Echocardiography in intensive care

Published on 27/02/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2665 times

Chapter 23 Echocardiography in intensive care

Echocardiography refers to a group of interrelated ultrasound applications used to examine the heart and great vessels. The reflected and processed sonic waves are displayed on a monitor and the images can be stored on videotape or disk. Echocardiography includes: (1) two-dimensional (2-D) anatomical imaging; (2) M-mode echocardiography, usually obtained with 2-D guidance; and (3) Doppler techniques. Echocardiography is a safe non-invasive technique which is integral to clinical cardiology. More recently, echocardiography has evolved into a powerful diagnostic and management tool in critically ill patients, especially in cardiovascular emergencies of uncertain cause.^1,²

Cardiologists, although highly skilled in echocardiography, may not appreciate the complex pathophysiology of critically ill patients in intensive care, and they have time commitments elsewhere. Echocardiography is operator-dependent; optimal image acquisition requires both technical knowledge of the machine’s capabilities as well as a degree of manual dexterity. As a general rule, echocardiography is not an efficient technique for monitoring haemodynamic indices over the medium term (hours–days) in the intensive care unit (ICU). Echocardiography can help determine whether and when continuous haemodynamic monitoring should be commenced.³ Repetitive trend measurements, such as stroke volume, cardiac output and pulmonary artery wedge pressure, are more suited to the pulmonary artery catheter. Moreover, intensive care staff are skilled and experienced in its use. Echocardiography technicians are not present in ICU, especially out of hours; intensivists must be able to act as their own sonographers as well as interpreting and reporting on the echographic images obtained. It is important that intensivists acquire all requisite skills so that echocardiography becomes more widely available.

PRINCIPLES AND TECHNICAL CONSIDERATIONS ⁴

Echocardiography uses the reflection of sound waves at tissue boundaries to construct a 2-D image of cardiac structures. The human ear hears only sound waves with a frequency between 20 Hz and 20 kHz; higher frequencies are referred to as ultrasound. Echocardiography uses sound in the frequency 1–10 MHz. A piezo electrical crystal generates and receives the ultrasound waves.

Acoustic window	Image plane
Transthoracic
Parasternal	Long axis
	Short axis
Apical	Four-chamber
	Two-chamber
	Long axis
Subcostal	Multiple
Transoesophageal*
Transgastric	e.g. Short axis
	Long axis
Deep transgastric	e.g. Long axis/five-chamber
Lower transoesophageal	e.g. Four-chamber
	Two-chamber
	Long axis
Upper transoesophageal	e.g. Short axis of aortic valve
	Long axis of aortic valve
	Right ventricular inflow–outflow

M-MODE ECHOCARDIOGRAPHY

Rapidly repeated transmit and receive cycles allow production of images with good resolution. M-mode echocardiography complements 2-D echocardiography. The beam is not scanned but stays in a fixed position, allowing the display of structures along the beam as a function of time. The rapid sampling rate makes identification of thin moving structures, such as valve leaflets, relatively easy. The cursor is usually directed through a 2-D image, giving a one-dimensional slice or ‘icepick’ view. Distance (depth) is on the vertical and time is on the horizontal axis (Figure 23.1).

Figure 23.1 M-mode echocardiogram guided by M-mode cursor on the two-dimensional (2-D) image (see top). The transducer is anterior with the cursor directed through the right ventricle (RV), intraventricular septum (IVS), mitral valve (MV) and posterior wall (PW) of the left ventricle. The mitral valve leaflets are thickened and there is reduced diastolic slope of the anterior leaflet and the posterior leaflet moves anteriorly during diastole due to commissural fusion.

DOPPLER ECHOCARDIOGRAPHY ^4,⁵

Doppler echocardiography is vital for obtaining haemodynamic information and is an integral part of every echocardiographic study.

The Doppler effect is based on changes in sound frequency that occur when a sound source moves towards or away from an observer. One classic example is that of an ambulance siren; as it comes toward a listener, the sound frequency increases (higher pitch) and after the ambulance moves away from the observer the sound frequency decreases (lower pitch).

The Doppler shift is the difference between frequencies that are transmitted and received by a transducer after striking red blood cells.

The Doppler equation is the mathematical relationship between the Doppler shift and the velocity of red blood cells that produce it.

where V = velocity of blood flow, C = speed of sound in soft tissue (1540 m/s), Δf = Doppler (frequency) shift:difference in frequency between received (f_r) and transmitted (f_t) ultrasound and θ = angle between ultrasound beam and direction of blood flow (Figure 23.2). Echocardiographic machines contain computers that automatically calculate the Doppler shift which is then entered into the Doppler equation. Blood flow velocity (m/s) is calculated and displayed on the monitor.

Figure 23.2 The Doppler equation.

It is very important that the ultrasound beam is parallel or nearly parallel with the direction of blood flow. Ifθ equals 0 then cosign θ equals 1; however, if θ > 20° then the velocity of blood flow will be significantly underestimated. Doppler data are processed and a spectral display plots instantaneous blood velocities over time. Blood flow velocity can be expressed as peak velocity or mean velocity throughout a cardiac cycle (velocity–time integral: VTI) (Figure 23.3).

Figure 23.3 Velocity–time integral (VTI: cm/s) = area under the velocity curve: sum of velocities (cm/s) during the ejection time (s); t₁ = time one, t₂ = time two, etc.

The most common uses of Doppler are pulsed-wave (PW), continuous-wave (CW) and colour-flow Doppler (CFD).

THREE-DIMENSIONAL

Three-dimensional (3-D) echocardiography is an evolving technique. Current equipment provides ‘real-time’ 3-D images (Figure 23.4a). The acquisition of a complete volume data set typically requires four sequential cardiac cycles. This 3-D data set must then be viewed and analysed. The full 3-D volume can then be ‘cropped’ so that the interior is exposed (Figure 23.4b). The 3-D volume set can be ‘opened’ in any imaging plane, providing both long- and short-axis views of cardiac structures. The 3-D volumetric probe can also be used for simultaneous acquisition of true real-time biplane (Figure 23.5a) and triplane (Figure 23.5b) images.

