A structured 5-question approach for echocardiographic evaluation of undifferentiated shock

Abstract

Shock is the clinical state whereby the patient’s circulation fails to meet the body’s metabolic demands. Both acute illnesses that lead to shock and therapies employed to support the shocked patient (including intravenous fluids, vasoactive infusions and ventilatory support) may result in changes to the normal function of the heart. The conventional approach to interpretation of echocardiography in the steady physiological state needs modification when this test is applied to the shocked patient. In this paper we explain when and why modifications are required, and provide echocardiographers already established in outpatient practice with a practical approach to performing and reporting echocardiograms in shocked patients.

Supplementary information

The online version contains supplementary material available at 10.1186/s44156-026-00129-z.

Background

In this paper we provide echocardiographers established in outpatient practice with a structured approach to assessing the adult patient with undifferentiated shock.

This guidance reflects the expert opinion of two Consultant led Critical Care Echo services in the UK. The guidance is intentionally pragmatic and prioritises those views and measurements that are found within the British Society of Echocardiography’s (BSE) minimum dataset [1] already familiar to experienced echocardiographers.

This guidance relates to the performance of ‘comprehensive’ echocardiography. Intentionally we have not covered the use of Emergency Echo Protocols which follow their own minimum dataset, are supported by bespoke training packages and are recorded on bespoke reporting templates.

The field of Critical Care Echocardiography (CCE) is vast and developing rapidly. We have deliberately focused on the use of echocardiography in the assessment of shock. We do not discuss the broader practice of critical care echo or imaging cardiovascular mechanical support devices.

Defining shock

Shock is a clinical state of critical tissue hypoxia, caused by:

• Reduced oxygen delivery.

• Increased oxygen demand.

• Impaired oxygen utilisation.

• Or a combination of these factors.

Shock is diagnosed based on evidence of end-organ dysfunction. Whilst many shocked patients are hypotensive; a patient may be shocked despite normal or elevated blood pressure. Undifferentiated shock refers to the situation where the cause(s) of the shock in not yet diagnosed.

The causes of shock are divided into four main categories, according to the dominant cause [2, 3]:

• Cardiogenic.

• Hypovolaemic.

• Obstructive.

• Distributive.

In practice it is common for multiple causes of shock to coexist.

The role of echocardiography in shock

Echocardiography is the preferred first-line modality to investigate the cause of undifferentiated shock [4–6]. The role of echocardiography is to:

• Identify the cause(s) of shock, in order to start appropriate management.

• Quantify the likely contribution of different causes to the patient’s clinical state.

• Identify and assess contributory chronic cardiac lesions.

• To track progress over time (to assess for disease progression and/or response to therapy).

Decisions on patient management should never be taken based on the echocardiogram in isolation. Rather, the echocardiogram forms just one part of the comprehensive clinical assessment, and echocardiographers must work alongside the wider clinical team to integrate findings and inform treatment decisions.

Introducing the structured five-question approach

We propose a systematic questioning approach to the assessment of the shocked patient. Answering the following five key questions will facilitate integration of echocardiography into the primary assessment:

1. What is the stroke volume and cardiac output?

2. If the cardiac output is abnormal, what is the most likely cause(s)?

3. Is left atrial pressure likely to be elevated?

4. Is pulmonary artery pressure and/or pulmonary vascular resistance likely to be elevated?

5. How is the situation evolving?

A similar question-and-answer approach has previously been employed to guide echocardiographers assessing critically ill patients in the FUSIC HD EEP [7].

The main body of this paper covers the five key questions, and an approach to answering these questions in all shocked patients (key measurements and assessments are summarised in Fig. 1). For certain presentations and pathologies, there are special considerations that the echocardiographer needs to make in their assessment. These ‘special cases’ are covered in Appendix A of the supplementary materials.

Fig. 1.

Flowsheet highlighting key measurements and assessments to be made when assessing the shocked patient. SVI, stroke volume index; CI, cardiac index; LVOTO, left ventricular outflow tract obstruction; LAP, left atrial pressure; LA, left atrium; IAS, inter-atrial septum; PAP, pulmonary artery pressure; PVR, pulmonary vascular resistance; TR Vmax, tricuspid regurgitation maximum velocity; PASP, pulmonary artery systolic pressure; PH, pulmonary hypertension; PAAT, pulmonary artery acceleration time; IVS, inter-ventricular septum; RAP, right atrial pressure; TAPSE, tricuspid annular plane systolic excursion; RV, right ventricle; PA, pulmonary artery. *A chronically dilated LA may remain dilated during an acute drop in preload (and LAP), as may be seen in haemorrhagic shock.

Echocardiographic assessment in context

There are three major factors influencing loading conditions in shocked patients already receiving critical care support:

• Vasoactive medication.

• Invasive or non-invasive ventilation.

• Volume status.

Having a basic understanding of how these factors influence the heart is useful in placing the echocardiogram in context.

Given the importance of these factors, it is useful to record haemodynamic data from the time of the study, especially if serial studies are performed (and this should be included in the report). Table 1 summarises the demographic and haemodynamic data that should be recorded for each study.

Table 1.

Demographic and haemodynamic data to be reported

Demographic data
Height (cm), actual body weight (kg), ideal body weight (kg), body surface area (m2)
Haemodynamics
Heart rhythm and rate (bpm), systemic blood pressure (mmHg), central venous pressure (mmHg)
Vasoactive infusions
Drugs administered and dose (units are drug dependent)
Mechanical circulatory support
Modality of support and device settings
Respiration (spontaneously breathing patients)
Respiratory rate (bpm), work of breathing (‘normal’ or ‘increased’)
Respiration (patients receiving ventilatory support)
Modality of support (for example Continuous Positive Airway Pressure via face mask, Non-Invasive Ventilation via face mask, or invasive Mechanical Ventilation), respiratory rate (bpm), Positive End Expiratory Pressure (cmH2O), tidal volume (ml), tidal volume indexed to ideal body weight (ml/kg), Peak Inspiratory Pressure (cmH2O), Plateau Pressure (cmH2O)

Assessment in patients receiving pharmacological cardiovascular support

Principles underpinning the use of vasoactive infusions in the shocked patient:

• Patients in a shock state require therapies to support the failing circulation with the aim of limiting consequent organ dysfunction

• Vasoactive drugs affect the circulation in a number of different ways:

  • ○ Inotropy: change in myocardial contractility.
  • ○ Vasoconstriction/vasodilatation: change in vascular tone.
  • ○ Chronotropy: change in heart rate.
  • ○ Lusitropy: change in myocardial relaxation.

