Article Text

Download PDFPDF

Indications and appropriate selection of novel oral anticoagulants in patients with atrial fibrillation
  1. Michael Ghannam,
  2. Aman Chugh
  1. Department of Cardiovascular Medicine, University of Michigan, Ann Arbor, Michigan, USA
  1. Correspondence to Michael Ghannam, Department of Cardiovascular Medicine, University of Michigan, Ann Arbor, Michigan,USA; mousajab{at}

Statistics from

Learning objectives

  • Understand the major landmark trials of the novel oral anticoagulants and how these agents compare to dose-adjusted warfarin.

  • Learn to identify which patients with atrial fibrillation are candidates for novel oral anticoagulants.

  • Understand how patient specific factors can guide selection between these newer agents.


Atrial fibrillation (AF) is the most common cardiac arrhythmia and the incidence and prevalence of this disease are rising.1 2 It is estimated that by the year 2050 there will be over 7.5 million patients with AF in the USA alone.3 This disease carries a significant risk of morbidity and mortality driven in large part by the development of heart failure and thromboembolism.4 Oral anticoagulation therapy is effective in reducing risk of stroke and systemic embolism and is the standard of care for patients undergoing either a rhythm or rate control strategy.

For years warfarin has been the mainstay of stroke prophylaxis. More recently, direct-acting oral anticoagulants (DOACs) have entered the marketplace.5 Their site of action within the coagulation cascade is outlined in figure 1.6 7 These newer agents have rapidly been incorporated into modern practice,8 though there remains a role for warfarin among various subgroups. Selecting an appropriate agent is a complex decision and must take account a patient’s comorbidities, preference for dosing and monitoring, concern over the lack of a specific antidote, and importantly, financial constraints.

Figure 1

Overview of the coagulation system and sites of action of novel oral anticoagulants. The extrinsic pathway is activated by the exposure of tissue factor-expressing cells to blood through vascular injury. The intrinsic pathway is activated by the binding of factor IXa to factor VIIIA on anionic cell surfaces to form the intrinsic tenase complex. Activated platelets provide binding sites for this interaction. These processes then proceed through a common pathway ultimately resulting in activated factor IIa (thrombin) and the conversion of soluble thrombin to fibrin.6 These sites of action of direct-acting oral anticoagulants are shown below (adapted from Mega and Simon7 with permission from Elsevier Saunders).

Meta-analyses have compared the safety and efficacy of DOACs with warfarin for prevention of thromboembolic disease in non-valvular AF.9–11 Overall these newer agents are associated with an improved safety profile and reduced stroke rate compared with warfarin, and many guidelines support the use of newer agents over warfarin non-valvular AF (table 1).4 12–14 To date, there are no studies directly comparing the newer agents against one another, though indirect comparisons have been performed.15 The purpose of this review article is to discuss indications and appropriate selections of the novel oral anticoagulants in patients with non-valvular AF.

Table 1

Summary of pertinent guideline recommendations from major writing groups regarding the selection of warfarin versus novel oral anticoagulants

Background on newer agents


Similar to bivalirudin and argatroban, dabigatran is a direct thrombin inhibitor, which prevents the conversion of fibrinogen to fibrin thus preventing the development of thrombus. It was evaluated in the landmark RE-LY study, which compared two doses of dabigatran against warfarin with a goal international normalised ratio (INR) of 2.0–3.0.16 Dabigatran 150 mg (D150) twice daily was superior to warfarin with respect to the primary outcome of stroke or systemic embolism: 1.11% in the D150 group (relative risk (RR) 0.66; 95% CI 0.53 to 0.82; p<0.001 for superiority) vs 1.53% per year in the dabigatran 110 mg twice daily group (RR 0.91; 95% CI 0.74 to 1.11; p<0.001 for non-inferiority) vs 1.69% in the warfarin group. There was no significant increase in the rate of major bleeding. A reduced dose of dabigatran (75 mg twice daily) for patients with an estimated creatinine clearance (CrCl) of 15–30 mL/min has been approved based on pharmacokinetic data from RE-LY compared with a small study of patients with compromised renal function.17


