem (unlicensed)

c. digoxin

NHS England guidance on items which should not routinely be prescribed in primary care recommends that

amiodarone should NOT be offered for long-term rate control.

Beta-blockers

Atenolol is the first line option in North of Tyne and Gateshead formulary, however practical advice below

could be followed:

For people with AF alone: atenolol may be preferred as it is a once daily dose and is less expensive

than other beta-blockers.

For people who have AF and have had a previous myocardial infarction (without heart failure):

propranolol (standard release), atenolol or metoprolol (standard release), may be preferred.

For people with AF and heart failure: bisoprolol or carvedilol may be preferred.

For people with AF and diabetes mellitus: a cardio-selective beta-blocker (such as atenolol,

bisoprolol or metoprolol) is preferred. Avoid beta-blockers in people who experience frequent

hypoglycaemia.

Calcium channel blockers

Refer to North of Tyne and Gateshead formulary for preferred long acting preparation

Rhythm Control Pharmacological and/or electrical rhythm control should be considered for people with atrial fibrillation

whose symptoms continue after heart rate has been controlled or for whom a rate-control strategy has not

been successful

Patient preference, co-morbidities, risks of treatment and likelihood of re-occurrence of AF should be taken

into account before deciding on treatment.

Pharmacological Treatment Options

Beta-blocker (see above for choices)

If this is contra-indicated or unsuccessful the suitability of the following drugs should be considered:

Flecainide – Should not be used in patients with known ischaemic or structural heart disease

Propafenone – Should not be used in patients with known ischaemic or structural heart disease

Disopyramide

Sotalol – should only be initiated in secondary care for rhythm control

Amiodarone and dronedarone

In the 2019 review of items which should not routinely be prescribed in primary care NHS England have

recommended that prescribers should not initiate amiodarone or dronedarone for any new patient. In

exceptional circumstances there may be a clinical need to use either amiodarone or dronedarone. Initiation

must only be undertaken in a cooperation arrangement with a multidisciplinary team and/or other

healthcare professional. Amiodarone and dronedarone can only be started by a specialist and if they are to

be continued it should be under a shared care agreement.

Starting warfarin in AF – slow loading regimen:

Initiation of DOACs in AF:

Appendix 1 – Clinical system creatinine clearance calculators EMISWEB

To access the calculator:

From the consultation screen click the button in the centre of the EMIS ribbon for ‘run template’

If the template has been used by yourself previously, it will appear in your picking list. If it has not been used

you will need to search for the correct template. In the search box type at least 4 characters from the word

creatinine and click the magnifying glass symbol to search the template picker.

Dependent upon individual practice set up, there may be more than one option available that meets the

search criteria. If more than one option appears select ‘Estimated Creatinine Clearance (Cockcroft Gault)’

In order for the Cockcroft Gault equation to be accurate we require a recent height and weight. If these have

not been updated recently or the patient’s weight has changed significantly recently these will need to be

re-measured. The template has boxes to input these values if needed.

The BMI can be calculated within the template. Then select serum creatinine from the creatinine level drop

down box and input the serum creatinine value into the value box below. Once this detail is added click the

calculate button under the box where you have input the serum creatinine value.

As you can see in the example below if the patient is on dabigatran and has a body weight equal to or over

120% of their ideal body weight, the calculator automatically substitutes ideal body weight for the actual

body weight. The weight used to calculate creatinine clearance is explained in the grey box below the

creatinine clearance value.

CrCl in overweight patient on dabigatran

Where a patient is using either apixaban, edoxaban or rivaroxaban, the template will calculate the creatinine

clearance using actual body weight. Even if they weigh over 120% of ideal body weight.

CrCl in overweight patient on apixaban, edoxaban or rivaroxaban

As with all clinical templates and decision aids this is not a substitute for individual clinical judgement and

any decisions on a patient’s treatment should be made taking into account any other factors such as co-

morbidities and medication which may have an effect on the proposed course of treatment.

SystmOne Select Clinical Tools>Renal Disease Calculations

The calculator will then appear. Notice that the dates of the values do not appear- You must check that these are recent before saving the calculation. Check this in pathology and radiology or from the journal