• The choice of vasoactive drugs will depend upon the cause of shock (no one drug is inherently superior, but some drugs may be better suited to support the circulation in certain clinical contexts).

• Vasoactive drugs are typically administered in the lowest dose possible to achieve a pre-specified hemodynamic end-point (for example, the Surviving Sepsis guidelines suggest that vasopressors – drugs that increase arterial tone leading to a rise in systemic blood pressure – should be titrated to achieve a mean arterial blood pressure of 65 mmHg in adults with septic shock) [8].

• The introduction or discontinuation of a vasoactive drug, a change in drug dosage or change in the patient’s pharmacodynamic response to a drug are all expected to produce a change within the circulatory system that should be anticipated and assessed.

Considerations when assessing the patient receiving vasoactive infusions:

• A single vasoactive agent may affect the circulation in multiple ways simultaneously.

• The predominant effect is dependent upon not only the choice of agent, but also the relative contributions of different modes of shock.

  • ○ For example, in a patient with ventricular failure the addition of noradrenaline might cause blood pressure to rise, but stroke volume to fall due to increased afterload.
  • ○ In another patient, in whom vasoplegia is the predominant mode of shock, noradrenaline may produce venoconstriction, increasing venous return, leading to a rise in both blood pressure and stroke volume [9].

• Inotropes are initiated with the intention of increasing the force of ventricular contraction and hence stroke volume, in patients with cardiogenic shock due to ventricular dysfunction. When an echocardiogram is performed in a patient receiving inotropes, it is important to recognise that the observed ventricular function represents the current supported state only (and that the ‘true’ ventricular function may be more impaired than is observed).

• Ventricular function is influenced by afterload. In patients with vasoplegia, decreased arterial tone and therefore lowered afterload, may cause ventricular function to appear ‘normal’, with dysfunction only being unmasked when vasopressors restore vascular tone [10].

A summary of the key effects of a number of commonly used vasoactive agents can be found within Appendix B of the supplementary materials.

Assessment in patients receiving respiratory support

Principles underpinning the use of positive pressure ventilation [11]:

• Positive pressure ventilation (PPV) refers to any technique whereby a ventilator generates a positive pressure greater than the pressure within the patient’s airways, which leads to the movement of gas from outside the patient, into the patient’s lungs, known as insufflation.

• Movement of gas out of the patient’s lungs is known as exsufflation, and occurs when the pressure within the ventilator drops below that within the patient’s airways.

• PPV can be delivered non-invasively via a face mask or hood, or invasively via an endotracheal tube or tracheostomy tube.

• The positive pressure generated by the ventilator may be entirely responsible for the pressure gradient known as ‘mandatory mechanical ventilation’, or the patient may make some of their own respiratory effort by generating a degree of negative pleural pressure, which then triggers the ventilator to provide additional support. This is known as supported ventilation.

Heart-lung interactions during positive pressure ventilation:

• In ‘normal’ unsupported spontaneous breathing some of the negative pleural pressure generated by the patient is transmitted to the cardiac chambers and intra-thoracic vessels, altering pressure gradients and hence the flow of blood with each breath cycle. This is called respirophasic flow.

• ○ On inspiration blood is drawn into the right-heart, leading to a rise in RV stroke volume and a corresponding fall in LV ejection volume.

• ○ On expiration flow into the right heart is reduced causing a reciprocal fall in RV stroke volume and allowing a rise in LV ejection.

• In patients receiving mandatory positive pressure ventilation respirophasic flow variation is reversed, such that RV flow is greatest during expiration, whilst LV ejection and flow is greatest during inspiration.

• In patients receiving supported ventilation, the combination of patient generated negative intra-thoracic pressure and ventilator generated positive pressure means that the direction and magnitude of respirophasic changes can be unpredictable.

Considerations when assessing the patient receiving positive pressure ventilation [12]:

• It is essential to know and document which form of ventilatory support the patient is receiving: is a mandatory mode or a supported mode being employed (as this will influence how respirophasic flow variation should be interpreted)?

• The impact of mandatory ventilation on reducing right heart filling and flow during inspiration is magnified if the underlying volume status is low.

• PPV of any form will result in a rise in Pulmonary Vascular Resistance (PVR), which increases RV afterload, leading to increased RV cavity volume in both systole and diastole, with a reduction in RV ejected volume. The gradient between the pulmonary veins and the LA is expected to remain constant (as both are within the chest, and therefore both are exposed to any change in intra-thoracic pressure). However, left ventricular preload will decrease as a down-stream consequence of reduced venous return.

• PPV creates a pressure gradient between the intra-thoracic aorta, and the down-stream extra-thoracic aorta. This has the effect of decreasing LV afterload, promoting forward flow. In normally functioning hearts the potential benefit of decreased afterload is offset by the reduction in venous return to the left atrium (hence decreased LV preload). However, in patients in whom LV failure is a significant contribution to the shock state, PPV can lead to an increase in cardiac output and decrease in myocardial oxygen demand.

• The application of PPV may alter the apparent severity of valve lesions. For example, the reduction in LV preload and afterload may lead to a reduction in functional Mitral Regurgitation (MR) severity [13].

• In spontaneously ventilating patients with cardiac tamponade there is exaggerated respirophasic flow variation across the mitral and tricuspid valves, with a greater degree of flow variation indicating a more haemodynamically significant pericardial effusion [14, 15]. When PPV is applied, not only is the normal heart-lung interaction reversed, but also blunted [16, 17]. Trans-valvular flow variation as a sign of cardiac tamponade may be absent during PPV, and its absence should not be used to exclude the possibility of tamponade [18].

Further detail of the terms used to describe PPV techniques and settings can be found within Appendix C of the supplementary materials.

