Statistics from Altmetric.com
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.
To learn the optimal management of patients with out-of-hospital cardiac arrest (OOHCA).
Perform basic life support and advanced cardiac life support (ACLS).
Lead and coordinate the actions of an ACLS team.
Appreciation of the importance of working in a team with lay persons, paramedics and other medical personnel during resuscitation.
Recognising the urgent management and triage of OOHCA survivors.
Emergency and peri-resuscitation echocardiography.
To appreciate the literature supporting an early invasive strategy investigating OOHCA survivors.
To understand optimal protocols and patient pathways to improve outcome.
Diagnostic work-up and risk stratification of survivors.
Causes of cardiorespiratory arrest, identification of patients at risk and early implementation of corrective treatment of reversible causes.
Curriculum topic: Acute cardiovascular care
The scale of the problem
Out-of-hospital cardiac arrest (OOHCA) remains a leading cause of death in developed countries.1 In spite of clear improvements in treatments and patient pathways, all-comer survival is still <10%,1 ,2 with significant variability in registry outcomes dependent on geography, sophistication of ambulance services and the clinical experience of treating centres.3 Between 2013 and 2014, of 28 000 reported OOHCA cases in England, the overall survival to hospital discharge was only 8.7%.4 However, these data encompass all patient groups, including those with and without return of spontaneous circulation (ROSC), those with and without ‘shockable heart rhythms’ (SHR) and patients with highly variable time delays to implementation of resuscitation—all important factors that significantly skew results and outcomes. These factors also challenge the interpretation of the available literature.
The ‘chain of survival’
Approximately 80% of OOHCAs occur at home and 20% in public places.5 The current rate of initial bystander cardiopulmonary resuscitation (CPR) in England is reported as 43%,6 including both spontaneous bystander-initiated CPR and bystander-performed CPR prompted by Emergency Services over the telephone. Approximately 20% of OOHCA patients are in an SHR (ie, treatable by defibrillation) by the time the emergency team arrives. SHR mostly comprises ventricular fibrillation (VF), which, according to the ‘Utstein’ definition, should be distinguished from asystole with a deflection of >1 mm in amplitude on the surface ECG.7 Survival in this cohort is significantly higher than in those patients with non-SHR,8 although early defibrillation is crucial, with a 10% decrease in survival with every minute of delay.9
The concept of the ‘chain of survival’ has been introduced to standardise OOHCA protocols used worldwide.10 Time to first emergency response, witnessed cardiac arrest, effective bystander CPR, presence of an initial SHR, early defibrillation and prehospital ROSC are all established key factors in determining outcome.
Prehospital strategies to facilitate early diagnosis and treatment of OOHCA patients have proven important in improving survival and outcomes. Public education programmes and use of automatic external defibrillators (AEDs) have both been suggested to improve outcomes in VF arrests. However, the effectiveness of AEDs is strongly related to a large number of variables including operator experience, location, the emergency response system, ongoing maintenance and evaluation,12 thus making it difficult to extrapolate results for the general population.
Once ROSC is obtained and the airway secured, 100% inspired oxygen should be titrated to the lowest level required to achieve an arterial oxygen saturation of ≥94%. Hyperventilation or ‘overbagging’ the patient is common and can reduce cardiac output through metabolic derangement and venous impedance. Hyperoxia and hyperventilation can additionally worsen reperfusion injury and global cerebral damage.13
Although advanced life support (ALS) is often assumed to improve clinical outcomes, its routine adoption is controversial. A recent large observational study showed that patients with OOHCA receiving basic life support had higher survival rates to hospital discharge and at 90 days compared with those who received ALS and were less likely to experience poor neurological outcomes.14 Accepting the limitations of potential bias associated with these observational registries, this confirms previous reports15 suggesting that prehospital care should not delay swift transfer to a Cardiac Arrest Centre, where definitive diagnostic and therapeutic interventions can be offered immediately (eg, percutaneous coronary intervention for acute myocardial infarction or induced therapeutic hypothermia).