Figure 23.4 (a) Real-time volumetric scan (four cardiac cycles) merged into three-dimensional (3-D) volume data set. This large 3-D volume is not identifiable as a cardiac structure. (b) The full 3-D volume has been ‘cropped’, revealing the interior right ventricle, tricuspid valve, right atrium, left ventricle, mitral valve and left atrium.

Figure 23.5 (a) True real-time biplane (apical four-chamber and two-chamber views), and (b) triplane (apical four-chamber, two-chamber and apical long-axis views). Note: images obtained without moving the probe.

SAFETY

Ultrasound at the power levels used clinically to image cardiac structures has no known adverse biologic effects. Transoesophageal echocardiography (TOE) in trained, experienced hands is a well-developed and safe procedure. Nevertheless, TOE is semi-invasive and serious complications and even deaths have occurred. TOE is contraindicated in the presence of oesophageal stricture or neoplasm. Caution should be exercised in other situations such as previous oesophageal surgery or total gastrectomy. TOE should be deferred in patients with any symptoms of dysphagia until the cause has been found. Oesophageal varices, severe coagulopathy and significant cervical spine abnormalities are relative contraindications to TOE. Routine antibiotic prophylaxis does not appear to be necessary before TOE in intensive care patients.

TRANSTHORACIC (TTE) AND TRANSOESOPHAGEAL ECHOCARDIOGRAPHY

TTE and TOE are complementary ultrasound techniques, each with its own strengths and weakness (Table 23.2). For example, a vegetation imaged on the mitral valve of a septic patient will usually require a TOE to evaluate complications of endocarditis.

Table 23.2 Transthoracic echocardiography (TTE) and transoesophageal echocardiography (TOE): advantages and disadvantages

TRANSTHORACIC ECHOCARDIOGRAPHY

Recent technological advances such as harmonic imaging have greatly improved transthoracic surface imaging. TTE images may be inadequate in some critically ill patients, especially in intubated mechanically ventilated patients, or patients with chronic lung disease, obesity, chest trauma or surgical wounds. Inability to position such patients appropriately is a further limiting factor.

TRANSOESOPHAGEAL ECHOCARDIOGRAPHY ^6,⁷

The transducer is mounted at the tip of a flexible gastroscope-like probe which can be manoeuvred to various positions in the oesophagus and stomach close to the heart (Figure 23.6).

Figure 23.6 Transducer position in transoesophageal echocardiography.

The modern multiplane probes allow rotation of the ultrasound scanning plane from 0 to 180°, which provides a mirror-image orientation. Multiple tomographic imaging planes can be obtained without the necessity for further probe manipulation. Compared to TTE, TOE provides additional information in 32–100% of intensive care examinations and unexpected new diagnoses in 38–59%, leading to significant changes in treatment.⁸ Approximately 20% of patients with unexplained hypotension require surgery on the basis of new TOE findings.^9,¹⁰

INDICATIONS FOR ECHOCARDIOGRAPHY ^11,¹²

The indications for echocardiography broadly fall into three categories:

1 morphologic diagnosis of specific cardiac/great-vessel pathology (usually guided by clinical suspicion)

2 haemodynamic assessment and monitoring of cardiac function

3 others, such as evaluation of patients with atrial fibrillation for the presence of left atrial (LA) clot prior to cardioversion

Guidelines for the use of echocardiography are based on expert consensus and observational studies. There are no randomised trials that specifically test the impact of echocardiography on outcome. In intensive care practice, echocardiography is used to evaluate clinical syndromes that suggest a diagnosis or, more often, a number of possible diagnoses. For example, haemodynamic instability in a patient with major blunt chest trauma suggests cardiac contusion or pericardial effusion causing tamponade. Nevertheless, the echocardiographic examination might reveal hypovolaemia with hyperdynamic LV systolic function or valvular damage with severe regurgitation.

Some common syndromes in which echocardiography is especially useful are listed in Table 23.3.

Table 23.3 Indications for echocardiography in the intensive care unit according to clinical syndrome

