2 Research recommendations

The Guideline Development Group has made the following recommendations for research, based on its review of evidence, to improve NICE guidance and patient care in the future.

2.1 Cognitive behavioural therapy for people with atrial fibrillation

What is the clinical and cost effectiveness of cognitive behavioural therapy (CBT) compared with usual care for people with newly diagnosed atrial fibrillation?

Why this is important

There is currently little evidence to support psychological care for people with atrial fibrillation. Assessing the effectiveness of CBT in lowering people's levels of anxiety, reducing episodes of planned or unplanned care and improving their quality of life will help determine whether CBT should be an essential component of atrial fibrillation services. [new 2014]

2.2 Rate control drug treatment for people aged 75 and over with atrial fibrillation

What is the comparative effectiveness of the 3 main drug classes used for rate control (beta‑blockers, calcium‑channel blockers and digoxin) in people aged 75 and over with atrial fibrillation in controlling symptoms, improving quality of life and reducing morbidity and mortality?

Why this is important

Atrial fibrillation is the most common arrhythmia in people aged 75 and over, with a prevalence of more than 15%. This guideline recommends rate control of atrial fibrillation as the treatment of choice. However, there are no good‑quality randomised controlled trials (RCTs) comparing the 3 main drug classes (beta‑blockers, calcium‑channel blockers and digoxin) used for rate control, and no studies specifically addressing people aged 75 and over.

Drug treatment for rate control in people aged 75 and over with atrial fibrillation is particularly challenging because of comorbidities. For example, heart failure is prevalent in this age group but RCTs comparing beta‑blockers with digoxin in people aged over 70 are of low quality. Although these RCTs suggest no advantage of beta‑blockers compared with digoxin for rate control, and an increased incidence of hospitalisations with heart failure, current guidelines propose beta‑blockers as first‑line therapy for rate control. Other conditions such as chronic kidney disease, ischaemic heart disease, valvular heart disease, concomitant heart conduction disorders, dementia, pulmonary disease, hypo‑ and hypertension and frailty might also affect the choice of drugs for this age group. Optimal treatments for people with these comorbidities are not known.

Optimising drug treatment for atrial fibrillation in this age group has the potential to reduce hospitalisations and the need for services such as GPs and specialist nurses to manage secondary symptoms, with consequent economic benefits. [new 2014]

2.3 Case volume as an indicator of quality for people offered left atrial catheter ablation

What is the effect of case volume on complications and outcomes after left atrial catheter ablation?

Why this is important

As interest in left atrial catheter ablation for atrial fibrillation increases, more clinicians are taking up this procedure. Many people offered left atrial catheter ablation want to know whether they will receive safe and effective treatment. If increased experience and case volume are associated with improved outcomes, the case volume of a centre or a clinician is an easily measurable parameter that people with atrial fibrillation could use to help judge the quality of the procedure they are likely to receive. [new 2014]

2.4 Non‑vitamin K antagonist oral anticoagulants

Do people with atrial fibrillation whose anticoagulant control is poor, or is predicted to be poor, with warfarin benefit from changing to one of the non‑vitamin K antagonist (non‑VKA) oral anticoagulants?

Why this is important

Trials of the non‑VKA oral anticoagulants have shown that the degree of benefit of these agents compared with warfarin may depend on the time in therapeutic range (TTR) of the warfarin group. These trials assessed the degree of benefit in relation to the mean TTR for the warfarin group in that country.

However, the inference of benefit is based on a number of assumptions. It is unclear that the population TTR can be extrapolated to decision‑making in an individual. If, for example, an individual's low TTR is a result of poor compliance, it is unlikely that compliance will improve with a non‑VKA oral anticoagulant and uncertain whether a non‑VKA oral anticoagulant will offer any benefit. Moreover, the threshold of TTR at which a non‑VKA oral anticoagulant might offer benefit is unclear. The same question can be extended to include people before they start warfarin treatment, using criteria that prospectively identify those likely to have poor control on warfarin.

A study with 2 arms should be carried out. One arm should randomise people with atrial fibrillation, whose control is poor with warfarin, to either continue warfarin treatment or change to a non‑VKA oral anticoagulant. The other arm should randomise people newly diagnosed as having atrial fibrillation, who have not previously had anticoagulant therapy and in whom poor anticoagulant control is predicted (using the SAMe‑TT2R2 score[9]), to have treatment with either warfarin or a non‑VKA oral anticoagulant. Outcomes should include stroke and other thromboembolic complications, major haemorrhage and death. [new 2014]

2.5 Stroke risk assessment

Can routine data from UK primary care databases clarify stroke risk in people with atrial fibrillation according to baseline risk factors and treatment?

Why this is important

There are several scores available to predict stroke risk in people with atrial fibrillation. Most have been derived from secondary care populations and validated in non‑UK populations. The availability of routine primary care databases such as CPRD (The Clinical Practice Research Datalink) and THIN (The Health Improvement Network) provides the opportunity to assess these risk tools, and the impact of treatment on risk, in a non‑selected UK population.

A prospective cohort study should be carried out to establish baseline risk in people with atrial fibrillation, using established risk scores, and to prospectively evaluate the outcomes of stroke and mortality, taking into account treatment and changes in risk over time.

The results would help determine the most effective means of providing stroke prevention in a non‑selected general practice population and establish the discriminatory value of existing stroke risk scores. [new 2014]



[9] The SAMe‑TT2R2 score is defined as: 'Sex female, Age <60 years, Medical history (at least 2 of the following: hypertension, diabetes, coronary artery disease/myocardial infarction, peripheral arterial disease, congestive heart failure, previous stroke, pulmonary disease, hepatic or renal disease), Treatment (interacting drugs, for example, amiodarone for rhythm control): all 1 point; plus current Tobacco use (2 points) and Race (non‑Caucasian, 2 points)'.

Apostolakis S, Sullivan RM, Olshansky B et al. (2013) Factors affecting quality of anticoagulation control among patients with atrial fibrillation on warfarin: the SAMe-TT₂R₂ score. Chest 144: 1555–63.

  • National Institute for Health and Care Excellence (NICE)