In-hospital treatment of OOHCA survivors: optimum patient pathways
Approximately one in four patients successfully resuscitated from OOHCA and admitted to hospital survives to discharge.16 ,17 The likelihood of survival appears to correlate with the number of OOHCA cases treated by any individual institute18 ,19—a volume–outcome relationship found in many other areas of cardiology. The leading causes of in-hospital death in this group are central nervous system injury and myocardial failure.20 ,21 Perhaps not surprisingly, effective postcardiac arrest hospital care has been shown to reduce early mortality caused by haemodynamic instability and complications from multiorgan failure and brain injury.22–24 Multiple physiological systems are affected following cardiac arrest, and multidisciplinary, goal-directed therapies have been introduced as part of a ‘bundle of care’ designed to improve patient survival.25–27 A pivotal study in defining which strategies might be integral to this was published in 2007 by Sunde et al.28 These authors demonstrated that a more aggressive approach based on the use of mild therapeutic hypothermia, early coronary angiography (CAG) and revascularisation of culprit coronary disease increased survival from 26% to 61%, with excellent neurological outcomes.29
Is immediate CAG recommended for all OOHCA survivors?
Evidence from the literature
Obstructive coronary disease is found in >70% of OOHCA survivors, as determined by autopsy reports and premortem angiographic data.30 ,31 The 2012 European Society of Cardiology guidelines on ST-segment elevation myocardial infarction (STEMI) suggest that all OOHCA survivors with ECG evidence of STEMI undergo immediate CAG with a view to primary percutaneous cardiovascular intervention (PCI), as well as those patients with a high index of suspicion for ongoing ischaemia, even where there is no clear ST-segment elevation.32 This is now supported in expert consensus guidance published more recently by the European Association for Percutaneous Cardiovascular Intervention33 (figure 1).
However, several key questions remain, including the exact timing of CAG, the benefits of ‘early’ over ‘delayed’ revascularisation versus medical therapy alone, and the pharmacotherapy strategy employed, where intervention is deemed appropriate. There are no randomised studies in this specific area, as OOHCA survivors have either been systematically excluded or significantly under-represented in major randomised interventional STEMI34 and non-ST-segment elevation acute coronary syndrome (NSTE-ACS) trials.35 Nevertheless, a number of small observational studies looking at the use of an early invasive strategy have been published over the past decade.28 ,29 ,36–63 A synoptic review of these studies confirms this approach as feasible and safe, and also highlights an increase in uptake of this strategy over recent years (table 1). A meta-analysis of 3981 OOHCA patients from the USA, of whom 19% and 17% underwent CAG and PCI within 24 h of admission respectively, confirmed that a favourable neurological outcome was independently associated with early CAG and reperfusion.64 A more recent meta-analysis confirmed this trend by showing an overall survival in the acute CAG group of 58.8% versus 30.9% in the control group, with improved neurological outcomes (58.0% vs 35.8%).65 Once again, bias may be an important confounder in these non-randomised studies, where good clinical decision-making may have diverted patients in better condition (therefore more likely to survive) towards an earlier, more interventional approach.
As most observational OOHCA studies have included comatose patients, coma on admission should not per se be a contraindication for an early invasive strategy. However, unfavourable prehospital features related to cardiac arrest and initial resuscitation indicating a remote likelihood for neurological recovery, as well as known comorbidities indicating poor life expectancy, should be strongly considered in clinical context before any invasive strategy, regardless of post-resuscitation ECG findings.66
Optimum timing of angiography and other urgent investigations
The rationale for early revascularisation in OOHCA survivors is no different from patients without OOHCA—to restore coronary perfusion, secure unstable coronary lesions and to reduce ongoing ischaemia as a substrate for arrhythmia and pump failure. Unless there is a clear alternative explanation for the patient's cardiac arrest, all OOHCA survivors to hospital admission should be considered for early CAG67—hence the importance of the ‘ER stop’ which facilitates this decision. The ECG is still the frontline diagnostic tool to start the decision process. All OOHCA survivors with ST-segment elevation on the post-resuscitation ECG should be considered for immediate CAG and transferred directly to a designated Heart Attack Centre for further management. Patients with a presumed acute coronary syndrome without ST-segment elevation should ideally undergo CAG within 2 h. This timeframe is considered crucial in enabling careful assessment by a multidisciplinary team to exclude other possible causes of OOHCA (figure 2).