Clinical syndrome	Findings	Comments
Hypotension
Acute myocardial infarct
No murmur	LV RWMA(s)	Usually severe ↓ LV systolic function
	RV RWMA	RV ↓ > LV ↓
	Hypovolaemia	‘Empty’ LV cavity, systolic function largely preserved
	Acute MR	Often no murmur with normal LA size
New murmur	Papillary muscle rupture (partial or complete)	‘Good’ LV function with severe MR
	VSD	RWMA. High-velocity left-to-right systolic jet
Rarely	Cardiac rupture/tamponade	Acute pericardial tamponade, usually fatal
	LV pseudoaneurysm	Containment of rupture
Cardiothoracic surgery	Cardiac tamponade	Often localised. There may be no ‘echo-free space’ due to clot compressing the heart. Cardiac filling pressures may be normal
	(New) RWMA	May be due to graft occlusion, air embolism
	Hypovolaemia	‘Empty’ LV with RWMA
	Global LV ↓	Stunning or long-standing cardiomyopathy
	Valvular dysfunction	Often long-standing but may be worsened by surgery, e.g. ischaemia to a papillary muscle of mitral valve
Dynamic LVOT obstruction	LVOT pressure gradient, SAM, MR	Inotropes/IABP/hypovolaemia worsen LVOT obstruction
Trauma	Hypovolaemia	‘Empty’ LV with vigorous contraction
	Cardiac contusion	RWMA (RV > LV)
	Valvular injury	Most common aortic valve (AR) or mitral valve (MR), occasionally tricuspid valve (TR)
	VSD/ASD	Occasionally
	Cardiac tamponade	More common in penetrating chest injuries
	Ruptured thoracic aorta	Widened mediastinum (90%) at isthmus of aorta. TOE required
Sepsis	Often normal LV systolic function with ‘empty’ LV	Possible global/regional LV depression
Sepsis	Infective endocarditis – vegetations/abscess/regurgitation	TOE more sensitive than TTE for imaging vegetations/abscess/fistula
‘Isolated’ hypotension (‘hypotension ?cause’)	Global LV ↓ or RWMA	Cardiomyopathy or stunning
	Valvular dysfunction	Usually chronic, occasionally acute (e.g. ruptured papillary muscle of mitral valve)
	Dynamic LVOT obstruction
	Acute cor pulmonale	Pulmonary embolism – usually dilated right heart with depressed systolic function, sometimes with clot in a proximal pulmonary artery
Sepsis (source?)	Vegetations, regurgitation ± abscess	Infective endocarditis until proven otherwise
Sepsis (source?)	Normal TOE examination	Infective endocarditis unlikely. If clinically indicated serial TOE
Systemic emboli (source?)	LA/LAA clot	Usually enlarged LA and atrial fibrillation TOE usually required for diagnosis
	LV clot	Usually associated with RWMA or global LV depression. TTE for apical clot
	Aortic atherosclerotic	TOE essential for diagnosis plaques
	Vegetations – aortic or abscess	Septic?
	Clot – prosthetic aortic or mitral valve	Associated prosthetic valve dysfunction
	Patent foramen ovale with paradoxical embolism	RA pressure > LA pressure (e.g. IPPV with high PEEP). TOE (bubble contrast) usually necessary for diagnosis
	Tumour (e.g. LA myxoma)	Uncommon
Pulmonary oedema (cause?)	↓ LV systolic/diastolic function	Isolated LV diastolic dysfunction not uncommon
	Valvular dysfunction (MR, MS, AR, AS)	If flail leaflet/ruptured papillary muscle suspected then TOE indicated
	Intracardiac shunt
	Normal	Suggests non-cardiac cause (e.g. ARDS)
Dyspnoea/hypoxia without pulmonary oedema (dyspnoea cause?)	Pulmonary embolism (dilated right heart chambers ± clot in pulmonary artery)	Infers moderate to large clot burden i.e. submassive/massive PE
	Cardiac tamponade	Usually other clinical signs of tamponade present
	Miscellaneous e.g. chronic cor pulmonale, RVH, constrictive pericarditis, diastolic dysfunction, intracardiac shunt, RV volume overload
Chest pain of uncertain aetiology (chest pain cause?)	RWMA	Infers presence of coronary artery disease
	Dissecting aortic aneurysm (intimal flap, true/false lumen)	TOE more sensitive than TTE
	PE (dilated right heart chambers ± clot in pulmonary artery)	Moderate to large embolus
	Pericarditis	Effusion often too small to diagnose with echocardiography
	Aortic stenosis	Clinical signs of stenosis may be absent

AR, aortic regurgitation; ARDS, acute respiratory distress syndrome; AS, aortic stenosis; ASD, atrial septal defect; IABP, intra-aortic balloon pump; IPPV, intermittent positive-pressure ventilation; LA, left atrium; LAA, left atrial appendage; LV, left ventricle; LVOT, left ventricular outflow tract; MR, mitral regurgitation; MS, mitral stenosis; PE, pulmonary embolism. PEEP, positive end-expiratory pressure; RA, right atrium; RV, right ventricle; RVH, right ventricular haemorrhage; RWMA, regional wall motion abnormality; SAM, systolic anterior motion; TOE, transoesophageal echocardiography; TR, tricuspid regurgitation; VSD, ventricular septal defect.

As a general rule, unless there is gross haemodynamic instability or some other overriding reason, the echocardiographic examination should be structured around a basic set of 2-D views. This applies to both TTE and TOE. The focus of the TOE examination should be on the most important specific clinical question. The second priority should be other potential pathology in the differential diagnosis. A printed report should be performed after each echocardiographic examination and should be stored and available for future reference.¹³

VALVULAR HEART DISEASE ^14,¹⁵

Echocardiography is the ‘gold standard’ for evaluating valvular heart disease. 2-D echocardiography provides excellent imaging of all cardiac valves. Various Doppler techniques allow accurate haemodynamic evaluation of the valves.

VALVULAR STENOSIS

Narrowing or stenosis of any heart valve obstructs blood flow, increasing velocity and causing a pressure gradient across the valve. Evaluation of valvular stenosis requires: (1) imaging of the valve to define the morphology and mobility of the valve cusps; (2) some quantification of the degree of stenosis; and (3) the effect of pressure overload on relevant cardiac chambers. Transvalvular pressure gradients (ΔP) can be estimated by Doppler techniques (see above). Note that in conditions of low cardiac output, pressure gradients will be low, thereby underestimating the severity of stenosis.

Aortic stenosis

The most common aetiology of valvular aortic stenosis is degeneration and calcification of the valve apparatus, although bicuspid aortic valve and rheumatic heart disease may also cause aortic stenosis. Thickening/calcification of relatively immobile valve leaflets causes valvular obstruction and pressure overload, resulting in LV hypertrophy. Severe aortic stenosis may occasionally cause sudden collapse, cardiogenic shock or pulmonary oedema of unknown cause.

Assessment of severity

Visual assessment of aortic valve cusp thickness, calcification, mobility and degree of LV hypertrophy will often indicate the physiological significance of the aortic stenosis. Doppler echocardiography is essential for assessment of severity: (1) peak aortic velocity; (2) mean transvalvular pressure gradient; (3) aortic valve area; and (4) ratio of LVOT velocity/aortic valve velocity (V₁/V₂).

• Peak aortic velocity (normal = 1.7 m/s) (Figure 23.7). In the presence of thickened immobile aortic valve cusps, a peak velocity = 4.5 m/s (i.e. peak pressure gradient 81 mmHg) almost certainly indicates severe stenosis.