Rivaroxaban is a factor Xa inhibitor approved for the treatment of non-valvular AF and venous thromboembolic disease. Its clinical safety and efficacy was demonstrated in the ROCKET-AF trial.18 This was a large, multicentred randomised controlled trial (RCT) that compared fixed-dose rivaroxaban with dose-adjusted warfarin in patients with non-valvular AF. Over one-half of the study population had suffered a prior stroke or systemic embolism as compared with 20% and 19% in RE-LY and ARISTOTLE (which evaluated apixaban; see below), respectively.16 19 The primary endpoint of combined stroke or systemic embolism occurred in the rivaroxaban group at a rate of 1.7% per year, and in the warfarin group at a rate of 2.2% per year (HR 0.79; 95% CI 0.66 to 0.96, p<0.001 for non-inferiority). An intention-to-treat analysis demonstrated non-inferiority versus warfarin but was negative for superiority. There was no statistically significant increase in the rate of major or non-major bleeding with rivaroxaban as compared with warfarin (HR 1.03; 95% CI 96 to 1.11, p=0.44). Importantly, rivaroxaban was associated with a lower rate of intracranial haemorrhage (0.5% vs 0.7%, p=0.02) and fatal bleeding (0.2% vs 0.5%, p=0.003) as compared with warfarin.


Apixaban is also a direct factor Xa inhibitor and is approved for the treatment of non-valvular AF. The ARISTOTLE trial was a double-blinded RCT that compared apixaban and warfarin in patients with non-valvular AF.19 The trial was designed to demonstrate non-inferiority with respect to warfarin for the combined endpoint of stroke or systemic embolism. A key secondary objective was to test for superiority with regard to the primary outcome as well as rates of major bleeding and death from any cause. The annual event rate was 1.27% in the apixaban group as compared with 1.60% in the warfarin group (HR 0.79; 95% CI 0.66 to 0.95; p<0.001 for non-inferiority; p=0.01 for superiority). This was accompanied by a statistically significant decrease in major bleeding (2.13% vs 3.09% per year, HR 0.69; 95% CI 0.60 to 0.80; p<0.001), and the rate of death from any cause (3.52% and 3.94%, HR 0.89; 95% CI 0.80 to 0.99; p=0.047).


Edoxaban is the third direct Xa inhibitor to be approved for the use of stroke and systemic embolism reduction in non-valvular AF. The landmark RCT evaluating its use was the ENGAGE AF-TIMI 48 trial which compared two dosing schemes of edoxaban with dose-adjusted warfarin.20 Two doses of edoxaban were evaluated, 60 and 30 mg taken once a day. The dose was halved if any of the following characteristics were present: estimated CrCl of 30–50 mL/min, body weight of 60 kg or less, or the use of verapamil or quinidine (potent P-glycoprotein inhibitors). Edoxaban was non-inferior to warfarin for the primary efficacy endpoint of stroke or thromboembolism (1.5% per year vs 1.18% per year; HR vs warfarin 0.70; 97.5% CI 0.63 to 0.99; p<0.001 for non-inferiority, p=0.02 for superiority). Major bleeding rates were also lower with edoxaban for both the high-dose and low-dose dosing groups (3.43% per year with warfarin compared with 2.75% per year with high-dose edoxaban; HR 0.80; 95% CI 0.71 to 0.91; p<0.001) and low-dose edoxaban (HR 0.47; 95% CI 0.41 to 0.55; p<0.0001). The rates of life-threatening bleeding, intracranial bleeding, and major bleeding plus clinically relevant non-major bleeding were all lower with edoxaban. The risk of major gastrointestinal (GI) bleeding, however, was higher in the high-dose edoxaban group. These agents and their respective trials are summarised in table 2.

Table 2

Summary of randomised controlled trials of novel oral anticoagulants for use in stroke and systemic embolisation prophylaxis in atrial fibrillation

Indications for use

General considerations

With the exception of selected low-risk patients, systemic anticoagulation for prevention of stroke or thrombotic events is widely endorsed by multiple society guidelines. Most guidelines endorse the use of a DOAC over the use of dose-adjusted warfarin for the majority of patients with AF (table 1). Multiple risk assessment tools have been validated to estimate an individual’s risk for thromboembolic events. The most widely endorsed system is CHA2DS2-VASc (C, congestive heart failure; H, hypertension; A2, age >75 years; D, diabetes mellitus; S2, stroke (doubled); V, vascular disease; A, age >65 years; Sc, sex category, ie, female gender). It is recommended that patients with a CHA2DS2-VASc score of 2 or greater be treated with an oral anticoagulant.4 12 These risk assessment scores should be applied to all patients with non-valvular AF regardless of the subtype (paroxysmal, persistent or permanent) or treatment strategy (rate or rhythm control).4