Routine CT imaging of the brain, thorax and abdomen may be considered, when initial diagnostic uncertainty remains. It is important that all patients undergo early emergency bedside echocardiography to facilitate these decisions. A handheld echo can assess ventricular function, regional wall motion abnormalities, valvular heart disease, tamponade, aortic root dissection or acute RV changes due to massive pulmonary embolism.68 CT brain imaging should not delay coronary investigation unless a cerebrovascular event is suspected as a cause of the cardiac arrest or trauma involving the head raises the possibility of intracranial bleeding. Once obvious confounding diagnoses are excluded, all remaining patients should then proceed directly to CAG as soon as possible according to guidelines for high-risk NSTE-ACS35 (figure 2).
When is coronary revascularisation indicated?
A recent meta-analysis of angiographic data in resuscitated OOHCA patients studied acutely confirms a high prevalence of ‘significant’ coronary disease, ranging from 59% to 71% of patients.69 Studies addressing different revascularisation strategies in relation to angiographic characteristics, completeness of revascularisation and adjunctive therapies are lacking. It is currently recommended that clear ‘culprit’ lesions defined variably by acute occlusion, evidence of thrombus or correlation with ECG and echo data, are managed immediately by PCI (just as they would be in other scenarios), in order to decrease the incidence of recurrent cardiac arrest, to reduce infarct size and to improve haemodynamic stability.33 Optical coherence tomography (OCT), due to its ability to provide unique information on plaque composition and presence of thrombus, can be useful in identifying ‘culprit’ lesions if clinical and angiographic evidence is lacking. In STEMI patients, OCT has a higher sensitivity than angioscopy and intravascular ultrasound in detecting plaque rupture.70
The current literature does not support the treatment of bystander lesions at the time of OOHCA angiography, unless there is evidence of ongoing electrical or haemodynamic instability.71 To what extent this will change in light of recent data supporting complete revascularisation in non-OOHCA STEMI populations is unclear,72 although it is again recognised that there are additional physiological considerations between these different groups.
Coronary artery spasm has been associated with lethal arrhythmias and sudden death, though its diagnosis remains a challenge, particularly in patients acutely unwell with heightened adrenergic states.73 ,74 Where clinical suspicion is high, provocation testing using either intracoronary acetylcholine or ergonovine can be considered at repeat angiography once the patient is stable and recovered.75 If spasm is considered likely, use of calcium channel blockers is recommended together with a lower threshold for implantable defibrillators.76 ,77
Optimum PCI strategy: access choice, stent selection and antiplatelet agents
The attempt to define the optimal PCI strategy in OOHCA survivors is challenging given the lack of data. Radial access has been associated with a lower mortality in patients with STEMI and no OOHCA.78 For this reason, using the radial approach as the default strategy in OOHCA survivors, if the operator is experienced, may be preferable. However, femoral access may have advantages in haemodynamically compromised patients requiring a percutaneous assist device or temporary pacing wire.
Similar to arterial access site, there are no published data regarding stent selection in the setting of PCI in OOHCA patients, so careful consideration of the need for longer-term dual antiplatelet therapy but potentially better long-term clinical outcome should be made when deciding to deploy a drug-eluting stent in OOHCA survivors.
Stent thrombosis has been suggested to be more frequent in OOHCA survivors,79 although conflicting data exist.80 The main pharmacokinetic characteristics of antiplatelet agents currently used in the setting of PCI are summarised in table 2. Since comatose survivors are intubated and mechanically ventilated, administration of crushed tablets via a nasogastric tube remains the only option.80 The challenges of drug administration and the higher risk of stent thrombosis, which may be exacerbated by therapeutic hypothermia (see below), suggest the need for the administration of an intravenous antiplatelet agent to ‘bridge’ the delayed effect of oral P2Y12 inhibitors. There is recent evidence that administration of eptifibatide in this setting results in profound platelet inhibition for at least 22 h.81 Similarly, it has been demonstrated in patients with STEMI and no cardiac arrest that a bolus of abciximab without additional infusion might be sufficient.82 Cangrelor, a novel intravenous P2Y12 inhibitor, is a promising agent in this setting due to its rapid onset (and offset) of action and its effectiveness in platelet inhibition,83 but data are lacking.
