Medical emergencies: atrial fibrillation and myocardial infarction


Correspondence to: J. Ball


In this, the first of two article on medical emergencies, we discuss the definitions, epidemiology, pathophysiology, acute and chronic management of atrial fibrillation and acute myocardial necrosis in the peri-operative and intensive care settings.

For the purposes of this article, we define a medical emergency as an acute pathophysiological process that without immediate medical treatment, will result in severe single, or multiple, organ injury, and/or death. In this series of two articles, we have attempted to describe the immediate management issues for a number of medical conditions. The topics covered are either very common and/or those whose management lacks a clear, evidence based guideline applicable to patients in intensive care or equivalent environments. We offer pragmatic advice based upon published evidence (however limited), reasoned thinking and experience.

Cardiovascular emergencies

Cardiovascular system failure is termed shock and can result in ischaemic injury or infarction. It is also worth noting that reperfusion injury should be considered part of the same pathophysiological continuum. In physiological terms, there are broadly six components that can fail, either singularly or in any combination. Those elements are: the circulating blood volume (preload); blood composition (viscosity and oxygen carrying capacity); heart rate (HR) and rhythm; myocardial contractility and relaxation; vascular tone (principally arterial or afterload); and the microcirculation (functional capillary density and flow rate). Regardless of the aetiology, shock requires immediate management that should be directed at all of the components affected and the underlying cause. Given the innate connectivity between these six components, a systematic approach is essential in all cases.

Specific cardiovascular emergencies 1: atrial fibrillation

Atrial fibrillation (AF) is the commonest cardiac arrhythmia. In the Western world, its incidence increases with age being prevalent in 0.7% of patients aged 55–59, rising to 17.8% in patients aged over 85 [1]. Such data underestimates the true prevalence as an unknown proportion of patients have asymptomatic and/or paroxysmal AF. The incidence of first diagnosis or new onset AF in the peri-operative period is reported as 5–10% in non-cardiac surgery [2] and 10–65% for cardiac surgery [2]. The epidemiology of AF in the intensive care unit (ICU) is discussed in detail elsewhere [3]. Overall, the proportion of these patients that suffer related shock is unknown but, anecdotally, is significant.

Atrial fibrillation is the manifestation of diverse cardiac abnormalities (electrical, structural, metabolic, neurohormonal and molecular alterations) in multiple pathological conditions both acute (acute coronary syndrome, pulmonary embolus, systemic inflammatory response syndrome/sepsis) and chronic (heart failure, hypertension, diabetes mellitus, hyperthyroidism and ageing) (Figs 1 and 2). It is thought that AF requires both a substrate and a trigger to be initiated. The dominant substrate is atrial remodelling whilst ischaemia, inflammation and catecholamine-stress are the principal triggers. AF may be maintained by re-entrant circuits or the failure of depolarisation to proceed uniformly in a 1:1 conduction ratio throughout the atrial tissue. In order for a wave of depolarisation or a re-entrant circuit to fail to self terminate, it must proceed along its course at a speed slow enough to allow depolarised tissue to repolarise (i.e. beyond the refractory period). This requires some degree of structural change in the atrium – for example dilatation or tissue fibrosis that occurs as a consequence of the chronic conditions listed above. At a functional level, fibrillatory activity limits calcium entry into the cell and promotes potassium efflux, both of which favour rapid atrial reactivation. A detailed review of the pathophysiology of AF can be found elsewhere [4, 5].

Figure 1.

 Schematic illustration of various factors involved in induction and perpetuation of AF by generating substrates and triggers for AF. Reproduced with permission from [5]. AF, atrial fibrillation; APD, action potential duration; CHF, congestive heart failure; ERP, effective refractory period; HCN, hyperpolarization-activated cAMP-gated non-selective cation channel; ICaL, L-type Ca2+ current; If, funny current; IK1, inward rectifier K+ current; INCX, Na+/Ca2+ exchanger current.

Figure 2.

 Presence of atrial fibrillation (AF) in different stages of the cardiovascular disease continuum. Reproduced and modified with permission from [27]. CAD, coronary artery disease; LV, left ventricle.

Atrial fibrillation in context

Newly diagnosed or new onset AF, and poorly controlled (fast) paroxysmal or chronic AF in the peri-operative period or in the context of an acute severe illness, are associated with significant short and long term morbidity, prolonged hospital stay and excess mortality. In short, AF warrants timely attention and a systematic approach. AF can be either the precipitant, or the consequence, of shock, the combination requiring emergency treatment. The resulting ventricular response rate is classically fast but may be slow.

Fast AF: consequences and management

As HR increases, ventricular filling time (diastole) progressively shortens, making atrial systole an increasingly important component of end-diastolic ventricular volume and hence both stroke volume and ejection fraction, and fast AF (> 110 beats min−1) may result in the dramatic reduction of both. To further exacerbate the situation, the resulting shock stimulates a catecholamine response that further drives the tachycardia. Thus, the first priority when managing fast AF is rate control, with restoration of sinus rhythm (SR) a secondary goal, although both are early and important goals. Diagnosis and timely treatment of the precipitating cause of the AF is equally important. Systematic assessment and management of all the contributory components of shock is also essential.

In addition to shock and acute or decompensated chronic heart failure, the secondary consequences of fast AF include: the provocation of focal or global myocardial ischaemia (especially in the presence of proximal coronary artery flow limitation); degeneration into lethal ventricular arrhythmias; and, if left uncontrolled (for hours to days) a reversible, tachycardia induced, cardiomyopathy [6].

We propose the following stepwise treatment plan for the management of patients with fast AF (Table 1); in the context of an acute severe illness or during the peri-operative period, we define this as, ‘a witnessed rhythm change from SR or sinus tachycardia (ST) to fast AF with a HR > 110 beats.min−1’. Our approach is also applicable to patients in chronic AF with inadequate rate control.