Assessment under different loading conditions

Principles underpinning manipulation of ‘volume status’ during critical illness:

• Critically ill patients presenting with shock can be expected to experience marked swings in their circulating volume as their disease progresses, and as a result of treatments that are provided.

• In critical illness fluid may be:

  • ○ Lost from the body, for example in haemorrhage, vomiting and burns.
  • ○ Retained within the body but lost from the circulation to the tissues, for example sepsis-associated capillary-leak syndrome.
  • ○ Retained within the circulation but pooled with venous capacitance vessels, as seen in the vasoplegia of distributive shock.

• The administration of additional fluid to restore tissue perfusion is potentially life-saving. However, this may come at the cost of venous congestion and impaired organ function [19].

• Resuscitation can be described in four distinct phases [20]:

  • ○ Salvage/Rescue: immediate life-saving phase, during which intra-venous fluids are likely to be administered, potentially in large quantities.
  • ○ Optimisation: the shock state is now compensated and tissue perfusion has been at least partially restored; the potential benefit of any additional fluids must be offset against the potential negative consequences.
  • ○ Stabilisation: the shock state is now resolved, and the patient will typically be held in a neutral fluid balance, whilst treating the underlying disease that precipitated the shock state.
  • ○ De-escalation: having recovered from the underlying disease process the patient must expel the additional fluid received during earlier resuscitation phases.

Considerations when assessing patients under different loading conditions:

• Cardiac function is dependent upon both preload (variously defined as either the myocardial sarcomere tension or length, immediately prior to contraction) and afterload (the sum of forces opposing ventricular ejection). The heart will behave differently as these variables shift.

  We can consider the impact of alterations of loading parameters in isolation, but the reality is likely to be more complex, with preload, afterload and inherent contractility all potentially affected during acute illness.

• A decrease in preload will result in a reduction in end-diastolic cavity volume, and usually in a reduction in the force of contraction and lower ejected volume.

• In some patients with ventricular failure, myocardial sarcomeres have become stretched beyond their optimal length for maximal force of contraction. In this setting a decrease in preload may be therapeutic, and lead to an increase in force of contraction.

• In low afterload states, for example in a hypotensive patient with distributive shock, ventricular contractility is expected to increase with a smaller ventricular end-systolic volume and an increased ejected volume.

• Conversely as afterload increases, end-systolic volume is expected to increase and the ejected volume decrease.

• Whilst it is possible to anticipate the changes that are expected, what is anticipated does not always match what is observed. For example, in a study of septic patients who remained hypotensive following volume resuscitation, the addition of noradrenaline increased afterload and increased LV contractility and ejection [21]. In these patients it was postulated that contractility may have improved due to a rise in diastolic blood pressure and thus increased coronary perfusion [22].

• What is actually observed should be reported, since the variables affecting observed cardiac function and output are multiple, complex and interacting.

A practical approach for using echocardiography to help predict if the administration of intravenous fluids is likely to result in a significant increase in cardiac output can be found within Appendix D of the supplementary materials.

Question 1: What is the stroke volume and cardiac output?

Summary

• Stroke volume index (SVI) and cardiac index (CI) should be reported for all shocked patients.

• SVI below age and sex specific cut-offs should be reported as ‘low stroke volume’ (Table 2).

• CI 1.8-2.19 L/min/m² should be reported as ‘low cardiac output’.

• CI ≤ 1.8 L/min/m² should be reported as ‘critically low cardiac output’.

• When CI is ≥2.2 L/min/m², but SVI is low AND heart rate is >100 bpm, this should be reported as ‘cardiac output preserved, despite low stroke volume, in the context of tachycardia’.

• When CI is <2.2 L/min/m2, but SVI is within normal limits AND heart rate is <60 bpm, this should be reported as ‘low cardiac output despite preserved stroke volume, in the context of bradycardia’.

Table 2.

Lower limits of normal for stroke volume indexed by body surface area (ml/m²), as assessed using the Doppler measurement technique

	18–40 years	41–65 years	>65 years
Men	27.3	24.6	24.3
Women	25.5	24.2	23.5

Reproduced (modified) with permission from: Patel et al.; WASE Investigators. Normal Values of Cardiac Output and Stroke Volume According to Measurement Technique, Age, Sex, and Ethnicity: Results of the World Alliance of Societies of Echocardiography Study. J Am Soc Echocardiogr. 2021 Oct;34(10):1077–1085

Considerations when assessing stroke volume and cardiac output

1. National and international guidelines on chamber quantification do not quote normal reference intervals for stroke volume (SV), as determined using the Doppler method [23, 24]. The World Alliance of Societies of Echocardiography (WASE) study derived normal reference intervals from 1450 patients in 15 countries [25]. We recommend using the lower bounds derived from these reference intervals to identify patients with a ‘low stroke volume’ (Table 2).

2. Cardiac output (CO) is estimated by multiplying the average stroke volume by the heart rate, and this value should then be indexed to BSA (Fig. 2). The lower boundary of the normal reference interval for cardiac index (CI), in health and at rest, is ~2.5 L/min/m² in young adults, falling slightly in older individuals [26, 27].

• 3There is no single cut-off value that describes ‘enough’ cardiac output for all individuals in all health and illness states [28], and CI estimates must be interpreted in the clinical context. Many individuals living with chronic cardiac disease (with reduced LVEF) may have a resting CI < 2 L/min/m² and yet are not manifestly ‘shocked’ [26], since they have adapted to this low output state.
• 4Patients in a shocked state may experience either an increase or a decrease in the body’s demand for oxygen [29, 30].
• 5Cardiac Index is not a binary index for shock or critical illness. In a study of patients with acute myocardial infarction, almost 90% of patients with evidence of clinical hypoperfusion had CI estimated at <2.2 L/min/m2 [31, 32], and this value has been used as a threshold to diagnose cardiogenic shock in therapeutic trials [33, 34]. However, in this same study ~25% of patients with CI estimated at <2.2 L/min/m2 had no clinical features of impaired tissue perfusion.
• 6It is essential to report both the SVI and CI and to document the patient’s heart rate and rhythm at that time. When the stroke volume is low, the patient may mount a tachycardia to maintain adequate tissue perfusion. Whilst the CI may fall within a ‘normal’ or ‘acceptable’ range, this comes at the expense of increased myocardial oxygen demand [35, 36] and concomitant changes in pulmonary and systemic venous pressures [37–40]. Therefore, ‘normal’ CI cannot be considered ‘normal’ in the context of low SVI and tachycardia.
• 7It is possible for patients with bradycardia to have a low cardiac output despite having a normal SV. Although tachycardia is a common physiological response to critical illness [41] not all patients will be able to mount this response [42, 43].