The benefits of profound inhibition of platelet reactivity should be weighed against the increased risk of bleeding due to possible traumatic injury related to chest compression and possible head injury, as well as to the unpredictable effect of these agents in the presence of acute metabolic disturbance, which is usually observed in OOHCA survivors.
When is ventricular support required?
Cardiogenic shock is present in 30–40% of OOHCA survivors, developing most often within 4–6 h of the index event, due to the combined effect of the cardiac insult itself, with subsequent global myocardial hypoxia sustained during the period of resuscitation. LV support may be required as a bridge to definitive revascularisation, supporting the stunned myocardium and improving peripheral perfusion. Nevertheless, as with all cardiogenic shock trials, irrespective of preceding arrest, no device has yet been shown definitively to alter prognosis.
A number of support devices are available for use (figure 3). Among these, only the intra-aortic balloon pump (IABP) has been investigated in this clinical setting. Its feasibility and safety have been demonstrated in comatose survivors of OOHCA with cardiogenic shock,41 but no prognostic benefit was seen in the SHOCK II IABP trial; a randomised controlled trial of IABP over standard medical care, in which around 40% of enrolled patients were resuscitated OOHCA survivors.84
Newer devices include extracorporeal membrane oxygenation, Impella and TandemHeart devices. These all have unique technical features, physiological effects and potential hazards, as summarised in table 3. At the current time, since outcome studies are lacking, outside of a clinical trial it is suggested that they only be considered in conscious patients on admission, or comatose patients with no severe pre-arrest comorbidities and a high likelihood of survival with favourable neurological outcome.
Is there still a role for hypothermia?
Evidence from the literature
It is known that transient interruption of cerebral blood flow for minutes during cardiac arrest enables free radical generation during reperfusion, activation of deleterious enzyme cascade pathways and, ultimately, cerebral injury.85 Cooling is considered beneficial at the molecular level, by attenuating cell membrane destruction, reducing calcium release and the deleterious formation of reactive oxygen species (figure 4); this occurs due to a slowing down of cellular metabolism at a rate of 6–7% for each 1°C decrease in temperature.86 ,87 The cascade begins within minutes of cardiac arrest, so early interventions to help promote neuroprotection are vital if a meaningful neurological recovery is to be achieved.
The evidence base, however, is controversial. In 2002, two randomised controlled trials demonstrated that induction of mild hypothermia for 12 or 24 h increased survival and improved neurological outcomes in OOHCA patients with an SHR.88 ,89 Following this, hypothermia gained widespread use and was enthusiastically included in international guidelines,27 with the suggestion of favourable results, even when applied to less selected cohorts of patients than those in the original trials.90 Furthermore, even when considering a more detailed picture of OOHCA survivors’ neurological outcome, patients treated with hypothermia showed a better cognitive recovery after OOHCA, and, if employed at the time of the cardiac arrest, most could return to their previous employment.91
A recent study by Nielsen et al92 has questioned the precise strategy of induced hypothermia. These authors randomised 939 patients with ROSC after CPR to targeted temperature management at either 33 or 36°C. Somewhat surprisingly, survival (51%) and good neurological outcome (47–48%) did not differ significantly between the groups. Furthermore, post hoc analysis has demonstrated targeted management to 33°C was associated with haemodynamic alterations with decreased heart rate, elevated levels of lactate and greater vasopressor support, all independently associated with increased mortality.93 The precise reason for these results is unclear,94 but it is acknowledged that hypothermia itself (or the depth and extent of cooling) may not be the ultimate goal—rather, the prevention of hyperthermia in patients who have widespread systemic inflammatory activation.
At the present time, until more defined studies can address this issue, strict hypothermic targets are no longer recommended beyond the more general goal of maintaining body temperature at or below 36°.
Technical aspects: who should be cooled and how?