Table 1. Management of the patient with fast atrial fibillation (AF) and shock.
Consider crystalloid bolus 3–5−1 over 5 min
If this results in >15% decrease in HR or increase in MAP, repeat; otherwise stop all i.v. maintenance fluid
Fast AF may be precipitated by hypovolaemia and by volume overload
If not fluid responsive, consider noradrenaline infusion to a target MAP of 60–70 mmHg (assuming central venous access is in place)Noradrenaline is probably the vasopressor with least side effects in this setting. In the absence of central venous access, phenylephrine (100 μg boluses titrated to response) is arguably a better choice than metaraminol but there is no clinical evidence to guide this decision
Consider rapid infusion of MgSO4 8–12 mmol (2–3 g) over 1–5 min
If poorly tolerated (>15% decrease in MAP):
 (a) give a crystalloid bolus (as above)
 (b) progressively slow the infusion rate (max. 20 min)
There is conflicting evidence regarding the efficacy of this intervention (as monotherapy) [9–11]. There is no evidence of harm.
However, in the presence of hypokalaemia (with active replacement) [8] and/or as an adjunct to any of the drug therapies listed below, magnesium therapy is probably synergistic [9]
Common side effects are hypotension, flushing, nausea and lethargy
If > 15% decrease in HR consider a continuous infusion of 4–8 mmol.h−1 (1–2 g). The optimal duration of this therapy/daily dosing regime remains unknown Simultaneously aim for a serum potassium of 4.5–5.5 mmol.l−1. This target has been extrapolated from a variety of cardiac studies and is reviewed here [7]
If the patient has been in AF for < 48 h, consider synchronised DC cardioversion (with appropriate analgesia/sedation and airway management) starting with 100 J [10]In the event of no rhythm change, rapidly deliver a second shock of the same energy as the first shock should lower the defibrillation threshold. A third shock of a higher energy, classically 200 J (ranges reported 150–360 J) can be tried but is reported as having a low rate of success.
In the event of cardioversion to sinus rhythm or sinus tachycardia, rapid reversion to AF is seen in 50–80% of published series. If this occurs, continue to optimise fluid and electrolyte status, support MAP with vasopressor therapy and commence rate controlling drug therapy, before considering further attempts at DC cardioversion
Attempt rate control with
 Esmolol: Load with 500 μ−1 over 1 min then commence infusion at 200 μ−1.min−1. This can be down titrated in response to bradycardia (HR < 70 beats.min−1). Arguably, vasopressor doses should be increased to resolve any hypotension. Doses > 200 μ−1.min−1 can be tried but reports suggest little efficacy.First choice in the shocked patient due to efficacy, (a rapid rate control to between 70–110 within 10 min in ∼ 70% of patients), relative haemodynamic stability, and a favourable pharmacokinetics (half life 9 min, rapidly and extensively metabolised via red blood cell esterases to an inactive metabolite)
 If unavailable, consider metoprolol: load with 2.5–15 mg i.v. over 3–5 min. If tolerated, follow i.v. loading with an immediate enteral dose of a suitable β-adrenoceptor antagonist. If the enteral route is unreliable or unavailable repeated dosing at 6 hourly intervals is recommended. Alternatively, a continuous i.v. infusion can be given, with a starting range of 0.4–2.5 mg.h−1, and titrated to responseThis drug is widely available, has a relatively short elimination half life of 3–4 h, and almost complete hepatic metabolism to inactive metabolites; so it is safe in acute kidney injury. There is limited and conflicting evidence to guide the choice of long term therapy, but carvedilol maybe superior to bisoprolol or metoprolol [11, 12]. Sotalol is unique amongst β-adrenoceptor antagonists in exhibiting class II and III anti-arrhythmic properties. There is some, albeit limited, evidence to suggest this action has significant clinical benefits in some settings [13–15]
 If inadequate rate control and/or unacceptable hypotension/fall in cardiac output:
 Consider continuing esmolol/metoprolol (albeit at a reduced dose) and giving a second agent
 Consider digoxin loading dose 500 μg over 30 minThis slows the ventricular rate primarily by increasing parasympathetic tone on the atrioventricular node. However, any condition associated with a high sympathetic tone easily overcomes this effect, rendering digoxin ineffective as monotherapy [16]. It is however synergistic with β-adrenoceptor antagonists (recommended as an adjunct in chronic AF) but has a narrow therapeutic concentration window and unfavourable pharmacokinetics
Consider amiodarone: load intravenously, preferably via central venous access (as it is highly irritant) with 150 mg over 15 min then 150 mg over 45 min (or 300 mg over 1 h), followed by 900 mg in 23 h. Additional bolus dosing can be attempted up to a maximum total dose of 1600 mg per 24 hAmiodarone is synergistic with β-adrenoceptor antagonists (in both the acute and chronic settings). The i.v. preparation is made in a solvent mixture of polysorbate 80 and benzyl alcohol, both of which are potent negative inotropes. Rapid IV loading may result in hypotension due do this negative inotropy and some vasodilatory effects. Treat by slowing the rate of infusion or stopping. Due to the unpredictable and very variable pharmacokinetics (volume of distribution) total loading doses in excess of 10 g over 1–3 weeks are required to ensure consistent therapeutic levels. Enteral bioavailability is 35–65% of the i.v. dose. The i.v. preparation can cause an acute hepatitis. Chronic therapy may cause pulmonary fibrosis, cirrhosis, thyroid dysfunction, skin photosensitivity and corneal deposits. As monotherapy, amiodarone achieves rate control in 50–70% of patients and only cardioverts ∼ 50% to sinus rhythm. The median time to cardioversion of AF to sinus rhythm is 7 h. In short, this should be second line/synergistic therapy. Load aggressively, switching to the enteral route (400 mg tds) at the earliest opportunity. The median elimination half-life is 48 days (range 26–107 days). As the risk and extent of end-organ toxicity seems related to cumulative dose, consider early cessation. However, this remains the most effective agent for long term rhythm control
  Alternative single or co-therapies with β-adrenoceptor antagonists
  Verapamil: i.v. bolus 1.25–10 mg over 2–10 minSynergistic with β-adrenoceptor antagonists but moderately potent negative inotrope with variable elimination half-life 0.5–6 h. Co-therapy with β-adrenoceptor antagonists can result in complete AV block (very rare). Diltiazem is probably equally efficacious but the i.v. preparation is not available in the UK
  Clonidine: i.v. infusion 1–10 μ−1.h−1Synergistic with β-adrenoceptor antagonists. Very limited evidence base [17, 18]. Infusion doses can be increased up to 25 μ−1.h−1 and slow bolus doses of 10–20 μ−1 have been described as being safe with a surprisingly low incidence of hypotension and bradycardia [17]
If, despite attempts at therapy optimisation as above, inadequate rate control and/or AF persists, consider a further attempt at synchronised DC cardioversion (as above).

When considering the management of AF in this context, it is important to consider five questions:

  • 1 Is the patient shocked with signs of end-organ hypoperfusion ± a mean arterial pressure (MAP) < 60 mmHg or ≥ 15% drop in MAP with rhythm change (Table 1)?
  • 2 Is rate related myocardial ischaemia demonstrated on the 12-lead ECG?
  • 3 Is there any evidence of pre-excitation on the 12-lead ECG, e.g. Wolff–Parkinson–White syndrome? If so, atrio-ventricular node blocking drugs should not be given as these do not block accessory pathways and may result in an increase in ventricular response rate and/or degeneration into ventricular fibrillation.
  • 4 Is there an identifiable precipitating event that warrants urgent investigation or empirical therapy?
  • 5 Is there occult (or overt) hyperthyroidism?