• Mean transvalvular pressure gradient. Aortic valve velocity integrated over time is displayed on the monitor and the spectral display traced. A mean pressure gradient of 50 mmHg suggests severe aortic stenosis. From a practical point of view, a patient with normal LV systolic function, a peak aortic velocity gradient of 3.0 m/s and a mean pressure gradient of 30 mmHg is extremely unlikely to have critical aortic stenosis, whereas a patient with thickened aortic leaflets, restricted movement of cusps, aortic velocity > 4.5 m/s and a mean pressure gradient = 50 mmHg will almost certainly have critical aortic stenosis. Difficulties in assessment of aortic stenosis severity arise in patients with depressed LV function and a peak aortic velocity not greatly increased, typically about 3.0–3.5 m/s, and a transvalvular gradient of 25–30 mmHg. Such patients may have mild to moderate aortic stenosis and unrelated LV dysfunction or critical aortic stenosis with severe depression of LV function; dobutamine stress echocardiography is a useful test to distinguish one condition from another.

• Aortic valve area (normal 3–4 cm²). In severe aortic stenosis the aortic valve area is 0.75 cm². The aortic valve area is estimated using the continuity equation, which is based on the principle that, if the volume of blood flow upstream from an orifice is known, then measuring the velocity of blood immediately downstream from the orifice allows calculation of its cross-sectional area. According to the formula: A₁ × V₁ = A₂ × V₂, where A₁ = LVOT cross-sectional area,V₁ = LVOT VTI, A₂ = aortic valve cross-sectional area, V₂ = aortic VTI. The equation can be rearranged: A₂ = A₁ × V₁/V₂.

• Ratio of LVOT velocity/aortic valve velocity (V₁/V₂) provides a ‘dimensionless’ measurement of aortic stenosis severity which is independent of cardiac output. A velocity ratio (V₁/V₂) of < 0.25 implies an aortic valve area of 0.75 cm ² (i.e. severe aortic stenosis).

Figure 23.7 Continuous-wave Doppler recording of an aortic stenosis jet in a patient with critical stenosis. The jet velocity is 589 cm/s (5.89 m/s). The calculated maximum pressure gradient is 139 mmHg. Mean pressure gradient is 94 mmHg.

Accurate estimation of maximum aortic jet velocity is best obtained by TTE. However, TOE allows planimetry (tracing) of the aortic valve orifice in the short-axis view which correlates well with catheter-derived aortic valve area.¹⁶ TOE planimetry may be useful in critically ill patients when other data are inconsistent or inconclusive.

Mitral stenosis

Significant mitral stenosis can occasionally contribute to haemodynamic instability in critically ill patients. Dyspnoea has usually been present for many years, so that mitral stenosis has been diagnosed prior to admission to intensive care. Mitral stenosis secondary to rheumatic heart disease produces typical M-mode (see Figure 23.1) and 2-D images of thickened and fused mitral valve leaflets with diastolic doming, resulting in a ‘hockeystick’ appearance of the anterior leaflet, together with an enlarged LA (Figure 23.8). In the short-axis view the mitral valve appears as a ‘fish mouth’ orifice, the area of which can be estimated (normal 4–6 cm²) (Figure 23.9) using planimetry.

Figure 23.8 Two-dimensional (2-D) echocardiogram of a patient with mitral stenosis. The left atrium (LA) is enlarged. There is doming of both anterior (AMVL) and posterior (PMVL) leaflets with the typical ‘hockeystick’ appearance of the anterior leaflet (transthoracic echocardiography parasternal long axis). LV, left ventricle.

Figure 23.9 Two-dimensional (2-D) echocardiogram of a patient with mitral stenosis. There is diminished separation between the anterior and posterior leaflets, with a typical ‘fish-mouth’ appearance. The valve area by 2-D planimetry is 0.99 cm² (transthoracic echocardiography parasternal short axis).

Assessment of severity

In addition to visualisation and planimetry of the valve, accurate assessment of mitral valve stenosis requires Doppler echocardiography: (1) peak velocity (E) across the mitral valve is increased (normal =1.3 m/s). In severe mitral stenosis, the E velocity is usually > 2 m/s; (2) transvalvular pressure gradient: the mean pressure gradient between the LA and LV is usually > 10 mmHg in severe stenosis; (3) pressure half-time (P_t ½) is the time it takes for the peak pressure gradient to halve. The rate of pressure decline across a stenotic orifice is determined by its cross-sectional area (i.e. the smaller the orifice, the slower the rate of decline). Pressure half-time > 220 ms indicates severe mitral stenosis (mitral valve area = 1 cm ²)¹⁷ (Figure 23.10). Pulmonary artery pressure should also be estimated (see below).

Figure 23.10 Pressure half-time (Pt_½) measurement in a patient with mitral stenosis. The Pt_½ corresponds to a valve area of 0.94 cm².

Tricuspid stenosis and pulmonary stenosis

These disorders can also be diagnosed and quantified using similar echocardiographic techniques.

VALVULAR REGURGITATION

Evaluation of valvular regurgitation requires: (1) assessment of the valve morphology; (2) estimation of severity of regurgitation; and (3) effects of volume overload on relevant cardiac chambers. Mild valvular regurgitation of the mitral, tricuspid and pulmonary valves is common (70–90%) and such findings have no clinical significance. Trivial aortic regurgitation is found in only about 5% of examinations.

Aortic regurgitation

2-D echocardiography may show the cause of the aortic regurgitation, for example, valve destruction caused by infective endocarditis or a dilated aortic root with aortic dissection or, most commonly, degenerative calcification disease of aortic valve.

Assessment of severity ¹⁸

CFD imaging allows some grading of severity of regurgitation by comparing the width of the regurgitant jet to the LVOT area (Figure 23.11). If the jet occupies more than 60% of the LVOT area, then severe aortic regurgitation is probably present.

Figure 23.11 Severe aortic regurgitation; aliased jet occupies entire left ventricular (LV) outflow tract (transoesophageal echocardiography long axis). LA, left atrial.