The European Resuscitation Council Guidelines still recommend hypothermia for all comatose OOHCA survivors regardless of the initial rhythm, although they acknowledge the lesser evidence base for patients presenting with non-SHR.95 There are four stages of hypothermia: (1) initiation, (2) maintenance, (3) rewarming and (4) return to normothermia.96 Hypothermia should be initiated as soon as possible since a 20% increase in mortality has been described for every hour of delay.97 Cooled intravenous saline will decrease body temperature by 1°C within 30 min,54 help to prevent post-resuscitation hypotension and can be delivered simply by first responders or emergency department personnel. Ice bags and cooling blankets are simple and effective, but are difficult to titrate to target temperature, whereas temperature-regulated surface and endovascular devices allow easier temperature control during the maintenance phase and prevent rapid temperature changes during rewarming.98
Contraindications and adverse effects of cooling
Hypothermia is generally considered contraindicated in intracranial or severe extracranial haemorrhage, refractory hypotension, severe sepsis and pregnancy. As always, the risk of any intervention should be balanced against the expected prognosis and neurological outlook of the patient. The literature is currently not adequate enough to guide clinical practice or allow guidelines or protocols—each patient should be managed according to the specific clinical scenario.
Early post-resuscitation status and ongoing hypothermia have the potential to alter the body’s response to pharmacological intervention. These factors affect P2Y12 absorption and metabolism leading to delayed onset of action not only after clopidogrel loading99 but also with newer agents, such as prasugrel and ticagrelor;100 these demonstrate a better pharmacological profile to clopidogrel, but also appear to be sensitive to body temperature with respect to platelet inhibition. However, a recent study has shown that when compared to clopidogrel, ticagrelor provides a more rapid (within 4 h) and sustained reduction (6 days) in platelet reactivity in patients receiving therapeutic hypothermia.101
Intravenous Cangrelor may represent an attractive pharmacological option for rapid platelet inhibition in this setting, but evidence is still lacking.102 Further studies and new pharmacological options should be encouraged in this field.
What comes next? The ongoing care of the OOHCA survivor
Providing appropriate care after OOHCA remains one of the most challenging aspects of managing survivors and helping their families. Neurological assessment relies on physical examination, electroencephalography, neuroimaging, sensory stimulatory evoked potentials (SSEPs) and occasionally biomarkers. However, the positive and negative predictive values of these tests are variable and ill-defined; the presence of an early abnormality does not always indicate a poor long-term prognosis, and vice versa. However, some data suggest that two abnormal findings, such as incomplete recovery of brainstem reflexes and bilaterally absent SSEPs, have a higher specificity for poor neurological recovery.103
A recognised potential bias in observational neurological outcome studies is the fact that patients perceived at high risk of a poor outcome are not given the chance of survival and neurological recovery. Moreover, most tests are insufficiently powered to exclude an acceptable false-positive rate. Current American Heart Association guidelines recognise this dilemma and recommend that neurological prognostication should be delayed until at least 72 h after the return to normothermia.27
A recent meta-analysis by Sandroni et al104 identified a series of early predictors of poor neurological outcome in comatose patients resuscitated from OOHCA treated using therapeutic hypothermia. Studies such as this suggest that an integrated approach of clinical and EEG predictors provides the most useful prognostic information.
Patients who survive sustained ventricular tachycardia (VT) or VF in the first days of a STEMI are believed to have a similar long-term prognosis as those who did not, assuming their ventricular function is not significantly impaired.32 Therefore, current guidelines on the management of STEMI do not recommend the routine implantation of implantable cardioverter defibrillator (ICDs) in patients with OOHCA survivors with VT or VF in the context of acute infarction up to 48 h. Much of these data derive from the thrombolysis era, but a recent study enrolling 4653 consecutive PCI-treated STEMI patients has supported this in contemporary practice, with no negative influence of VT/VF in the acute phase of STEMI on the long-term prognosis.105
A previous meta-analysis of three randomised studies including a large number of OOHCA survivors showed a 27% reduction in the relative risk of dying (absolute reduction of 3.5% per year) associated with ICD implantation, and this was due almost entirely to a 50% reduction in arrhythmic death.106 Patients with an LVEF of <35% benefited significantly more from an ICD than those with a greater EF, and this is reflected by current National Institute for Health and Care Excellence guidelines for all patients with acute coronary events.107 However, since a large number of patients could fall into the ‘grey zone’ with moderate LV impairment, novel predictors of ICD effectiveness in ischaemic cardiomyopathy, such as myocardial scar extension and characterisation,108 should be taken into account in the clinical decision process.