The place for class I agents (flecainide and propafenone) in the ICU setting is limited to patients with evidence of pre-excitation, in whom AV nodal blocking drugs (β-adrenoceptor antagonists, non-dihydropyridine calcium channel antagonists and digoxin) are contra-indicated. However, class I agents are contra-indicated in the presence of significant ischaemic heart disease or abnormal left ventricular function. In such circumstances amiodarone or sotalol should be used. The newly licensed class III anti-arrhythmic, vernakalant, should be considered an alternative to class I agents with identical limitations [19].

Newer/novel agents

Dronedarone, an analogue of amiodarone, with a better side-effect profile, may be the preferred chronic therapy in certain patients but is contra-indicated in those with New York Heart Association (NYHA) category 3/4 heart failure; hence, the decision to use it is best left to expert review after the acute illness phase [20].

A number of i.v. agents are emerging including landiolol and ibutilide but their place in the overall scheme of AF management remains to be determined. A significant number of new chronic therapies for AF are in the late stages of development.

Slow AF

Less commonly, AF may be associated with a slow ventricular response rate; this is usually the consequence of drug therapy but may also be in association with innate pacemaker or conduction abnormalities. Slow AF may be associated with chronotropic incompetence and inadequate cardiac output despite a normal stroke volume and ejection fraction. Drug therapy with positive chronotropes, synchronised DC cardioversion (see notes above) or cardiac pacing may be required to reverse shock in this setting.


AF may also predispose to intra-atrial thrombi generation with consequent embolic complications such as acute stroke, bowel and limb ischaemia. Thus in persistent AF, early consideration of the optimal anticoagulation strategy is vital. Unless contra-indicated, therapeutic dose, once daily, subcutaneous (s/c) injection of low molecular weight heparin (LMWH) is probably the optimal strategy. Consideration of immediate therapy with i.v. unfractionated heparin (UFH) 5000 IU is advocated in many guidelines before emergency electrical or chemical cardioversion, followed immediately by either an UFH infusion or treatment dose s/c injection of a LMWH.

Following a first episode of AF, either as a solitary event or in the context of an acute severe illness, the following questions should be addressed:

  • 1 Is the heart structurally normal (echocardiogram)?
  • 2 If cardioverted to SR:
    • (a) Was the AF an isolated event or merely the first recognition of paroxysmal AF (clinical history and 24 h tape)?
    • (b) If the latter, then what is the optimal strategy to maintain SR (‘pill in the pocket’ vs chronic drug therapy vs ablation procedure)?
  • 3 If in persistent AF:
    • (a) What is the optimal rate control strategy? There is conflicting evidence to answer this question, although a more ‘lenient’ target for both resting (< 110 beats.min−1) and submaximal exercise (< 140 beats.min−1) maybe associated with better quality of life. For a review of this topic see [21–23].
    • (b) Should an attempt at delayed DC cardioversion be considered?
    • (c) Should an ablation procedure be considered?
  • 4 What are the risks for and against long term oral anticoagulation? Use the CHA2DS2-VASc score (to estimate stroke risk) and the HAS-BLED score (to assess haemorrhage risk) to determine an evidence-based therapeutic strategy [19]. For a review of scoring systems see [24]. Factoring in the presence of any structural abnormalities to this process is vital [25]. Importantly, the place of novel oral anticoagulant therapy (factor Xa and direct thrombin inhibitors) over that of vitamin K antagonists in the long term management of patients with paroxysmal or chronic AF continues to evolve. For a review see [19, 26].
  • 5 What ‘upstream’ therapies, defined as treatments that can reduce any of the substrates for AF (Figs. 1 and 2), should be modified or instituted? This includes determining whether coronary artery disease is a contributory factor and developing an appropriate strategy to diagnose and treat this.

Finally, in patients at high risk of developing acute AF, there is a growing body of evidence to suggest that primary prevention is possible and results in significant improvements in clinically important endpoints, not merely the incidence of acute AF. The best studied population is patients undergoing coronary revascularisation. For reviews of this topic see [30, 31].

Specific cardiovascular emergencies 2: Myocardial infarction with or without shock


It is clearly understood that a myriad of pathologies and clinical scenarios can result in myocardial injury, only one of which is atherosclerotic, ischaemic heart disease. Accordingly, in 2007, an international task force, created a, ‘universal definition of myocardial infarction,’ [28]. The critical difference between this document and previous definitions are that five sets of diagnostic criteria were set and five types of myocardial infarction (MI) defined (Tables 2 & 3). The central tenet of the diagnosis of acute MI is the detection of myocardial cell death due to prolonged ischaemia. Myocardial cell death is categorised pathologically as coagulation and/or contraction band necrosis, not the demonstration of coronary artery flow limitation or thrombosis. The time from the onset of myocardial ischaemia to cell death, by necrosis, takes between 20 min and 4 h (or even more) depending upon the degree and persistence of ischaemia, the individual cell’s demand for oxygen and nutrients, and that cell’s resistance to supply demand imbalance (termed preconditioning) [28].

Table 2. Diagnostic criteria for acute myocardial infarction (MI), modified from [28].
  1. CABG, coronary artery bypass grafting; cTn, cardiac troponin; ECG, electrocardiogram; LBBB, left bundle branch block; PCI, percutaneous coronary intervention; URL, upper reference limit.

The term myocardial infarction should be used when there is evidence of myocardial necrosis in a clinical setting consistent with myocardial ischaemia. Under these conditions, any one of the following criteria meets the diagnosis for MI:
 1. Detection of rise and/or fall of biomarkers of myocardial necrosis (preferably cTn) with ≥ 1 value above the 99th percentile of the upper reference limit together with evidence of myocardial ischaemia with ≥ 1 of the following:
 • Symptoms consistent with myocardial ischaemia
 • ECG changes indicative of new ischaemia or infarction (Table 3)
 • Imaging evidence of new loss of viable myocardium or new regional wall-motion abnormality
 2. Sudden, unexpected cardiac death, involving cardiac arrest, often with symptoms suggestive of myocardial ischaemia, and accompanied by presumably new ST elevation, or new LBBB, and/or evidence of fresh thrombus by coronary angiography and/or at autopsy; but with death occuring before blood samples could be obtained or at a time before the appearance of cardiac biomarkers in the blood
 3. For PCI in patients with normal baseline cTn values, elevations of cardiac biomarkers above the 99th percentile of the upper reference limit are indicative of periprocedural myocardial necrosis. By convention, increases of biomarkers > 3 × 99th-centile of the upper reference limit have been designated as defining PCI-related MI. A subtype related to a documented stent thrombosis is recognised
 4. For patients with normal baseline cTn undergoing CABG with periprocedural myocardial necrosis: by convention, increases of biomarkers by > 5 × 99th-percentile upper reference limit and either new pathological Q waves or new LBBB, or angiographically documented new graft or native coronary artery occlusion, or imaging evidence of new loss of viable myocardium, have been designated as defining CABG-related MI
 5. Pathological findings of an acute MI
Table 3. Clinical classification of different types of myocardial infarction (MI), modified from [28].
  1. CABG, coronary artery bypass grafting; LBBB, left bundle branch block; PCI, percutaneous coronary intervention.