Fig. 2.

Left ventricular outflow tract (LVOT) velocity time integral (VTI) measurements from a patient in cardiogenic shock following acute anterolateral myocardial infarction. Three consecutive ejections have been measured. LVOT diameter was measured at 2.2 cm, and LVOT area estimated at 3.80 cm². Given the patient’s body surface area (BSA) was estimated at 2.02 m², the Doppler estimate of indexed stroke volume (SVI) is low at 21.3 ml/m², with a low indexed cardiac output (CI) of 1.77 L/min/m²

Question 2: If the cardiac output is abnormal, what is the most likely cause(s)?

Summary

• A systematic approach should be used to evaluate and describe all visualised structures within the heart. For each identified lesion, consider if it is expected to:

  • ○ Reduce onward flow - downstream consequences.
  • ○ Increase back-pressure - upstream consequences.

• Identified structural and/or functional abnormalities may be entirely acute, chronic or represent an acute deterioration in the context of chronic disease.

• Ventricular systolic function should be described quantitatively where possible.

• Left ventricular regional wall motion abnormalities should be reported using the 17-segment model.

• All patients should be evaluated for left ventricular outflow tract obstruction (LVOTO), and either the presence or absence of LVOTO should be reported: LVOTO can occur de novo in patients responding to shock with a tachycardia due to either profound hypovolaemia, subaortic hypertrophy of the LV septal wall or a combination of both.

• All patients who develop shock following acute MI should be evaluated for the presence of an ischaemic ventricular septal defect, which may require dedicated off-axis views.

• Acute and chronic valve lesions should be described using existing cut-offs, but with awareness that the change in loading conditions which occurs during shock may lead to underestimation of severity.

• Evidence of low preload should be specifically sought and reported.

• Echocardiographers should be aware that multiple forms of shock may coexist within the same patient.

Considerations when assessing ventricular function

1. The echocardiogram report should describe the presence and severity of ventricular systolic impairment. Where possible quantitative measures should be used.

2. Left Ventricular Ejection Fraction (LVEF) is a measure familiar to all healthcare professionals working with shocked patients, and severely reduced left ventricular ejection fraction correlates with poor outcomes for shock of both cardiac and non-cardiac causes [44–46]. It should therefore be reported.

3. Whilst LVEF should be reported, there are several reasons why it may be mis-leading in the assessment of critically ill patients [47]:

   • ○ LVEF is the ratio of stroke volume (SV) to left ventricular end-diastolic volume (LVEDV), and therefore abnormalities of the LVEDV may result in a LVEF that is misleading. LV remodelling and volume status are the two biggest factors potentially affecting this ratio [48–50]. When LVEDV is low it is possible for LV stroke volume to be low, despite LVEF falling within the normal range.
   • ○ LVEF is blind to the direction of blood flow. A normal LVEF does not guarantee adequate blood is reaching the systemic circulation if a significant proportion of the ejected volume leaves the LV via a regurgitant mitral valve or ventricular septal defect (VSD).
   • ○ LVEF is influenced by LV afterload, which is in turn dependent upon systemic arterial elastance: low arterial elastance occurs when the arterial system accommodates a relatively large stroke volume for a relatively small increase in arterial blood pressure – as may be seen in distributive shock. This is termed vasodilatation and can occur as a consequence of disease or as a result of therapies (for example in response to inodilators such as dobutamine or milrinone). A ‘normal’ LVEF may be falsely reassuring when systemic arterial elastance is low, and impaired LV contractility may be unmasked when vasoconstrictors are used to restore arterial tone [51, 52].
   • ○ The relationship between LVEF and mortality in critical illness is not linear, but rather ‘U-shaped’. Both severe LV impairment and hyperdynamic LVEF (≥70%) are independently associated with worse outcomes [44, 53, 54].

4. In patients with impaired LV contractility, the pattern of impairment, whether global or regional, should be described. Whilst the presence of left ventricular regional wall motion abnormalities should prompt consideration of coronary artery disease with or without acute coronary syndrome, it is not uncommon for critically ill patients with LV systolic impairment of a ‘non-cardiac’ aetiology to present with regional impairment which may or may not follow a coronary pattern [55].

5. Conditions which affect right ventricular contractility can produce different patterns of impairment [56]. Relying on a one-dimensional measure of contractility in a single plane (such as Tricuspid Annular Plane Systolic Excursion – TAPSE) may be insensitive, and risks missing patients with right-heart dysfunction [57, 58].

6. Right Ventricular Fractional Area Change (RVFAC) assesses change in RV cavity during systole in more than one plane [59]. However, this technique still does not consider the contribution from contraction at the RV infundibulum and suffers from inter-observer variability [60].

Assessment for dynamic left ventricular outflow obstruction

1. Echocardiographers should actively search for the presence of dynamic left ventricular outflow tract obstruction (LVOTO) (Fig. 3). Critical illness, and the drugs used to support the circulation during critical illness, may result in the combination of hypercontractility and a low LV cavity volume, which predisposes to the development of left ventricular outflow obstruction in patients with or without underlying structural abnormalities [61].

• 2LVOTO may also occur in the context of overall impaired LV systolic function, when impairment is regional (for example in an anterior myocardial infarction, or a stress cardiomyopathy with a classical takotsubo pattern of regional wall motion abnormalities – in both scenarios there may be basal septal hypercontractility).
• 3If LVOTO is not recognised there is the potential for vasoactive infusions to be increased, paradoxically worsening the shock state [62].

Fig. 3.