The velocity of the regurgitant jet is directly related to the pressure gradient between the aortic valve and the left ventricle. In severe aortic regurgitation, especially acute, aortic diastolic pressure falls rapidly and LV end-diastolic pressure increases, resulting in a rapid decay of the pressure gradient: an aortic regurgitant pressure half-time (250 m/s) suggests severe regurgitation. Flow reversal downstream during diastole in the descending aorta indicates severe aortic regurgitation.

Mitral regurgitation

As with aortic regurgitation, echocardiography is useful in detecting the aetiology and mechanism of mitral regurgitation, such as annular dilation (poor coaptation of leaflets), endocarditis (leaflet destruction/perforation) and chordal or papillary muscle rupture (inadequate leaflet support). In mitral regurgitation, the left ventricle ejects blood both into the aorta and backwards into the LA. Early in the course of mitral regurgitation the increase in stroke volume is achieved by an increase in LV ejection fraction (EF). Thus, a hyperdynamic normal-sized left ventricle with normal-sized LA implies acute mitral regurgitation. Over time mitral regurgitation causes enlargement of the LA and eventually the left ventricle dilates and LV function becomes less hyperdynamic. This apparent normalisation of LV function suggests impaired contractility and increased LA pressure. Pulmonary artery pressure also increases over time.

Assessment of severity ¹⁸

Severity of mitral regurgitation is assessed by CFD imaging of the regurgitant jet. The severity of mitral regurgitation is directly proportional to the size of the regurgitant jet within the LA (Figure 23.12). Mitral valve colour flow jet reaching the posterior wall of the LA suggests severe mitral regurgitation, as does systolic flow reversal in the pulmonary veins. Colour flow mapping of the narrowest cross-sectional area of a regurgitant jet (vena contracta) can estimate the severity of the mitral regurgitation¹⁹ and this can be done in < 1 min.²⁰ Combining 2-D imaging with spectral Doppler imaging can be used to quantify valvular regurgitation by measuring regurgitant volume (volume of blood that leaks through an incompetent valve) and regurgitant fraction (fraction or percentage of stroke volume that regurgitates).

Figure 23.12 Severe mitral regurgitation secondary to flail posterior mitral valve leaflet (transoesophageal echocardiography four-chamber).

Another approach for estimating the regurgitant volume is to examine the flow pattern on the LV side of the mitral valve. Proximal isovelocity surface area (PISA): PISA calculations, regurgitant volumes and fractions can be estimated.

Tricuspid regurgitation and pulmonary regurgitation

These disorders can be diagnosed and evaluated using similar principles; for example, severe tricuspid regurgitation causes systolic flow reversal in the hepatic veins.

INFECTIVE ENDOCARDITIS

Infective endocarditis should be suspected in any critically ill patient with sepsis and no obvious source of infection. The echocardiographic hallmark of endocarditis is the presence of vegetation(s) which appear as an echogenic chaotically moving mass attached to a heart valve (Figure 23.13). The Duke criteria²¹^–²³ integrate clinical, microbiological and echocardiographic information in suspected infective endocarditis. Echocardiographic findings consistent with vegetation or new endocardial infection (abscess, prosthetic valve dehiscence, etc.) are major diagnostic criteria. In some situations, clot, tumour or marantic vegetation may mimic endocarditis. In a patient with endocarditis, echocardiography will: (1) image the vegetation and assess its size; (2) diagnose complications such as paravalvular abscess and fistula; (3) examine underlying valve morphology (e.g. bicuspid aortic valve); (4) diagnose and assess severity of associated valvular regurgitation; (5) assess cardiac function; and (6) image other heart valves. Large vegetations (> 10 mm) increase the risk of embolic events.²⁴ Despite advances in imaging, TTE is still insensitive compared to TOE²⁵ for detection of native valve vegetations and fails to demonstrate about 50% of them. TOE accuracy is much better, with a sensitivity and specificity of about 90%. In one study the negative predictive value was 86%.²⁶ TTE may be adequate to exclude vegetations in patients when clinical suspicion for endocarditis is low²⁷ provided that image quality is adequate.

Figure 23.13 Large vegetation (Veg) on the anterior mitral valve leaflet. The patient presented with septic shock and ‘meningitis’ and no obvious cardiac murmur (transoesophageal echocardiography long axis). LV, left ventricle.

About 25% of patients with Staphylococcus aureus septicaemia have infective endocarditis, even in the absence of obvious clinical signs.²⁸ TOE is essential for the diagnosis and detection of associated complications.

PROSTHETIC VALVES

Echocardiographic examination of prosthetic valves requires knowledge of the various types of valves (bioprosthesis, homograft, mechanical). Blood velocity isnearly always increased through normal prosthetic valves and trivial regurgitation is characteristic of most mechanical valves. Such jets are always small, may be multiple and occur within the sewing ring of the valve. Paravalvular leaks are always abnormal and associated with dehiscence; infective endocarditis should always be considered. Suspected prosthetic valve endocarditis mandates TOE and serial examinations are often necessary. A characteristic rocking motion of the valve is diagnostic of dehiscence. TOE is superior to TTE for assessing mitral regurgitation in patients with prosthetic mitral valves, as acoustic shadowing by the mechanical valve of the LA causes poor imaging. In most instances, adequate echocardiographic assessment of prosthetic valves requires combined transthoracic and transoesophageal imaging.

VENTRICULAR FUNCTION AND HAEMODYNAMIC ASSESSMENT

In critically ill patients, echocardiography, especially TOE, has evolved as an independent approach in the assessment and monitoring of cardiovascular haemodynamics. Cardiac systolic function is directly visualised and other parameters can be measured.