Type 1Spontaneous MI related to ischaemia due to a primary coronary event such as plaque erosion and/or rupture, fissuring, or dissection
Type 2MI secondary to ischaemia due to either increased (myocardial) oxygen demand or decreased (coronary) supply (e.g. anaemia, arrhythmias, systemic hypertension, systemic hypotension, coronary artery spasm, coronary embolism)
Type 3Sudden unexpected cardiac death, including cardiac arrest, often with symptoms suggestive of myocardial ischaemia, accompanied by presumably new ST elevation, or new LBBB, or evidence of fresh thrombus in a coronary artery by angiography and/or at autopsy; but death occurring before blood samples could be obtained or at a time before the appearance of cardiac biomarkers in the blood
Type 4aMI associated with PCI
Type 4bMI associated with stent thrombosis, as documented by angiography or at autopsy
Type 5MI associated with CABG

Of these universal definitions, it is the first set of diagnostic criteria (Table 2) together with the recognition of, and clear differences between, Type 1-‘spontaneous’ and Type 2-‘secondary’ MI, that are of major importance in the peri-operative and ICU settings (Fig. 3).

Figure 3.

 Our suggested spectrum of diagnostic entities for myocardial ischaemia, myocardial infarction and cardiac troponin (cTn) > 99th percentile in the peri-operative and ICU settings. ‘?’– reflects the common diagnostic uncertainty in differentiating between Type 2 myocardial infarction and detectable cardiac troponin not due to myocardial necrosis. NSTEMI, non-ST elevation myocardial infarction; STEMI, ST elevation myocardial infarction; MI, myocardial infarction; cTn, cardiac troponin.

Type 1-‘spontaneous’ MIs fall within the clinical diagnosis of acute coronary syndrome (ACS), within which three distinct pathological diagnoses exist (Fig. 3):

  • 1 Unstable angina – which is defined as recurrent or persistent myocardial ischaemia at rest OR with minimal exertion, due to coronary artery atherosclerotic stenosis but without myocardial necrosis. The diagnosis is based upon a clinical presentation and confirmed by specific, new ECG criteria (Table 3) together with the exclusion of alternative aetiologies for the ECG changes and a series of normal measurements of a biomarker of myocardial necrosis [29].
  • 2 Non-ST elevation MI (NSTEMI) – which is defined as myocardial necrosis as a direct result of a sudden and critical, reduction in coronary artery blood flow most commonly, but not exclusively, due to acute thrombosis, induced by a ruptured or eroded atherosclerotic coronary plaque, with or without concomitant vasoconstriction. The diagnosis is based upon a clinical presentation and specific new ECG criteria (Table 4), together with the exclusion of alternative aetiologies for the ECG changes and at least one elevated measurement of a biomarker of myocardial necrosis [29].
  • 3 ST elevation MI (STEMI) – which is defined as myocardial necrosis as a direct result of a sudden, proximal and complete occlusion of a coronary artery most commonly but not exclusively due to acute thrombosis, induced by a ruptured or eroded atherosclerotic coronary plaque.’ The diagnosis is based upon a clinical presentation and specific new ECG criteria (Table 4) together with the exclusion of alternative aetiologies for the ECG changes. Elevated measurements of a biomarker of myocardial necrosis confirm but are not required to make the diagnosis since a time interval is required for cellular necrosis and its detection in peripheral blood [30].
Table 4. ECG criteria for the diagnosis of acute myocardial ischemia that may lead to infarction (in the absence of left ventricular hypertrophy or any bundle branch block). Adapted from [28].
  1. UA, unstable angina; NSTEMI, non-ST elevation myocardial infarction; STEMI, ST elevation myocardial infarction; BBB, bundle branch block.

First/early signsIncreased T-wave amplitude (hyper-acute) with prominent symmetrical T-waves in at least two contiguous leads
Increased R-wave amplitude and width (giant R-wave with S-wave diminution, most commonly seen in leads exhibiting ST elevation and tall T-waves) representing conduction delay in the ischemic myocardium
Pseudo-normalisation of previously inverted T-waves
Classical pathognomonic changes of UA/NSTEMINew horizontal or down-sloping ST depression ≥ 0.05 mV in two contiguous leads; and/or T wave inversion ≥ 0.1 mV in two contiguous leads with prominent R-wave or R:S ratio > 1. For a review of all causes (ischaemic and non-ischaemic) and patterns of ST segment depression and T wave inversion see [58]
Classical pathognomonic changes of STEMINew ST segment elevation at the J-point, in two contiguous leads with cut-off points: ≥ 0.2 mV in men or ≥ 0.15 mV in women in leads V2–V3 and/or ≥ 0.1 mV in other leads. The classical definition of STEMI also mandates that these changes persist for > 20 min. For a review of all causes (ischaemic and non-ischaemic) and patterns of ST segment elevation see [57]
New onset left BBB
Development of pathological Q waves (indicative of infarction) [28]
NotesThe J-point (the point where the QRS complex joins the ST segment) is used to determine the magnitude of the ST elevation
J-point elevation in men decreases with increasing age; however, this is not observed in women, in whom J-point elevation is less than in men
Contiguous leads means lead groups such as anterior leads (V1–V6), inferior leads (II, III and aVF), or lateral/apical leads (I, aVL, V5–V6). Supplemental leads maybe very informative such as V3R and V4R (which reflect the free wall of the right ventricle) and V7–V9 (which reflect the inferobasal area of the left ventricle)
For further detailed discussion see [28, 57, 58]

Type 2-‘secondary’ MIs, on the other hand, are the result of a heterogeneous set of pathologies (Table 2), in which the myocardial ischaemia is not the result of a sudden coronary artery thrombosis induced by acute changes in an atherosclerotic plaque.