Continuous wave (CW) Doppler trace across the left ventricular tract (LVOT), demonstrating obstruction due to systolic anterior motion of the anterior mitral valve leaflet (SAM). The patient collapsed with chest pain during a dental procedure. Echocardiography revealed basal septal hypertrophy, impaired left ventricular systolic function (overall LVEF 49%) in a pattern consistent with a takotsubo pattern stress cardiomyopathy (with apical akinesia and basal hyperkinesia), leading to dynamic LVOT obstruction

Assessment for acute ventricular septal defects

1. Acute ischaemic ventricular septal defects (VSD) are a rare but serious complication of myocardial infarction (occurring in ~ 1 in 500 cases of STEMI), with short-term mortality approaching 100% if untreated [63]. In patients who have late haemodynamic collapse in the days to weeks after acute MI it is essential to evaluate the entirety of the inter-ventricular septum using colour Doppler, from all available views with a special focus on the apex of the ventricles (apical VSDs are more commonly seen with anterior infarction, whilst inferior and lateral infarcts are more likely to cause basal defects [64]) (Fig. 4).

Fig. 4.

Modified apical 4 chamber view demonstrating an ischaemic ventricular septal defect (VSD), with left to right shunt

Considerations when assessing valvular function

1. Chronic valve disease can complicate shock of any aetiology [65].

2. Echocardiographers should be aware that changes in the patient’s haemodynamic state and loading conditions may alter the appearance of different valvulopathies, and hence severity grade. For example, low LV systolic pressure can occur in shock of any cause, and in a patient with mitral regurgitation the regurgitant volume will fall as LV-LA systolic pressure gradient falls [66].

3. Diagnosis and severity grading of aortic stenosis may prove challenging in the shock state where cardiac output is reduced. In the presence of a morphologically abnormal valve on 2D imaging an aortic valve area (AVA) <1 cm² as assessed by the continuity equation, or a dimensionless index (DI) <0.25 are suggestive of severe AS [67].

4. Conversely the severity of aortic stenosis may be over-estimated in states of increased flow (such as sepsis with a hyperdynamic circulation or with the use of inotropes) [68].

5. Acute aortic regurgitation (AR) may complicate type A aortic dissection or infective endocarditis, and produces haemodynamic instability by reducing the effective cardiac output and raising left ventricular end diastolic pressure. In the acute setting the left ventricle will not have undergone compensatory dilatation, therefore the ejected volume will not be elevated and the wide pulse pressure seen in chronic AR may be absent [69]. If left ventricular end-diastolic pressure exceeds left atrial pressure, then the mitral valve will close prior to the end of diastole, this may present as low velocity diastolic mitral regurgitation [70].

6. Chronic mitral stenosis may decompensate in conditions which result in tachycardia or the acute loss of sinus rhythm, since the reduced time available for diastole hampers both LV filling, and LA emptying, raising pulmonary pressures. Whilst tachycardia typically leads to an increase in trans-valvular mean pressure gradient, this value can be misleadingly low when the patient is in a low output state [71]. Direct observation of the valve and planimetry of the MV orifice become of paramount importance in altered loading conditions.

7. Common causes of acute mitral regurgitation include leaflet perforation complicating infective endocarditis, papillary muscle rupture complicating myocardial infarction and chordal rupture complicating myxomatous MV disease. The severity of mitral regurgitation may be under-estimated unless a thorough search for eccentric jets is conducted. The left atrial cavity size can be normal, as there has not been time for compensatory dilatation. The continuous wave Doppler envelope may be triangular, rather than parabolic, as the non-compliant and not yet dilated LA experiences a rapid rise in volume during systole [72].

8. Significant volume tricuspid regurgitation is usually functional, occurring in the context of right-heart chamber dilatation [73]. When the volume of regurgitation is great or the right ventricle is failing, rapid equalisation of right ventricular and right atrial pressure in systole can lead to a low velocity continuous wave Doppler envelope [73]. Severe TR can lead to organ dysfunction due to venous congestion [74, 75], therefore when present echocardiographers should evaluate venous drainage of the liver to look for blunting or reversal of forward flow during ventricular systole in the hepatic veins, using pulsed wave Doppler [73].

Assessment of low preload states

1. Low preload may be seen in both hypovolaemic shock, and distributive shock where vasodilation leads to blood pooling within the venous system [3]. There are many pathological processes which may lead to vasoplegia and distributive shock, including: cytokine release; catecholamine resistance; corticosteroid deficiency; drug toxicity; loss of sympathetic tone.

2. Echocardiography alone is likely to be inadequate in differentiating between these different causes of shock, and the echocardiogram must be incorporated into the wider clinical assessment.

3. As circulating volume decreases, so does end-diastolic LV cavity size [76, 77]. The range of LV cavity sizes within the normal reference range in healthy individuals is wide [24] and the range of baseline cavity sizes will be wider still when including individuals with chronic cardiac disease [76]. This makes it difficult to propose a single cut-off value for LV size that reliably diagnoses hypovolaemia. In the correct clinical context, and in the absence of an alternative cause, very low LV cavity size (LV end-diastolic area < 10 cm² and/or LV end-systolic area < 5 cm²) is often reported to be consistent with marked hypovolaemia [78–80].

4. Low inferior vena cava (IVC) end-expiratory calibre has been proposed as a marker of hypovolaemia. There is some evidence that suggests a correlation between right atrial pressure and inferior vena cava calibre, however there are many confounding factors that influence IVC calibre, beyond intra-vascular volume [81].

5. There is evidence that excessive IVC calibre change across the respiratory cycle may help identify hypovolaemic patients. However, the evidence is inconsistent [82, 83]. When reporting inferior vena cava calibre change over the respiratory cycle, as a sign of volume status, the work of breathing or level of ventilatory support should also be reported (as the magnitude of intra-thoracic pressure and volume changes will influence the magnitude of IVC calibre change).

6. In select patients there is the potential for echocardiography to be helpful in predicting if a bolus of intravenous fluids is likely to result in an increase in stroke volume and cardiac output. A description of this ‘fluid responsiveness’ assessment and algorithm are provided in the supplementary materials (Appendix D).