The EF is the percentage of the LV diastolic volume that is ejected with each heart beat (normal EF > 50%). It should be appreciated that EF is the result of complex interactions between ventricular loading conditions and contractile state, and may not always reflect true ventricular contractility. Nevertheless, EF is extremely useful and can be estimated using various techniques:

1 Visual inspection of 2-D images. Although widely used, this method is subjective.

2 The American Society of Echocardiography recommends the modified Simpson’s rule, where LV volumes are measured in two orthogonal views (apical four-chamber and apical two-chamber) (Figure 23.14a–d). The LV endocardial border is traced at end-diastole and end-systole. The ventricle is considered to be a series of discs which are summated. Calculation of LV end-diastolic and end-systolic volumes, although complicated, is easily computed using modern echocardiographic machines.

Figure 23.14 Left ventricular (LV) ejection fraction estimation using modified Simpson’s rule. The long axis of the LV is divided into a series of discs from base to apex; the LV endocardial border is traced in end-diastole (LVED) and end-systole (LVES) in two orthogonal views: apical (a) LVED, (b) LVES; four-chamber (c) LVED, (d) LVES two-chamber.

where EDV = LV end-diastolic volume and ESV = LV end-systolic volume.

Real-time 3-D echocardiography is more accurate than 2-D techniques for the quantification of LV volumes.²⁹

Regional left ventricular function ³⁰

Assessment of LV regional wall motion analysis is based on grading the contractility of individual segments: 1 = normal or hyperkinesis; 2 = hypokinesis; 3 = akinesis; 4 = dyskinesis (paradoxical systolic motion); 5 = aneurysmal. The American Society of Echocardiography divides the LV into 16 segments (six at the base, six at the mid ventricular level and four at the apex) (Figure 23.15).

Figure 23.15 Regional wall motion analysis. AS, Anteroseptum; Ant, anterior; Lat, lateral; post, posterior; Inf, inferior; Sep, septum.

Real-time 3D echocardiographic imaging is more sensitive than 2-D techniques for the detection of regional LV wall motion abnormalities.²⁹

Stroke volume and cardiac output

Doppler-derived blood flow velocities can be used to quantify cardiac output. The technique is based on the principle that the VTI of blood flow multiplied by the cross-sectional area (CSA) of its orifice yields an estimate of stroke volume (SV = CSA × VTI) (Figure 23.16a, b). The LVOT is the most commonly used site for stroke volume and cardiac output estimation. In general, correlations between echocardiographic and thermodilution-derived cardiac output have been reasonable, although neither method is a ‘gold standard’.

Figure 23.16 Calculation of stroke volume (SV) using the left ventricular outflow tract (LVOT). (a) LVOT diameter (LVOT_d) is measured from parasternal long axis view. LVOT_CSA = 0.785 × (LVOT_d)², where LVOT_CSA is the LVOT cross-sectional area. (b) LVOT_VTI measured using pulsed-wave Doppler (apical five-chamber or long-axis view), where LVOT_VTI is the velocity–time integral. SV = LVOT_CSA × LVOT_VTI.

Real-time 3-D is more accurate than 2-D echocardiography for measurement of LV volumes. Studies comparing 3-D echocardiographic LV volumes with magnetic resonance imaging (MRI) confirm the accuracy of 3-D echocardiography.²⁹

Preload (left ventricular end-diastolic volume)

The assessment of LV end-diastolic volume (LVEDV) is often important in critically ill patients. Significant hypovolaemia is inferred by LV end-systolic cavity obliteration. Although a presumptive diagnosis of hypovolaemia can often be made with such echo signs, other causes of end-systolic cavity obliteration, such as decreased systemic vascular resistance, severe mitral or aortic regurgitation, or ventricular septal defect, must be considered. Integrating the clinical picture together with CFD and, if necessary, quantitative assessment of preload, differentiates between the various diagnoses.

Modified Simpson’s rule technique is the most accurate echocardiographic method for measuring LVEDV/preload in clinical practice (Figure 23.14a, c).

Afterload

End-systolic wall stress (where wall stress = (pressure × radius)/2 × wall thickness), determined by echocardiography, provides an index of LV afterload. This is rarely used in clinical practice.

Diastolic function ³¹

The non-invasive assessment of LV diastolic dysfunction has seen major advances in recent years. Diastolic dysfunction is a disorder of LV filling, where the LV is unable to fill to a normal LVEDV without an abnormal increase in end-diastolic pressure. All patients with LV systolic dysfunction have some degree of LV diastolic dysfunction. Some patients with congestive cardiac failure have normal or near-normal LV systolic function with isolated diastolic heart failure. Gandhi et al.,³² using echocardiography, demonstrated that pulmonary oedema in patients with marked systemic hypertension is commonly due to isolated diastolic dysfunction. Mitral and pulmonary inflow patterns are Doppler techniques routinely used to diagnose diastolic dysfunction and assess severity.

E/E ¹ ratio

Early diastolic blood flow from LA to LV results in an E-wave. Doppler tissue imaging recordings from the lateral (or medial) mitral annulus during early diastole result in an E ¹. E¹ is a good index of LV relaxation. The E/E¹ ratio correlates well with LV filling pressures. An E/E¹ ratio > 15 is associated with LA pressures > 15 mmHg, whereas an E/E¹ < 8 is highly specific for normal LA pressure.³¹

RIGHT VENTRICLE (RV)

Evaluation of right-sided heart function is important in critically ill patients.

Chamber size

The normal shape of the RV is a crescent, curving in front of, and concave towards, the LV. A combination of image planes is necessary to assess RV size. Ventricular dilation may be due to right-sided volume overload secondary to conditions such as tricuspid regurgitation, pulmonary regurgitation or atrial septal defect. CFD and occasionally intravenous injection of a contrast agent such as agitated saline will help distinguish the aetiology.

Wall thickness

RV hypertrophy (due to pressure overload) leads to increased thickness of the RV free wall.

Systolic function

Right ventricular end-diastolic volume assessment is difficult because the shape of the RV and the presence of prominent trabeculations make the endocardial border difficult to outline. TOE is essential for assessing RV dysfunction following cardiac transplantation.