The necessity for this detailed set of definitions are that:

  • 1 Many common pathologies result in elevated biomarkers of myocardial necrosis (Fig. 3). However, differentiating between Type 1 MIs, Type 2 MIs and non-necrotic myocardial cellular processes [31] remains very challenging (see later section on biomarkers).
  • 2 Making this differentiation is vital, as only in Type 1 MIs, are there detailed, evidence-based management guidelines, which stress the time critical nature of successful medical and/or percutaneous interventions [29, 32]. By contrast, there is no consensus on the optimal management of Type 2 acute MI perhaps most especially as this diagnosis represents a heterogeneous set of aetiologies.
  • 3 There is a strong association between elevated biomarkers of myocardial necrosis (regardless of aetiology) and poor outcome following surgery (both cardiac [33] and non-cardiac [34]) and any acute severe illness [35–37].
  • 4In the peri-operative/ICU setting:
    • (a) Symptoms of myocardial ischaemia are rare or silent, making the detection of inducible myocardial ischaemia (angina), especially unstable angina, very difficult.
    • (b) Physiological signs that could be the result of acute myocardial ischaemia are common and non-specific, ranging from unexplained or persistent ST changes, new onset AF, decreasing PaO2/FIO2 ratio through to cardiogenic shock.
    • (c) 12-lead ECG changes consistent with new onset myocardial ischaemia (see later section on ECG and Table 4) may be caused by other pathologies or be absent [28].
    • (d) Imaging evidence of new loss of viable myocardium or new regional wall-motion abnormality requires equipment and technical expertise that is rarely rapidly available, especially out of hours.
    • (e) Acute MI, whether Type 1 or 2, and other causes of elevated biomarkers of myocardial necrosis, most commonly occur in the context of at least one other acute pathology, which may inhibit both diagnostic and therapeutic strategies or raise doubts about the risk-benefit ratio of specific interventions.


In the general population the once high incidence of STEMI is decreasing whilst the equally high incidence of NSTEMI remains unchanged or has even increased [38, 39]. Although the immediate and hospital mortality from STEMIs is marginally higher than that from NSTEMIs, four years after the event the mortality rate for patients with NSTEMI is twice that of STEMI patients [29]. Patients with ischaemic heart disease remain the focus of many large scale epidemiological studies, all of which demonstrate not only the importance of care during the index event but, critically, that the long term management is equally important and yet all too frequently appears to be suboptimal [29].

By contrast, and in part due to the diagnostic difficulties described in detail in the following sections, the incidence of Type 1 and Type 2 MI in the peri-operative/ICU setting is unknown. In our ICU, 52% of patients had at least one elevated level of a biomarker of myocardial necrosis during their stay [37]. We do not know the proportion of these that were the result of Type 1 MI, Type 2 MI or non-MI related causes.

Biomarkers in the diagnosis and prognosis of myocardial injury

Two separate groups of biomarkers are used and continue to be the focus of investigation, diagnosis, risk stratification, and outcome prediction for patients with acute MI: biomarkers of myocardial necrosis; and a second heterogeneous group of biomarkers including markers of myocardial stress or dysfunction, markers of neurohormonal activation, and markers of systemic inflammation.

Biomarkers of myocardial necrosis

The pivotal role of these markers in the diagnosis of MI is detailed above. There are four established markers of myocardial necrosis; myoglobin; creatine kinase myocardial band; and the cardiac troponins (cTn) T and I; detailed review of their biology, biochemical assays and clinical kinetics can be found elsewhere [31]. Of these, the cTns are considered superior due to their specificity and the development of widely available, reliable, and highly sensitive assays [40]. Cardiac troponin I (cTnI) is considered by some to be superior to cardiac troponin T (cTnT), due to its more rapid clearance from blood (4–7 days versus 10–14 days) although for both this is at least, in part, dependent upon renal function (see below).

Abnormal levels of cTn are defined as those above the 99th percentile of the normal population or upper reference limit with assays expected to achieve a high analytical precision, defined as a coefficient of variation of <10% [40].

Although elevated levels of cTn classically indicate cardiac myocyte necrosis, several other cellular processes including apoptosis, cellular release of cTn degradation products and increased cell wall permeability (following stretch or ischaemia) result in elevated plasma levels [31]. Consequently, a multitude of non-thrombotic and non-ischaemic causes of cTn elevation have been described, including direct myocardial injury from trauma or electrical cardioversion, myocardial wall stretch (pulmonary embolus, acute left ventricular failure, fluid resuscitation, positive pressure ventilation), exposure to toxins, local infection or inflammation (myocarditis), infiltrative disorders, sepsis, renal failure (acute and chronic), acute brain injury, exercise [41] and even skeletal muscle inflammation/injury [42].

The kinetics of both the rise and fall of plasma cTns may indicate a diagnostic pattern; thus serial cTn measurements are valuable as a rapid rise (> 50% increase in cTnI or > 100% increase in cTnT from baseline over 6–24 h), a high peak value (> 50 times the upper reference limit) and rapid fall (> 50% in 48–72 h) are indicative of acute MI [31, 40] with both the peak value and area under the cTn versus time curve indicative of the infarct size [43] and the long term prognosis [31].

Perhaps the commonest cause of falsely negative cTnT and falsely elevated cTnI is haemolysis [31]. In severe renal failure, both acute and chronic, cTn levels above the upper reference limit are a common finding and are thought to represent, to variable degrees, both impaired clearances of cTn degradation products and diffuse myocardial injury. However, to attribute such findings solely to the renal pathology mandates serial cTn measurements and the demonstration of a largely static cTn level against time [31]. As any level of cTn above the upper reference limit is associated with an increased mortality, regardless of the cause [37], and serial measurements contribute materially to differentiating the cause, we advocate routine daily measurement in all ICU patients. In those with levels above the upper reference limit, we advocate a proportionate diagnostic workup, starting with a 12-lead ECG. We have a low threshold for urgent (preferably, same day) bedside echo. We strive to differentiate Type 1 from Type 2 acute MI, or conclude a non-MI related cause. The presence of any patient factors that contra-indicate medical and/or percutaneous interventions (discussed below) guides our decision making process. We believe that it is essential to confirm or refute the diagnosis of Type 1 MI, as this should influence the long term medical management and prognosis of the patient if they survive their ICU admission.

Biomarkers that aid in the diagnosis and prognosis of ACS

The list of biomarkers that have been investigated to aid in the diagnosis and prognosis of ACS is very long with only a few emerging as clinically useful [31, 44]. The only biomarkers to emerge as useful from this vast body of work are B-type natriuretic peptide (BNP) and its precursor, N-terminal pro-B-type natriuretic peptide (NT-proBNP). They are markers of myocardial stress or dysfunction and are mainly produced by ventricular myocardium in response to increased wall stretch and volume overload, regardless of aetiology. Importantly, BNP is cleared mainly from the circulation by the natriuretic peptide C receptor and degraded by neutral endopeptidase, whereas NT-proBNP is cleared by the kidneys. Therefore, NT-proBNP concentrations inversely correlate with the glomerular filtration rate and increase with age. The half-life of BNP is only 22 min, whereas the half-life of NT-proBNP is much longer at approximately 120 min (with a normal glomerular filtration rate). In vitro, BNP is less stable than NT-proBNP if blood is not collected in plastic tubes containing EDTA as an anticoagulant [45].