Managing patients on warfarin can be complex due to the myriad drug–drug interactions, dietary counselling and restrictions, routine monitoring, and difficulty of maintaining the INR in a narrow therapeutic window.21 Newer oral anticoagulants offer an attractive solution to patients either unwilling or unable to adhere to these requirements. A concern with the use of DOACs, particularly voiced by patients, is the lack of a reversal agent. Protocols for emergency reversal of newer agents have been proposed though robust data for their efficacy are still needed.22–24 These concerns need to be weighed against the newer agents overall lower bleeding risks as compared with warfarin.

Valvular AF

All four landmark trials investigating the use of newer oral anticoagulation excluded patients with valvular AF16 18–20 though the exact criteria for a concomitant valve disorder varied substantially.25 The RE-LY trial excluded patients with a history of haemodynamically relevant valve disease excluding patients with, for example, severe aortic stenosis or regurgitation. Conversely, ROCKET-AF limited their exclusion criteria only to patients with significant mitral valve stenosis. Similar discrepancies exist regarding the inclusion or exclusion of patients with prosthetic valves or valve repairs.

The ARISTOTLE trial excluded patients with moderate to severe mitral stenosis or mechanical prosthetic valves; however, a subgroup analysis identified 4808 patients who had other forms of moderate or severe valvular heart disease as assessed by medical history or available echocardiography data.26 Patients with valvular heart disease derived benefit from the use of apixaban versus warfarin regarding reduction of stroke or systemic embolism with a comparable risk of major bleeding. Similar risk and benefit profiles were observed when the cohort was stratified by location of the valve disease (mitral, aortic or tricuspid), in patients with mild valve disease, or patients with previous valve surgery. There were low numbers echocardiography confirmed severe valvular disease which limits the broad applicability of this study. These conclusions should be considered hypothesis generating and warrant further investigation.

The RE-ALIGN trial was a phase II RCT which evaluated the safety and efficacy of dabigatran in patients with mechanical mitral or aortic valve replacement.27 The trial was terminated early due to an excess of adverse events including valve thrombosis, and ischaemic stroke in the dabigatran. Accordingly, neither dabigatran nor the other DOACs are approved in patients who require anticoagulation for valve replacement.

Hypertrophic cardiomyopathy

AF is commonly associated with hypertrophic cardiomyopathy (HCM) and these patients are at a particularly elevated risk of thromboembolic events. A recent pooled cohort study placed the prevalence of AF in patient with HCM at 22% with an incidence of thromboembolic complications of 3.8%.28 Given the high event rates, systemic anticoagulation is indicated after any documentation of AF independent of the CHA2DS2-VASc score.29 Warfarin has been shown to reduce the risk of stroke and embolisation although in a single community-based cohort study.30 In the absence of robust data on DOACs in this population, warfarin is generally recommended in patients with HCM.4 29

Cardiac amyloid

The prevalence of AF among patients with cardiac amyloidosis is about 15%.31 Autopsy series have revealed a high burden of intracardiac thrombosis (33% in one series32). Feng et al reported that left ventricular diastolic dysfunction (OR 15.2, 95% CI 7.5 to 999.99, p=0.008) and AF (OR 11.8, 95% CI 1.0 to 395.2, p=0.05) strongly predicted the presence of thrombus.33 Further, anticoagulation therapy was effective in preventing thromboembolic events. The anticoagulation protocol was not reported in this study, and there are no data on the use of DOACs in this population. Patients with cardiac amyloid are also at risk of bleeding in part due to the fragile blood vessel from amyloid deposition as well as coexisting coagulopathy.34 The decision to anticoagulate with either warfarin or a DOAC must carefully balance these risks.


The X-VeRT trial is the only prospective study that examined the utility and safety of one of the novel oral anticoagulants in patients undergoing transthoracic cardioversion.35 Over 1500 patients with AF of greater than 48 hours or of an unknown duration received either rivaroxaban or warfarin and underwent either an early (transoesophageal echocardiogram (TEE) guided) or delayed (3 weeks of uninterrupted anticoagulation) cardioversion. A net clinical benefit outcome (the composite of stroke, non-central nervous system systemic embolism, transient ischaemic attack, myocardial infarction, cardiovascular death and major bleeding) occurred in 1.06% of patients receiving rivaroxaban and 1.81% of patients receiving vitamin K antagonist (VKA) (risk ratio 0.49; 95% CI 0.14 to 1.69). Rivaroxaban taken 4 hours before TEE-guided cardioversion in the early group was deemed safe. Rivaroxaban was associated with a shorter time to cardioversion in the delayed group due to difficulty in maintaining a therapeutic INR for 3 weeks in the warfarin group.