Assessment of obstructive shock

1. Obstructive shock can be defined as a physical obstruction to blood flow and can occur within different locations:

   • Within the heart, for example due to an intra-cardiac mass.
   • Within the extra-cardiac circulation, for example due to pulmonary embolus.
   • Due to external compression, for example cardiac tamponade (Fig. 5) and tension pneumothorax.

2. Echocardiographic clues will largely depend upon the specific pathology and are described in the supplementary materials (Appendix A).

Fig. 5.

Apical 4 chamber viewing demonstrating systolic right atrial collapse in a patient with a malignant pericardial effusion. Pulsed wave (PW) Doppler across the left ventricular outflow track revealed stroke volume variation of ~120% (SV varied from 55 ml to just 14 ml). Cardiac output was just about maintained (4.3 L/min), but only in the context of sinus tachycardia

Co-existent shock states

1 Echocardiographers should be mindful that the assessment of the shocked patient is made more challenging as some diseases may result in multiple forms of shock co-existing. A classic example of this is septic shock, which may result in distributive shock due to vasoplegia, hypovolaemia due to capillary leak and cardiogenic shock due to myocardial depression. The relative contribution of these different forms of shock may shift during the course of the patient’s disease, as the sepsis state progresses and resolves [84].

Question 3: Is left atrial pressure likely to be elevated?

Summary

• Two key questions may be addressed separately:

  • ○ What is the current relaxation capacity of the ventricle?
  • ○ Is the LAP likely to be normal or elevated?

• Septal and/or lateral early diastolic mitral annular velocity (e’) below age and sex specific cut-offs should be reported as ‘evidence of impaired left ventricular relaxation’.

• The ratio of velocities, of early diastolic blood flow and average early diastolic mitral annular velocity (E/e’) correlates with LAP reasonably well in general ICU patients (including patients with septic shock). A value of > 14 may be considered as ‘evidence of elevated left atrial pressure’ in these patients.

• E/e’ correlates with LAP less well in patients with cardiogenic shock, is potentially misleading in these patients and should not be relied upon as a stand-alone measure.

• In the absence of mitral stenosis, trans-mitral E wave Deceleration Time (E DT) ≥150 ms may be considered ‘evidence that left atrial pressure is not elevated’, if consistent with other findings.

• When the E/A ratio can be measured, a value > 2 may be considered as ‘evidence of elevated left atrial pressure’ in patients with left ventricular impairment.

• LA cavity size may be normal when LAP has risen acutely: normal LA volume is not necessarily evidence of normal LAP.

• In patients with chronic LA dilatation, the LA is likely to remain dilated during acute falls in LAP, for example in hypovolaemic shock. Elevated LA volume is therefore not necessarily evidence of elevated LAP if the patient is hypovolaemic.

• Elevated TR Vmax (>2.8 m/s) in this setting should prompt a search for pulmonary and/or right heart pathology rather than being immediately attributed to raised LAP.

• Echocardiographers trained in lung ultrasound should perform dedicated lung views and report the presence of b-lines and pleural effusions.

Considerations when assessing left ventricular relaxation and LAP

1. Guidelines designed to assess the diastolic function of the left ventricle are principally concerned with assessment of stable out-patients presenting with the clinical signs and symptoms of heart failure [85–87].

2. The validity of the 2024 BSE guidelines [86] on diastolic dysfunction in patients with undifferentiated shock has not yet been tested. The 2016 joint ASE/EACVI (American Society of Echocardiography/European Association of CardioVascular Imaging) guidelines [85] have been tested in an ICU setting [88]. In approximately half of patients the algorithm was indeterminant, however, when patients were classified (as having normal or elevated LAP), specificity and sensitivity were both 74%. The 2016 joint ASE/EACVI guidelines [85], 2024 BSE guidelines [86] and 2025 ASE guidelines [87] have been tested in patients with cardiogenic shock following myocardial infarction, many of whom were supported with mechanical circulatory support [89]. In this context all three guidelines were found to be insensitive in identifying patients with elevated Pulmonary Capillary Wedge Pressure (PCWP).

3. Shocked patients may have pre-existing diastolic dysfunction, or may develop impaired LV relaxation as a component of their acute illness [90, 91]. It is reasonable to assess septal and lateral early diastolic mitral annular velocities (e’), for evidence of impaired LV relaxation (Table 3, Fig. 6).

4. Determining whether left atrial pressure (LAP) is elevated can be challenging in haemodynamically stable patients, and even more challenging during the dynamic changes in loading conditions that occur during acute illness [47]. Nevertheless, determination of whether the LAP is likely to be raised or not is a useful guide for clinicians caring for the patient, since this is a common and significant cause for pulmonary congestion and right ventricular dysfunction [92, 93].

5. Multiple studies have investigated the relationship between echocardiographic measures found within international guidelines, and LAP, in critically ill patients [94]. The ratio of trans-mitral E wave maximal velocity to early diastolic mitral annular velocity (E/e’) has been shown to have reasonable to good correlation with invasively measured PCWP in general ICU patients (including patients with septic shock) [95–99]. However, the correlation was poor when assessed in patients with (predominantly ischaemic) cardiogenic shock [100–102]. As this measure is one of the initial steps in both the BSE and ASE/EACVI diastolic function guidelines [85–87], it will be familiar to all echocardiographers performing comprehensive studies and should be reported in shocked patients, but not be relied upon in isolation.

6. Trans-mitral E wave Deceleration Time (E DT) has been shown to correlate with PCWP in general ICU patients [99, 103]. E wave DT values > 100-120 ms suggest raised PCWP is unlikely. In one study it aided discrimination between normal and elevated PCWP when E/E’ was within the ‘grey zone’ of values between 8 and 15 [97]. Current BSE guidance [86] recommends a cut-off of 150 ms to help discriminate between normal and elevated filling pressures.

7. It will frequently be impossible to measure the E/A ratio in shocked patients, as tachycardia and high-normal heart rates (leading to merging of E and A waves) and atrial fibrillation are relatively common. When the E/A ratio can be assessed in critically ill patients, then patients with elevated PCWP have been shown to have higher E/A ratios [88, 95, 97, 99, 103]. This measure is likely to be most useful in patients with impaired LV function.