Detection of regional wall motion abnormalities is a sensitive and specific marker of RV ischaemia or infarction.³³ The diagnosis of cardiac contusion, which occurs commonly after blunt chest trauma, is often unrecognised and can be readily made by TOE.³⁴

Pulmonary artery pressure

This is an important and routine part of assessing RV function. Some tricuspid regurgitation is present in > 90% of the adult population. The simplified Bernoulli equation is used: RV systolic pressure is obtained by adding right atrial pressure (RAP) to the transtricuspid gradient derived from the peak tricuspid regurgitation velocity (V_tr) measured by Doppler (Figure 23.17). In the absence of pulmonary stenosis, systolic RV pressure = systolic PA pressure.

Figure 23.17 Estimation of right ventricular systolic pressure (RVSP) and systolic pulmonary artery pressure (SPAP) in the presence of tricuspid regurgitation (TR). Peak TR velocity (VTR) measured by continuous-wave Doppler is 396 cm/s (3.9 m/s). Right atrial pressure (from central venous pressure) = 15 mmHg. Difference in pressure between RA and RV (using Bernoulli equation) is 4 × () (i.e. 4 × (3.9)² = 62). Therefore RVSP = 15 + 62 = 77 mmHg. In the absence of pulmonary stenosis, RVSP = SPAP.

CARDIOMYOPATHIES

A simple classification of the cardiomyopathies includes: (1) dilated; (2) hypertrophic; and (3) restrictive.

DILATED CARDIOMYOPATHY

All echocardiographic features of dilated cardiomyopathy are non-specific, but LV dilation and reduced globalsystolic function are characteristic. Both LV end-diastolic and LV end-systolic volumes are increased, with EF and fractional area of change uniformly reduced. Mural thrombus may be present. The RV may also be affected. There may be atrial enlargement. Due to the increased LVEDV, stroke volume and cardiac output may be preserved at rest. Significant mitral regurgitation, secondary to annular dilation and poor coaptation of the mitral leaflets, may be present. Diastolic dysfunction is invariably present and is an important prognostic feature. Pulmonary hypertension estimated from tricuspid regurgitation velocity may predict a poor prognosis.

HYPERTROPHIC CARDIOMYOPATHY

In critically ill patients, the diagnosis cannot be made without the assistance of echocardiography, where asymmetrical septal hypertrophy, the most common feature, can be readily identified and quantified, as can a variety of other hypertrophic patterns. In the elderly, a normal proximal septal bulge (‘sigmoid septum’) should not be confused with true hypertrophic cardiomyopathy.

Ventricular outflow tract obstruction

During systole, the anterior leaflet of the mitral valve moves anteriorly toward the interventricular septum and may obstruct the LVOT. The LVOT obstruction is dependent on loading conditions and may be absent at rest. Indeed, the dynamic nature of the LVOT obstruction is one of the hallmarks of this condition. The systolic anterior motion (SAM) of the leaflet distorts the mitral valve, causing mitral regurgitation. CFD (and PW Doppler) not only localise the site of LVOT obstruction, but also quantify the degree of mitral regurgitation. The severity of LVOT obstruction can be estimated by CW Doppler. Although the degree of obstruction may vary, a peak gradient of 50 mmHg indicates significant LVOT obstruction.

Dynamic LVOT obstruction without asymmetrical septal hypertrophy ³⁵

Dynamic LVOT obstruction with SAM can occur whenever a hyperdynamic state exists, especially in elderly patients with hypovolaemia and a septal bulge (‘sigmoid’ septum). The haemodynamic effects of this condition are identical to the LVOT obstruction of hypertrophic cardiomyopathy and can cause cardiogenic shock. Dynamic LVOT obstruction may be precipitated or worsened by positive inotropic or vasodilator therapy and intra-aortic balloon pumping. In hypotensive patients, a paradoxical haemodynamic response to conventional inotropic therapy is an indication for TOE. The LVOT gradient can usually be abolished by ensuring adequate volume replacement, avoiding positive inotropic agents and increasing afterload by administration of an agent such as phenylephrine (a pure α-agonist).

RESTRICTIVE CARDIOMYOPATHY

Primary restrictive cardiomyopathy is characterised by impaired LV filling secondary to idiopathic myocardial stiffening. Echocardiographic findings usually include marked biatrial enlargement with normal LV cavity size and wall thickness and normal or slightly reduced systolic function. A characteristic restrictive mitral inflow pattern is present.

Secondary restrictive cardiomyopathy has a similar abnormal diastolic filling pattern secondary to a known myocardial cause (e.g. amyloidosis).

PERICARDIAL DISEASE

PERICARDIAL EFFUSION AND TAMPONADE

Echocardiography is the accepted gold standard for both the diagnosis of pericardial effusion and its haemodynamic significance. Pericardial fluid or blood can be recognised on 2-D as an ‘echo-free’ space around the heart (Figure 23.18). Whenever a pericardial effusion is imaged, the possibility of tamponade should be considered. When the effusion within the pericardial sac is large enough, intrapericardial pressure exceeds RV diastolic pressure and cardiac tamponade occurs, causing impaired cardiac filling and low cardiac output. 2-D echocardiographic features of cardiac tamponade may include:

1 Moderate to large pericardial effusion (usually). However, it should be appreciated that a small rapidly accumulating effusion (50–100 ml) can increase intrapericardial pressure acutely and tamponade the heart. Conversely, a large, slowly increasing effusion, sometimes over 1 l in volume, may have little effect on intrapericardial pressure and cause no haemodynamic embarrassment.

2 The heart may swing excessively in the pericardial sac. Electrocardiographically this may manifest as electrical alternans.

3 Early diastolic collapse of the RV free wall is more specific (85–100%) but less sensitive (60–90%) for cardiac tamponade than

4 Late diastolic atrial collapse. Sustained (30% of cardiac cycle) RA collapse is more likely to be significant than brief collapse.