Post hoc data from ARISTOTLE, RE-LY and ROCKET-AF also support the use of newer anticoagulants in patients undergoing transthoracic or pharmacologic cardioversion.36–38 The largest dataset was derived from the RE-LY trial in which 1983 cardioversions were performed in 1270 patients. Stroke and systemic embolism rates within 30 days were low (0.77%, 0.30% and 0.6% in patients taking dabigatran at 110 mg and 150 mg, and dose-adjusted warfarin, respectively) with no statistically significant difference between groups. Post hoc data from ROCKET-AF and ARISTOTLE reached similar conclusions but these studies were hampered by a small number of patients undergoing cardioversion (n=285 patients in ROCKET-AF and n=540 patients in ARISTOTLE).

Catheter ablation of AF

Many operators perform catheter ablations of AF on uninterrupted warfarin39 40 while the optimal periprocedural management of DOACs is less understood. Lakkireddy et al 41evaluated the safety of uninterrupted dabigatran and found that it was associated with higher rates of major bleeding (6% vs 1%, p=0.019) as well as composite bleeding and thromboembolic complications (16% vs 6%, p=0.009) compared with warfarin use. In contrast, a strategy of interrupted dabigatran versus continuous or interrupted warfarin was studied in a retrospective manner by Somani et al 42who found no difference in the rates of thromboembolic or bleeding complications. A trial evaluating the safety and utility of apixaban for prevention of thromboembolic events during catheter ablation of AF is currently recruiting patients. The VENTURE AF trial was an exploratory RCT comparing uninterrupted rivaroxaban to uninterrupted warfarin.40 While not powered for statistical significance, adverse event rates were equally low in both groups. Currently, the preferred strategy is to perform left atrial ablation during uninterrupted oral anticoagulation with warfarin. In patients who are taking one of the novel agents, a reasonable approach is to discontinue the drug for two to three doses prior to the procedure and resume it after haemostasis.

Special populations

High stroke-risk patients

The mean CHADS2 score for patients enrolled in the RE-LY or ARISTOTLE trial was both 2.1 while patients in the ROCKET-AF group had a slightly higher mean score of 3.5. Subgroup analyses on patients with high stroke risk (generally defined as CHADS2 score ≥3) have been performed on RE-LY, ARISTOTLE and ROCKET-AF data sets.43–45 These post hoc analyses suggest the advantages of DOACs over warfarin extend to these high-risk populations as well.

High bleeding-risk patients

Identifying patients with an elevated bleeding risk can be performed with risk scores such as the HASBLED scheme, which incorporates various patient characteristics: H=hypertension, A=abnormal renal or liver function, S=stroke history, B=bleeding diathesis, L=labile INR, E=elderly (>65 years), D=drug/alcohol use. However, acting on these data can be difficult as these same patients often are at higher stroke risk as well.46 Banerjee et al 47performed a modelling analysis based on the Danish National Patient Registry to assess the net clinical benefit of warfarin, dabigatran, apixaban and rivaroxaban in patients. They concluded that the newer anticoagulants have a net clinical benefit in all patient groups with the exception of those with CHA2DS2VASC=0 and HAS-BLED ≥3. Patients with CHADSVASC≥2 and HAS-BLED ≥3 derived clinical benefit from all three newer oral anticoagulants. Apixaban appears superior to either dabigatran or rivaroxaban in terms of its risk of GI bleeding.48 As compared with warfarin, the HR of GI bleeding for apixaban was 0.89 (p=0.37), and for dabigatran was 1.5 (p<0.001). In ROCKET-AF, GI bleeding was reported as percentage and no HRs were available; however, GI bleeding occurred in 3.2% of patients on rivaroxaban as compared with 2.2% on warfarin (p<0.001).