In patients with suspected ACS, the addition of BNP and/or NT-proBNP measurements to routine investigations adds incremental diagnostic and prognostic value, principally the latter [31, 44, 46]. However, in a general ICU population these biomarkers have not been shown to have much value [47]. Despite these reports, several recent studies have successfully exploited the short half-life of these markers and made serial measurements over shot periods to assess response to interventions [48–50].

In summary, in peri-operative and ICU patient populations, serial BNP/NT-proBNP measurements in the context of a newly elevated or rapidly rising cTn may add both diagnostic and prognostic value and warrants further investigation. These markers may also have a role in the titration of heart failure and cardiogenic shock therapy [51, 52].

In patients presenting with ACS, C-reactive protein (CRP) demonstrates reliable, long term, prognostic information, especially if measured using newer, highly sensitive assays. However, in the context of the peri-operative or ICU patient, any cardiac cause of elevated CRP is indistinguishable from elevation due to other pathologies. Similarly, stress hyperglycaemia, renal impairment and anaemia have all been demonstrated to be sensitive and reliable prognostic markers of long term outcome in patients with primary ACS, but their utility in peri-operative and ICU patients with secondary ACS is likely to be none.

Electrocardiogram considerations

Clinicians may be alerted to the possibility of an acute MI by a variety of non-specific physiological signs ranging from unexplained or persistent ST changes, new onset AF, decreasing PaO2/FIO2 ratio through to cardiogenic shock. In theatre and ICU, continuous ECG monitoring for arrhythmia is routine as is 12-lead ECG for screening for signs of acute ischaemia or infarction (Table 4). However, despite being advocated in international, expert opinion based guidelines for many years [53], surveillance with automated, continuous ST segment monitoring remains limited [54]. The reasons are multifactorial but a lack of evidence of benefit and uncertainty regarding the specificity and implications of detected myocardial ischaemia are paramount. Furthermore, the optimal technique is debated [53, 55]. What is startling is the lack of data surrounding this topic. It is clear from continuous ST segment monitoring that episodic myocardial ischaemia is very prevalent in ICU patients and largely ignored [56]. Periodic 12-lead ECG misses the vast majority of transient ischaemic events and is additionally confounded by difficulties with interpretation; a trace unchanged from baseline does not exclude myocardial ischaemia or infarction, and ST segment or T wave changes suggestive of ischaemia or infarction occur in their absence due to other causes [57, 58]. In summary, in the peri-operative and ICU settings there is a reasonable case for automated, continuous ST segment monitoring using 6 to 12-lead ECG, but the positive and negative predictive values of detecting changes for myocardial ischaemia and infarction are undetermined and no consensus exists on optimal management.

Non-invasive cardiac imaging

Echocardiography offers a rapid, non-invasive, bedside technique to perform a detailed, quantitative study of cardiac structure and function. It has a clearly defined role in the diagnosis of acute myocardial ischaemia, infarction and shock [59]. In the context of an anaesthetised or ICU patient with ECG changes and/or elevated cardiac biomarkers suggestive of acute myocardial ischaemia or infarction, the presence of ventricular regional wall motion abnormalities is diagnostic, with very high positive and negative predictive value. However, the timing of the myocardial event cannot be accurately determined, especially if the patient has evidence of pre-existing ischaemic heart disease and in the absence of a recent echo (or equivalent imaging) that confirms that the regional wall motion abnormalities are new. To compound this further, detection of regional wall motion abnormalities requires considerable skill and experience. Currently, there is very limited provision of round-the-clock, rapidly available diagnostic echo services. Indeed, this is but one compelling reason for the development of anaesthetist and intensivist delivered echo services.

The sensitivity and specificity of echo in the ICU setting can be enhanced by repeated scanning, the use of echo contrast (for endocardial definition) and, potentially, by employing a pharmacological stress protocol.

Other potentially useful, non-invasive, imaging techniques include methods to detect coronary stenoses (CT and MRI coronary angiography) and methods that detect coronary flow limitation (pharmacological stress radioisotope and MRI studies) [60]. However, none of these, with perhaps the exception of CT, is practical in critically ill patients. Indeed, recent advances in CT imaging may offer additional information including regional wall motion abnormalities and perfusion defects at rest [61], albeit at the cost of a significant radiation dose.

Risk stratification

Patients with suspected or proven ACS must be amongst the most studied of any patient population. The range and complexity of available medical and percutaneous therapies has become bewildering, especially to the non-expert. Such therapeutic complexity and the necessity of ever more detailed patient selection for ongoing and future clinical trials has led to the development of a series of risk stratification tools. The two most widely used are ‘The Global Registry of Acute Coronary Events (GRACE) score’ (see and, ‘The Thrombolysis In Myocardial Infarction (TIMI)’ score (see A recent meta-analysis of these scores concludes that GRACE is the superior tool [62]. Both of these scores classify patients with classical ACS presentations into low, intermediate and high risk groups for both re-infarction and death. Not only do these classifications give reliable probabilities of future events, they are increasingly being used in international guidelines to direct therapeutic decision making. To our knowledge, there is not a single published study assessing the validity or utility of any of these scores in the peri-operative or ICU populations, in whom ACS is a secondary complicating pathology. This would appear to be an important study to undertake. In the interim, use of the GRACE score in such patients with suspected or proven ACS seems reasonable.

As with AF, the primary risk from aggressive interventions in patients with ACS is bleeding, most especially as this complication significantly worsens prognosis. In ACS the bleeding risk can be assessed using the CRUSADE score (see Again, this score is now being used in international guidelines to determine whether therapies recommended in patients at high risk of poor outcome are contra-indicated, as the bleeding risk outweighs the estimate benefit of the therapy. As this score currently represents the best estimate of therapeutic risk it should be used in peri-operative and ICU patients, despite the lack of validation in this patient population.

In summary, use of the GRACE and CRUSADE scores offers the best objective method for anaesthetists and intensivists to use in deciding on treatment strategies and in negotiating with cardiologists. Prospective studies using these scores in our patient population are needed.

Acute therapies and therapeutic controversies

What follows is a selective distillation of the current international guidelines [29, 32, 63], which we have attempted to put in the context of the peri-operative or ICU patient with a secondary diagnosis of suspected or proven ACS. In the first instance, the long established therapy with the acronym MONA (morphine, oxygen, nitrates and aspirin) should still be applied, albeit intelligently.

First aid

In the majority of cases (see definition above), it is likely that systemic opioids are already in use and therefore an assessment and, if necessary, intervention to achieve effective analgesia is all that is required.

Oxygen therapy, in particular hyperoxia, is commonly and perhaps thoughtlessly applied in peri-operative and ICU patient populations. There is some evidence to support the benefit of hyperoxia following brain injury [64, 65] but there is increasing evidence of harm, both global [66] and to specific organs [67], most especially its association with coronary vasoconstriction [68, 69]. The current balance of evidence, as assessed by the expert task force of the European Society of Cardiology, favours modest hypoxia over hyperoxia [29]; thus it seems reasonable to target oxygen saturations of 92–95% in any patient with a suspected ACS.