5 Respiratory variation in diastolic filling is manifested clinically by pulsus paradoxus. Doppler echocardiography demonstrates a similar phenomenon, namely exaggerated increase in tricuspid and decrease in mitral valve diastolic inflow with inspiration. A tricuspid valve inflow variation > 25% and a mitral valve inflow variation > 15% with respiration is consistent with tamponade. The inferior vena cava is usually dilated and its diameter does not vary with respiration in patients with cardiac tamponade. Doppler examination of the hepatic veins is also very important.

Figure 23.18 Massive (5 cm) pericardial effusion (PE) in a patient with tamponade (transthoracic echocardiography subcostal).

Echocardiography may also be used to guide pericardiocentesis by locating the optimal site for puncture and monitoring the residual effusion after drainage.³⁶ As 2-D echocardiography is a tomographic imaging technique, the needle tip is often difficult to visualise.

CARDIAC TAMPONADE OCCURRING AFTER CARDIAC SURGERY

The tamponade may be localised with differential compression of the right- or left-sided chambers and the clot is usually ‘echo-dense’ with little or no ‘echo-free’ space around the heart (Figure 23.19). Localised tamponade (e.g. of the RA or LA) is not uncommon in such patients. Not surprisingly, standard echocardiographic criteria, such as intermittent chamber collapse, are unreliable in detecting tamponade after cardiac surgery. The presence of 1 cm of pericardial separation (fluid/clot) in a patient with unexplained clinical deterioration (hypotension, decreased cardiac output) appears to be sensitive indetecting tamponade.³⁷ In the absence of tamponade, other causes of hypotension, such as unsuspected hypovolaemia, ventricular dysfunction or LVOT obstruction, may be diagnosed by echocardiography. Indeed, unnecessary reoperation may be prevented in some instances.

Figure 23.19 Localised clot compressing mainly the right heart in a patient after cardiac surgery (transoesophageal echocardiography four-chamber). LA, left atrium; RA, right atrium; RV, right ventricle.

CONSTRICTIVE PERICARDITIS

Echocardiography is of value in the diagnosis of constrictive pericarditis. There is often thickening of the pericardium, which may be imaged with 2-D echocardiography. Doppler studies of the tricuspid, mitral and pulmonary veins may show typical findings.

CARDIAC MASSES

Before any diagnosis is made of a cardiac mass, it is essential to rule out pseudomasses (i.e. artefacts or normal cardiac structures). TOE is usually more sensitive than TTE. Intracardiac masses include thrombi, tumours and vegetations.

INTRACARDIAC THROMBUS

Thrombi can occur in any cardiac chamber but are most common in the LV (often at the apex) and the LA (usually in the LA appendage). Low-flow states manifest as spontaneous echocardiographic contrast or echocardiographic ‘smoke’ and are precursors of clot formation and thromboemboli. LV clot (Figure 23.20), when it occurs after myocardial infarction, is associated with the regional wall motion abnormality. Clot may also occur in patients with dilated cardiomyopathy and severe global depression of LV systolic function. TTE is more sensitive than TOE for detection of LV apical clot. LA clot is much more common than RA clot and usually occurs in the setting of atrial fibrillation or mitral stenosis. TOE provides vastly superior images compared to TTE as LA clot is usually located in or near the LA appendage, which is poorly visualised with TTE. Conventional managementfor patients with AF = 48 h requires therapeutic anticoagulation for 3 weeks before planned cardioversion. Another approach necessitates LA clot exclusion by TOE. If there is no LA clot, then cardioversion may proceed immediately, otherwise cardioversion is delayed. This TOE-based strategy has been shown to be associated with low embolic rates and a shorter time to cardioversion compared to conventional strategy, but no difference in outcome.³⁸

Figure 23.20 (a) Left ventricular (LV) apical clot. There was an associated regional wall motion abnormality (transthoracic echocardiography apical four-chamber). (b) Clot in left atrial appendage. RV, right ventricle; RA, right atrium; LA, left atrium.

CARDIAC TUMOURS

LA myxoma is by far the most common benign cardiac tumour and can often be imaged by TTE, although superb images can be obtained with TOE. The tumour is usually attached by a pedicle to the interatrial septum. During diastole, the tumour may obstruct the mitral valve orifice. Rarely, other tumours, primary or metastatic, may be visualised.

AORTIC DISEASE

TOE is paramount for diagnosing aortic pathology in critically ill patients because of its accuracy, portability, real-time imaging and low complication rate.

AORTIC DISSECTION ³⁹

TOE is very useful for diagnosing aortic dissection. The sensitivity and specificity of TOE in diagnosis of aortic dissection are equivalent to that of computed tomography (CT) or MRI.⁴⁰ 2-D echocardiography will reveal the intimal flap (Figure 23.21) which is the hallmark of aortic dissection. CFD is useful in differentiating true from false lumen: (1) the false lumen is usually larger with slower flow; (2) the intimal flap pulsates in systole towards the false lumen because of the higher pressure in the true lumen. The entry point of the tear and the presence and extent of aortic regurgitation can also be evaluated.

Figure 23.21 Aortic dissection involving the ascending aorta – the intimal flap (arrows) impinges on the aortic valve (AV) cusps during diastole (transoesophageal echocardiography). LA, left atrium.

INTRAMURAL HAEMATOMA (IMH)⁴¹

IMH represents a variant of aortic dissection with virtually identical clinical presentation. Indeed, it may be an early finding in patients who develop classical aortic dissection or rupture. IMH cannot be diagnosed by angiography as there is no intimal flap. 2-D echocardiography identifies IMH as a (> 0.7 cm) circular or crescentric thickening of the aortic wall or an echolucent zone in the aortic wall. There may be a thrombus-like echo pattern of the aortic wall.

TRAUMATIC AORTIC RUPTURE

TOE is both sensitive and specific for the diagnosis of aortic trauma.⁴² The echocardiographic signs of aortic injury may include an intimal flap, aortic wall haematoma, aortic occlusion and fusiform aneurysm. Echocardiographic assessment for aortic injury requires an experienced operator.