Renal impairment

All of the newer oral anticoagulants undergo some degree of renal clearance and the recommended dose adjustments per the joint guidelines are shown in table 3. Sardar et al 49 performed a meta-analysis of RCTs examining newer oral anticoagulants (including trials investigating the use of the DOACs) and concluded that at the recommended dose reductions for moderate (estimated glomerular filtration rate (eGFR) of 30–49 mL/min) or mild (eGFR of 50–79 mL/min) renal impairment, these agents were effective and safe. In a recent meta-analysis by Ruff et al 50 (the first to include all four new agents), appropriately dosed DOACs as compared with warfarin did not demonstrate increased risks of major bleeding or ischaemic stroke in patients with various levels of renal dysfunction, including CrCl <50 mL/min.

Table 3

Dosing considerations for novel oral anticoagulants. Common P-gp inhibitors ketoconazole, verapamil, amiodarone, dronedarone, cyclosporine and erythromycin. Rifampin is a potent inducer of P-gp. Child-Pugh classification is a prognostic scoring system used to assess the severity of chronic liver disease, mainly cirrhosis, which incorporates clinical and laboratory variables (total bilirubin, serum albumin, severity of ascites, hepatic encephalopathy grade). Online calculator can be accessed at 5 52 69

All four landmark trials evaluating each of the newer agents included patients with stage III chronic kidney disease (CKD), corresponding to an eGFR of 30–49 mL/min. Subgroup analysis demonstrated superiority of dabigatran at a dose of 150 mg and non-inferiority of dabigatran at 110 mg, rivaroxaban, and apixaban for stroke prevention without an increased bleeding risk among patients with moderate CKD (eGFR of 25–50 mL/min for apixaban, 30–49 mL/min for dabigatran and rivaroxaban).51 A dose reduction of edoxaban to 30 mg daily for patients with moderate to severe renal impairment (30–50 mL/min) was shown to be superior to warfarin in the ENGAGE AF-TIMI study. Post hoc analysis has demonstrated an increased rate of ischaemic stroke with edoxaban 60 mg daily relative to warfarin (HR 2.16, 95% CI 1.17 to 3.97) in patients with a CrCl >95 mL/min).52 Edoxaban is contraindicated in such patients.

Patients with severe renal impairment were excluded from the randomised trials. Recently, the Food and Drug Administration approved the use of dabigatran 75 mg twice daily for patients with an eGFR of 15–30 mL/min though other regulatory agencies have not done so.51 Apixaban was also recently approved for patients on haemodialysis (HD) and is currently the only newer oral anticoagulant to be approved for patients on HD at a dose of 5 mg twice daily with a dose reduction to 2.5 mg twice daily for patients age ≥80 or body weight ≤60 kg.53 These latter two recommendations were made on the basis of pharmacokinetic and pharmacodynamic data, and were not derived from clinical trial data.54

Elderly patients

Anticoagulation with VKA in the elderly is challenging due to low body weight, poor diet, fluctuations in health, dementia and other comorbidities.55 Studies of elderly patients in long-term care facilities have demonstrated rates of time in appropriate therapeutic windows as low as 37%.56 Newer oral agents may reduce these barriers given their ease of dosing and more predictable response.

Elderly patients have historically been under-represented in AF trials examining stroke prophylaxis.55 Collectively, the landmark studies examining the newer oral anticoagulants enrolled over 19 000 patients age ≥75 years.16 18 19 A meta-analysis by Sardar et al 57examined the efficacy and safety of these agents in a subgroup of elderly patients and concluded that DOACs did not result in excess bleeding and were associated with equal or greater efficacy than conventional therapy. It appears as though rivaroxaban (OR 0.65, 95% CI 0.49 to 0.87) and apixaban (OR 0.61, 95% CI 0.49 to 0.76) are superior to warfarin therapy while dabigatran does not pose additional harm (OR 1.07, 95% CI 0.90 to 1.28). Concerns of increased bleeding risk in the elderly with newer oral anticoagulants were reported in case series and anecdotal evidence.58 59 However, on the strength of the report by Sardar et al as well as other meta-analyses,55 age alone should not preclude their use. Factors such as low body weight and renal dysfunction should still prompt concern for additional bleeding risks and appropriate dose adjustments.