8. Left atrial volume increases in response to chronically elevated LAP, and therefore in acute illness there will not have been time for the LA to dilate [102].

9. The position and motion of the inter-atrial septum (IAS) has been shown to reflect the pressure gradient between the left and right atria. It is usual for the mobile IAS to transiently bow from right to left during mid-systole, but this may be lost when LAP is elevated. If mid-systolic bowing towards the left atrium is observed, then it is unlikely that LAP is significantly elevated in the absence of raised RAP [104].

10. Whilst tricuspid regurgitation maximum velocity is elevated in patients with elevated pulmonary pressures due to left-heart disease (indicating post-capillary pulmonary hypertension), this measure alone does not discriminate between patients with elevated LAP and patients with pulmonary hypertension from any other cause [105], such as patients mechanically ventilated with ARDS (Acute Respiratory Distress Syndrome) [106].

11. Other measures that are found within recently published diastolic function assessment guidelines [86, 87] (for example LA strain) have been less exhaustively investigated in acutely unwell patients [94], and therefore their utility is less well understood.

12. Whilst it does not form part of the BSE’s minimum dataset, some echocardiographers may be experienced in performing lung ultrasound (US). Lung US can be performed rapidly and can reliably identify pleural effusions as well as signs of pulmonary congestion. In the correct context the presence of ≥ 3 ‘b-lines’ in multiple lung regions, is sensitive and specific for increased extra-vascular lung water, as seen in pulmonary congestion. However, this technique does not discriminate between pulmonary oedema and other causes of ‘interstitial syndrome’, including interstitial pneumonia, pneumonitis and diffuse parenchymal lung disease (Fig. 7) [107, 108]. When the echocardiographer is trained in lung US, this should be performed and reported to supplement their assessment [6].

Table 3.

Septal and lateral e’ values suggestive of impaired LV relaxation (cm/s), as assessed using tissue Doppler

		18–40 years	41–65 years	>65 years
Men	Septal	<7.0	<5.0	<4.0
	Lateral	<9.0	<6.0	<5.0
Women	Septal	<8.0	<5.0	<4.0
	Lateral	<11.0	<6.0	<5.0

Reproduced from: Robinson S, Ring L, Oxborough D, Harkness A, Bennett S, Rana B, Sutaria N, Lo Giudice F, Shun-Shin M, Paton M, Duncan R, Willis J, Colebourn C, Bassindale G, Gatenby K, Belham M, Cole G, Augustine D, Smiseth OA. The assessment of left ventricular diastolic function: guidance and recommendations from the British Society of Echocardiography. Echo Res Pract. 2024 Jun 3;11(1):16. doi: 10.1186/s44156-024–00051-2. PMID: 38825710; PMCID: PMC11145885

Fig. 6.

Tissue Doppler imaging (TDI) in a young patient who has developed septic cardiomyopathy following chemotherapy. There is evidence of both systolic and diastolic impairment (with the early diastolic mitral annular velocity falling below the cut-off for the patients age and sex)

Fig. 7.

Dedicated lung ultrasound of the right hemithorax with the curvelinear probe held in a transverse orientation (posterior chest to the left of the image). There is a moderate pleural effusion with confluent b-lines within the aerated lung

Question 4: Is pulmonary artery pressure and/or pulmonary vascular resistance likely to be elevated?

Summary

• Using the standard BSE flow chart to assess the probability of PH [105] is a reasonable first step during the assessment of right heart pressures in the shocked patient.

• When there is overt RV systolic failure or severe TR, TR Vmax may be below the usual diagnostic thresholds for PH.

• TR Vmax > 4 m/s suggests chronically elevated PVR (and is unlikely to be purely acute).

• PAAT < 105 ms is suggestive of elevated PVR when consistent with other findings.

• In patients with tachycardia, PAAT should be corrected for heart rate.

• When present, right-to-left displacement of the interventricular septum, as well as early systolic paradoxical movement of the IVS should be reported.

• The presence of RV dysfunction should be reported, using quantitative measures when image quality allows.

Causes of elevated PAP and PVR in the shocked patient

1. The leading causes of chronic pulmonary hypertension (PH) in the UK are left heart disease and chronic lung diseases [109, 110]. If left untreated the right heart, pumping against increased pulmonary vascular resistance (PVR) will adapt [105, 111]. Whilst RV systolic function is initially maintained, RV failure occurs in end-stage disease [111, 112]. In patients with chronic PH, shock may result from disease progression or represent decompensation of stable disease due to intercurrent illness [112].

2. In the critically unwell patient acute cor pulmonale (defined as RV failure as a consequence of an acute rise in PVR) may result through a number of different mechanisms. Hypoxic pulmonary vasoconstriction occurs in many acute severe respiratory conditions (for example pneumonia), and is an adaptive response directed at maintaining ventilation-perfusion matching by reducing flow to poorly ventilated lung [113]. Pulmonary embolism can lead to a rise in PVR through a combination of mechanical obstruction and the release of vasoconstricting mediators taking affect in the pulmonary vascular bed [114, 115]. Mechanical ventilation at higher lung volumes can raise PVR, through external compression of blood vessels [116, 117].

Considerations when assessing PAP and PVR

1. Current guidance on the echocardiographic risk assessment of PH centres on measurement of the TR jet maximum velocity (TR Vmax). In the absence of obstruction to flow leaving the RV, this correlates with pulmonary artery systolic pressure (PASP) [105]. Reasonable correlation has also been demonstrated in patients with cardiogenic shock [101], and this measurement should be made and reported.

2. TR Vmax may fall either due to severely impaired RV contractility or as TR severity increases to the point of being essentially free flowing (at which point the RV-RA pressure gradient is not maintained during systole) [118].

3. Whilst patients with chronic PH might experience very high pulmonary pressures, potentially surpassing systemic blood pressure, the untrained right ventricle, is unable to generate such high pressures [119, 120]. Therefore, when accurately measured TR Vmax exceeds 4 m/s, it is unlikely that the elevation in PVR is entirely acute [120, 121].