Patients with non-ischaemic cardiomyopathies and primary ion channelopathies should be considered for pre-discharge ICD implantation, weighed up against their neurological recovery and comorbidities.
In clinical practice, only a minority receives an ICD before hospital discharge, and the decision is often based upon the chances of neurological recovery. Martinelli et al109 studied 1489 hospitalised OOHCA survivors, of whom only 58 received ICD therapy before hospital discharge; unsurprisingly, recipients were generally younger and had better cerebral function.
Management of OOHCA survivor's families: psychological and genetic counselling
OOHCA survivors often require prolonged intensive care unit admissions, hospitalisation and rehabilitation. Their prognosis and neurological recovery can be uncertain. This situation strongly affects relatives, especially if the event is unexpected and the patient is young.110 Experiences of their next of kin's cardiac arrest are described as ‘strong’, ‘unexpected’ and ‘chaotic’;111 relatives require consistent support to prevent ‘burn out’.112 Rehabilitation and discharge planning should consider not only the patient but also the families involved, through education, support and counselling in order to reduce levels of depression and anxiety.
Family involvement is often required where inherited heart diseases are the suspected cause for OOHCA. In this setting, family screening is crucial in identifying relatives at potential risk of sudden cardiac death.113 This is a complex set of issues in a very heterogenous population; the psychological implications of screening, itself, the interpretation of test results—normal or abnormal—in a population of asymptomatic relatives is still a matter of complex debate.
OOHCA remains a major cause of mortality and morbidity in the Western world. Important predictors of a favourable outcome include presenting rhythm, early life support measures and facilitated transfer to a Cardiac Arrest Centre where immediate and comprehensive assessment and treatment can be undertaken. Patients with clear ECG changes of ischaemia (certainly those with ST elevation) should be offered immediate CAG with revascularisation where appropriate. Those in whom the diagnosis remains uncertain should be assessed in a resuscitation area (the ‘ER-stop’) with senior clinical review on arrival, including bedside echocardiography.
Patient protocols/pathways have been introduced under the auspices of the post-arrest ‘bundle of care’, but it is acknowledged that randomised controlled data to support many of the current interventions are sparse, and this will form the focus of much work ahead to improve the outcome of this complex and growing patient population.
Out-of-hospital cardiac arrest (OOHCA) is a major public health issue and, despite clear improvements in first assistance and early resuscitation, clinical outcomes are still poor.
Management of OOHCA survivors to hospital requires a rapid multidisciplinary evaluation of the patient in order to understand the best diagnostic pathway to follow.
An early invasive coronary strategy for OOHCA patients with evidence of ongoing ischaemia is feasible and is associated with significantly better clinical outcome. However, in the absence of dedicated trials in this area, many unanswered questions remain, including completeness of revascularisation, the safest access choice, the optimal antiplatelet strategy and use of LV support devices.
The benefits of hypothermia in improving clinical outcomes has recently been questioned, and its use should always take into account the patient's neurological prognosis and comorbidities.
Neurological recovery after OOHCA remains highly variable and challenging to predict.
ESC Core Curriculum Section: 2.23 Sudden Cardiac Death and Resuscitation
Relevant guidelines referenced
European Resuscitation Council guidelines for resuscitation 2010. Section 1: Executive summary
2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science
ESC guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation (2012)
ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation (2011)
European Resuscitation Council guidelines for resuscitation 2010. Section 4: Adult Advanced Life Support.
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: http://heart.bmj.com/misc/education.dtl
RCP credits: Log your activity in your CPD diary online (http://www.rcplondon.ac.uk/members/CPDdiary/index.asp)—pass mark is 80%.
EBAC credits: Print out and retain the BMJ Learning certificate once you have completed the MCQs—pass mark is 60%. EBAC/ EACCME Credits can now be converted to AMA PRA Category 1 CME Credits and are recognised by all National Accreditation Authorities in Europe (http://www.ebac-cme.org/newsite/?hit=men02).
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 All authors conceived the plan for the review manuscript, wrote and edited the manuscript and designed the tables and figures.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.