Triple therapy

It is not uncommon for patients to require treatment with multiple antiplatelet agents as well as an oral anticoagulant. Among patients with AF who have an indication for oral anticoagulation, 20%–30% have coexisting coronary artery disease.60 Patients with AF who undergo percutaneous coronary intervention require dual antiplatelet therapy (DAPT)61 as well as an oral anticoagulation for AF, so-called triple therapy (TT). The PIONEER AF-PCI trial suggested that rivaroxaban plus P2Y12 monotherapy at the time of coronary intervention leads to lower rates of clinically significant bleeding as compared with TT.62

A post hoc analysis of the RE-LY trial was performed which examined the 6952 patients who received concomitant aspirin or clopidogrel therapy at any time during the study period.63 The use of DAPT or TT did not affect the advantage of dabigatran over warfarin in preventing thromboembolic disease. Additional antiplatelet therapy did increase risk of major bleeding for both DAPT (HR 1.60; 95% CI 1.42 to 1.82) and TT (HR 2.31; 95% CI 1.79 to 2.98). When using a newer oral anticoagulant as part of TT regiment, it is recommended to use the lowest approved dosing to help prevent excessive bleeding.64 While lifelong oral anticoagulation therapy is usually warranted, patients should be frequently re-evaluated regarding their need for DAPT or TT.

The ongoing ORBIT-AF II study is a multicentre, prospective, outpatient disease registry which will specifically evaluate the use of target-specific antithrombotic agents in a real-world AF cohort and is poised to address these issues as well.65


Newer oral anticoagulants have demonstrated significant advantages over warfarin for preventing stroke and thromboembolic events with equal or lower rates of bleeding. These advantages persist across a wide range of patient populations. Major society guidelines have encouraged the use of these agents as first line over warfarin for a majority of patients with non-valvular AF (table 2), particularly in patients unable to maintain a therapeutic INR.

Selection between DOACs is limited by a lack of direct head-to-head comparisons as well as lingering questions raised by postmarket safety data. Pragmatic considerations such as dosing preference can be useful but other patient-specific factors should be taken into consideration. A suggested algorithm for selecting an appropriate DOAC is show in figure 2.66–68

Figure 2

Suggested algorithm for selecting anticoagulation therapy for patients with atrial fibrillation. See text for descriptions of CHA2DS2-VAScS and Child-Pugh scoring system. HAS-BLED and HEMORR2HAGES are risk factor assessment tools which have been validated to predict short-term bleeding risk of systemic anticoagulation (mostly dose-adjusted warfarin) in patients with AF.66 *INR goal for mechanical heart values may be higher depending on clinical factors.67 †See text for more information regarding the definition of ‘valvular atrial fibrillation.’ ‡Systemic anticoagulation for patients with CHA2DS2-VAScS of 1 is less defined, it is reasonable to consider either systemic anticoagulation or aspirin therapy in this population.68 AF, atrial fibrillation; CrCl, creatinine clearance; NOAC, novel oral anticoagulant; INR, international normalised ratio; P-gp, permeability glycoprotein, twice daily dosing; HOCM ,hypertrophic obstructive cardiomyopathy; ESRD, end stage renal disease; HD, hemodialysis.

Key messages

  • Direct-acting oral anticoagulants (DOACs) offer an alternative to warfarin for the majority of patients with non-valvular atrial fibrillation (AF). In large randomised trials these agents have been found to be as efficacious as dose-adjusted warfarin in preventing stroke or systemic embolism with an improved bleeding profile, including a lower risk of intracranial haemorrhage.

  • DOACs are not approved for use in valvular AF, particularly in patients with mitral stenosis and mechanical valves. There are little data to guide the use of DOACs in conditions such as hypertrophic cardiomyopathy or cardiac amyloidosis. There is growing evidence for the safety of DOACs in the setting of cardioversion and catheter ablation of AF.

  • To date, DOACs have not been compared with one another in a randomised, prospective manner which limits recommendations in choosing one agent over another. Patient-specific factors such as dosing preference (daily vs twice a day), renal function, liver function, gastrointestinal bleeding risks and drug–drug interactions can help direct providers.

You can get CPD/CME credits for Education in Heart

Education in Heart articles are accredited by both the UK Royal College of Physicians (London) and the European Board for Accreditation in Cardiology—you need to answer the accompanying multiple choice questions (MCQs). To access the questions, click on BMJ Learning: Take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to:

Please note: The MCQs are hosted on BMJ Learning—the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into Heart with their journal’s username and password. All users must also complete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.



  • Contributors The authors contributed equally in the creation of this manuscript. All authors read and approved the final manuscript.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.