Anti-ischaemic therapies

Nitrate therapy relieves anginal symptoms but has never been demonstrated (and barely investigated) to reduce infarct size or improve outcomes in ACS, although logically it should. Nitrates should be considered second-line therapy (after the use of beta adrenoceptor blocking drugs and/or non-dihydropyridine calcium channel blockade) in any peri-operative or ICU patient with suspected or proven ACS who manifests hypertension (defined as a systolic pressure of >140 mmHg). It is most effective if given as an intravenous infusion (0.5–12 mg.h−1) with a target systolic blood pressure of 90–120 mmHg.

In such cases, early consideration of active HR and blood pressure control with beta adrenoceptor blocking drugs should be undertaken. The clear indication is the presence of tachycardia with normo- or hypertension. Contra-indications are bradycardia and cardiogenic shock. Cautious i.v. loading using metoprolol or esmolol with early administration of a regular enteral dose of metoprolol or bisoprolol is recommended. In cases of known intolerance to beta adrenoceptor blocking drugs or inadequate efficacy, substitution or addition of a non-dihydropyridine calcium channel blocker (verapamil or diltiazem) should be considered. The relevant pharmacology is discussed in the AF section above. The target HR should be 70–90 beats min−1 and systolic blood pressure 90–120 mmHg.

Though unproven (and untested), logic would suggest that these therapies (including the use of nitrates) that reduce supply demand imbalance should be considered first line therapy in Type 2 MIs.

Inhibiting the propagation and de novo formation of coronary artery thrombus with antiplatelet therapies

Aspirin therapy remains a cornerstone of ACS treatment. Aspirin, rapidly (within minutes) and irreversibly blocks the production of thromboxane A2 in circulating platelets, thereby inhibiting one of the many, platelet activation pathways [70]. This effect can only be overcome by production of new platelets (with a half-life of 4 days) or platelet transfusion. Unless the patient is actively bleeding or severely coagulopathic, enteral aspirin should be given at a loading dose of 300 mg, with a daily maintenance dose of 75 mg. There is no proven benefit in re-loading patients already receiving maintenance dose therapy. It is worth noting that any concomitant use of other non-steroidal anti-inflammatory drugs should be stopped or avoided as these may block the effects of aspirin and be pro-thrombotic [70]. In ICU patients with unreliable enteral function, aspirin can be administered rectally or intravenously. However, neither of these forms are widely available. Resistance to the antiplatelet effects of aspirin may be a clinically important phenomenon but isn’t a discreet entity, has no clear definition or reliable method of diagnosis, and doesn’t appear to be a major issue, although the case for further investigation into these phenomena is compelling [70]. Beyond the issue of ACS in ICU patients, a recently published large, retrospective, cohort study investigated the possible association between chronic low dose aspirin therapy and in-hospital mortality in patients with systemic inflammatory response syndrome or sepsis [71]. The findings are compelling in that aspirin use was associated with a dramatic reduction in mortality. Whether aspirin has a role in the therapy of these ICU syndromes will require an appropriately designed and powered, randomised control trial but, at the very least, this study [71] suggests that aspirin appears to be safe in this patient population.

The additional benefit of blocking a second of the platelet activation agonists, adenosine diphosphate (ADP), has been repeatedly demonstrated. In current practice there is a choice of three agents, clopidogrel, prasugrel and ticagrelor, all of which block the ADP PY212 receptor, with more agents in the late stages of development. However, the optimal choice of agent and dosing regimen remain unclear. Clopidogrel, for which there is by far the greatest experience, and prasugrel are prodrugs that require conversion to active metabolites in the liver by cytochrome P450 isoenzymes. In addition, the absorption of these drugs is dependent upon specific carrier proteins. Thus, both absorption and essential metabolism are subject to wide inter-individual variation as a result of common genetic polymorphisms. Furthermore, both of these processes are subject to decreased efficiency as a consequence of chronic co-morbidities, especially diabetes mellitus, and other drugs, in particular proton pump inhibitors. The consequences of these undesirable and unpredictable pharmacokinetics has resulted in the need for rapid, point of care, therapeutic drug monitoring.

Unfortunately, assessing platelet function remains somewhat of a dark art [72] leaving the clinician with a best guess and a rather complex choice algorithm [29, 32]. These are heavily influenced by two factors: the bleeding risk (assessed using the CRUSADE score); and the decision whether or not to perform an angioplasty and stent implantation, especially a drug eluting stent [63], as the latter require a commitment to a minimum of 12 months’ dual antiplatelet therapy. Pragmatically, in the peri-operative and ICU patient populations, clopidogrel should be given with an enteral loading dose of 300 mg and a daily maintenance dose of 75 mg, providing a number of contra-indications have been excluded; these include active bleeding, severe coagulopathy, a CRUSADE score ≥ ‘high risk’, < 24 h since major surgery or likely to need major surgery within the successive 5–7 days. Clinicians should be aware that clopidogrel takes a minimum of 2–4 h to take effect, is irreversible, and a single loading dose effects platelet function for up to 10 days. Should problematic bleeding occur, the drug will affect any platelets transfused. Ticagrelor, by contrast, requires no metabolism, is active in 30 min and is reversible but is associated with a significantly higher bleeding risk than clopidogrel. Prasugrel, despite its potential susceptibility to unpredictable pharmacokinetics, appears to be the most potent of the three drugs, is irreversible, has the longest duration of effect and, unsurprisingly, the highest bleeding risk. Following an index event (with or without stent implantation) a minimum of 12 months’ therapy is recommended. Returning to the issue of proton pump inhibitors, the clinical significance of the drug interaction remains unclear, with advocates of both ‘avoid if possible’ and ‘use as routine prophylaxis against gastro-intestinal bleeding’. Finally, the cardiovascular risk of stopping a PY212 inhibitor within weeks of an index event, in order to reduce the risk of peri-operative bleeding or manage bleeding complications, is clinically significant. This cardiovascular risk is manifestly higher if a coronary stent, either bare metal or drug eluting, has been implanted, with the risks being significantly highly for the latter. Investigations into optimal bridging therapy are being conducted but, given the complexities, guidelines suggest seeking expert advice.

The final antiplatelet therapy that can be considered are the glycoprotein IIb/IIIa receptor inhibitors. There are three chemically distinct agents available: abciximab; eptifibatide; and tirofiban. They are all intravenous agents with relative short durations of action (2–4 h). Their use is currently only recommended in patients undergoing emergency percutaneous intervention although they may have a role in high risk patients (as assessed by the GRACE score) who exhibit signs of ongoing ischaemia and in whom dual antiplatelet therapy is unfeasible.

Inhibiting the propagation and de novo formation of coronary artery thrombus with anticoagulants

In ACS, combination antiplatelet and anticoagulant therapy is superior to either therapy alone. The choice of effective agents has become bewildering and includes an i.v. bolus then infusion of unfractionated heparin (UFH), LMWH (of which enoxaparin is the favoured agent), fondaparinux, direct factor Xa inhibitors (apixaban, rivaroxaban and otamixaban), and direct thrombin inhibitors (bivalirudin and dabigatran). On the basis of balancing efficacy and bleeding risk, fondaparinux is the favoured choice unless percutaneous coronary intervention and stent implantation are undertaken. Given as a s/c injection it is renally excreted with an elimination half-life of 17 h (assuming normal glomerular filtration rate). However, no therapeutic drug monitoring is available and its effects cannot be reversed, hence in the peri-operative and ICUs setting it is not the safest choice.

Although there is vast experience of using both UFH and LMWHs in the peri-operative and ICU settings, bivalirudin (given as an i.v. bolus then infusion) has more predictable pharmacokinetics and dynamics, can be monitored using standard coagulation tests, and has a short half-life (30–60 min). Its elimination is dependent upon renal function but as it can be monitored the dose can be adjusted [73]. There is no specific reversal agent but combination therapy using blood products and recombinant factor VIIa has been successfully used [74].

The optimal duration of anticoagulation therapy remains debatable and depends primarily upon whether percutaneous coronary intervention is undertaken or not. Guidelines suggest that anticoagulation therapy can be stopped a matter of hours after successful percutaneous coronary intervention but should otherwise be continued for 2–8 days.

Indications for diagnostic coronary angiography with intent to perform immediate revascularisation

The indications for emergency percutaneous coronary intervention are refractory angina or haemodynamic or electrical instability (attributable to probable ACS). The only absolute contra-indication is active bleeding or high risk of bleeding, such that antiplatelet and anticoagulation therapy cannot safely be given immediately and continued for months (antiplatelet therapy alone). However, in the peri-operative or ICU patient, ACS of this severity has to be considered in the wider context of the patient’s condition. In such circumstances, the risks of intervention are higher and the benefits smaller. It is worth re-iterating that making the diagnosis of ACS in such patients remains challenging.

In any patient who remains shocked despite fluid and vasopressor therapy, assessment of cardiac output is essential. Confirmation that there is a significant cardiogenic component to the shocked state should set in train a diagnostic plan whilst inotropic support is initiated. The principal diagnoses to consider are ACS, myocarditis (including septic myocardial depression), Takotsubo or stress cardiomyopathy, acutely decompensated chronic heart disease (ischaemic, hypertensive, valvular, cardiomyopathies), acute right ventricular failure due to acute pulmonary pathology, and traumatic myocardial contusion. If the clinical suspicion supports or cannot reasonably exclude the possibility of ACS as the aetiology of cardiogenic shock, then there is a case to be made for purely diagnostic emergency coronary angiography. Computer tomography coronary angiography may be a reasonable alternative.

A detailed discussion of the optimal pharmacological and mechanical supportive therapies for the various aetiologies of cardiogenic shock is beyond the scope of this article but has been recently reviewed [75, 76]. Much debate and surprisingly little evidence surrounds these issues. Intuitively, mechanical support for the failing pump should be better than stimulant drugs. The early use of intra-aortic balloon pumps is widely practiced and encouraged by international guidelines but their efficacy is by no means universal, and complications not uncommon. The future may lie in alternative approaches from therapeutic hypothermia to extra-corporeal cardiac or cardiopulmonary support.

The haemodynamically and electrically stable peri-operative or ICU patient with a suspected or proven ACS should undergo objective risk stratification [29] to help determine the potential benefits of early (within 12–24 h), intermediate (within 72 h), or deferred (pending non-invasive imaging and/or stress testing) percutaneous coronary intervention. It is worth noting from registry data that only a minority (30–40%) of patients who present with an ACS have single vessel disease. Of the remainder, probably only a minority have an obvious culprit lesion amongst these multiple stenoses. Despite impressive and ongoing technical developments in percutaneous coronary intervention, there is still a role for emergency coronary artery bypass surgery.

With the advent of primary percutaneous coronary intervention, ‘straight to catheter lab’ protocols and the seemingly limitless series of large scale percutaneous coronary intervention related trials, there is growing evidence that earlier interventions are beneficial. In hospitals with round-the-clock interventional cardiology, there is anecdotal evidence of a slowly increasing enthusiasm, or lowering of thresholds, to intervene in ICU patients with suspected ACS. However, there is a widespread perception that inpatients in general and ICU patients in particular are the poor cousins to patients presenting with ACS.

Other important considerations and therapies

The early and aggressive use of lipid lowering therapies, in particular statins, is well established in the management of ACS. The evidence for this comes principally from a single trial [77] and its substudies [78]. Indeed, it maybe the anti-inflammatory/pleiotropic effects of statins rather than their lipid lower effects that are important. Caution, with regard to hepatitis, rhadbomyolysis and drug interactions [79], is required during high dose therapy.

Both hyperglycaemia and high blood glucose variability are markers of severity of illness and associated with a worse outcome in patients with ACS [80]. Glycaemic control is essential but the balance of evidence favours a less ‘tight’ and more liberal target range of 5–11 mmol.l−1, with insulin therapy the preferred agent in the acute setting.

Blood pressure permitting, early and prolonged therapy (months) with angiotensin converting enzyme inhibitors (ACE-I) or angiotensin receptor blockers (ARB) is beneficial as these agents, in concert with beta adrenoceptor antagonists, promote successful cardiac remodelling. However, the risks of polypharmacy, especially in older patients, should be considered in the overall risk benefit equation. The adjunctive role of diuretics and aldosterone antagonists depends largely upon the extent of myocardial injury and systolic failure. It should be remembered that the guidelines stress the importance of up-titrating these drugs to the maximum tolerated dose or at least to clearly defined HR and blood pressure targets.

Anaemia is a potent risk factor for poor outcomes in ACS, including the risk of bleeding [81]. On the other hand, transfusion of packed red blood cells is not without risks [82]. Indeed, this paradox is apparent across all acute severe illnesses. In patients with ACS, the suggested threshold for packed red blood cell transfusion is a haemotcrit of < 25% or a haemoglobin concentration < 7 g.dl−1.

Surviving an ACS in the peri-operative or ICU setting should be considered the start, not the end, of a disease process that may require ongoing investigations, active optimisation of medical therapy, secondary preventative measures and cardiac rehabilitation. Failure to engage and invest in these interventions entirely negates the time, effort and cost of early and aggressive therapies.

Competing interests

No external funding or competing interests declared.