4. The pulmonary artery acceleration time (PAAT) shortens in patients with elevated pulmonary pressures [122] and helps distinguish between patients with pre-capillary PH with increased PVR, when the PAAT is expected to be short and exclusively post-capillary PH with normal PVR, where the PAAT is more likely to fall within normal limits [123]. PAAT < 105 ms is considered short [105, 124], and a subset of patients with short PAAT develop characteristic mid-systolic notching due to the reflection of elevated PVR back into the pulmonary artery [125, 126] (Fig. 8).

• 5Amongst critically ill patients the relationship between PAAT and estimated PAP is weak to modest [127–129]. PAAT < 105 ms can be considered as suggestive of elevated PVR, when consistent with other findings. In individuals with and without PH, PAAT shortens as HR increases [130]. Given tachycardia is a common finding in sick patients, when PAAT is <105 ms and HR exceeds 100 bpm, then PAAT should be corrected for heart rate [105, 128].
• 6Elevated PASP may be evident through right-to-left displacement of the inter-ventricular septum (IVS) during late systole. In healthy individuals LVSP exceeds RVSP throughout all of systole, and the IVS is pushed from left-to-right, leading to a circular cross section of the LV, and a crescentic RV, as viewed in the PSAX view. When PAP is elevated the duration of RV contraction increases, such that the RV continues to contract after the LV has begun to relax. This leads to a brief period at the end of systole when RVSP exceeds LVSP, and the IVS is pushed from right to left [131]. If the RV filling pressures are also elevated, then this right-to-left displacement continues throughout diastole, and it is only at the onset of systole when the LV pressure rises sharply, that the IVS pushes back towards the right. This pattern is described in patients with acute cor pulmonale and may contribute to shock by impairing LV filling [132]. The presence of RV dilatation should be described. In addition, when there is displacement of the IVS this should be noted and temporally described in relation to the period of the cardiac cycle. It may be described qualitatively or quantitatively (using the eccentricity index) [133].
• 7The RV and the PA are a ‘coupled unit’: when PAP and PVR rise, RV contractility must rise in concert to maintain RV ejection, known as coupling [134]. The ratio of TAPSE to PASP (as estimated by applying the simplified Bernoulli equation to the TR Vmax measurement and adding an estimated RAP) has been proposed as a surrogate measure of RV-PA coupling [135], with a low ratio indicating a failing RV in the context of PH. Low values of TAPSE/PASP have been shown to correlate with worse outcomes in a number of acute presentations including acute pulmonary embolus [135] and ARDS secondary to COVID-19 [136]. Different studies (investigating different patient populations) have identified different optimal cut-off values for predicting poor outcomes [137]. At present there is no universally agreed ‘normal’ range and interpretation of values must be context specific.
• 8Different pathologies result in different patterns of RV dysfunction: for example, in post-operative cardiac surgery patients the longitudinal shortening of the RV is disproportionately reduced, whilst in patients with compensated (coupled) PH, longitudinal function is relatively preserved, with a reduction in transverse motion [138–140].
• 9Akinesia of the mid RV free wall with preserved contractility at the apex (‘McConnell’s sign’) is sometimes seen in acute pulmonary embolus (PE) and particularly in haemodynamically unstable patients [141–144]. This sign is not specific to PE, but most often described in this context [145–148].

Fig. 8.

Pulsed wave (PW) Doppler trace taken from the right ventricular outflow track (RVOT), showing short pulmonary acceleration time (PAAT), that remained short when corrected for heart rate, and notched deceleration, consistent with elevated pulmonary vascular resistance. This was an unexpected finding in a patient admitted to ICU with diabetic ketoacidosis (DKA), who deteriorated 24hrs after admission. A CT pulmonary angiogram revealed massive pulmonary embolus and the patient underwent catheter directed thrombolysis with good effect

Question 5: How is the situation evolving?

Summary

• ‘Follow-up’ studies may be required whenever the patient’s condition changes significantly.

• ‘Follow-up’ studies may be used to assess the impact of a therapeutic intervention.

• ‘Follow-up’ studies may be abbreviated to answer a specific clinical question.

• ‘Follow-up’ studies should refer to previous studies and highlight any changes.

• Intra-/inter-observer variability should be considered when comparing measures made between different studies.

• In shocked patients in whom new (or presumed new) cardiac dysfunction is identified, it seems prudent to perform a convalescent study following resolution of the acute illness.

Considerations when performing follow-up studies

1. An individual patient’s haemodynamic status may go through many changes during the course of their critical illness. For example, a patient presenting to hospital with bacterial sepsis may be hypovolaemic at the point of first presentation, but following initial fluid resuscitation they may enter a period of hyperdynamic cardiac function as vasoplegia becomes the dominant form of shock. As their illness progresses, they may develop septic cardiomyopathy with reduced ventricular function leading to elevated pulmonary and systemic pressures and congestion [149]. Serial echocardiograms may be required to redefine the principal forms of shock throughout a single episode. In rapidly evolving shock, echocardiographic findings may vary hour to hour.

2. In critically ill patients who have undergone a comprehensive echocardiogram whilst shocked, it may be appropriate for subsequent studies to be targeted, tailored to the patient’s specific pathology: for example, a patient with acutely deteriorated left ventricular function may require multiple focused assessments of ventricular function over the coming days to weeks. The scope of a focused assessment must widen if new complications become evident.

3. Echocardiographic measurements (and visual assessments) are subject to a degree of both intra-observer and inter-observer variability [150–152]. When serial studies are performed the echocardiographer should consider if the apparent change is genuine, or reflects variability between measurements. One study, performed in acutely unwell patients, attempted to calculate the Least Significant Change (LSC) that would need to be observed, to be confident that observed change was genuine [153]. Reported estimates for LSC between studies performed by different operators included: LVEF, ±8%; LVOT VTI, ±14%; E/e’, ±21%; TAPSE, ±17%.

4. Many patients who develop acute ventricular failure in the context of an acute extra-cardiac illness are likely to experience partial or complete recovery of function in the short-term, but this is not universal [154, 155]. The long-term consequences of critical illness associated cardiomyopathy require further study, however it would seem prudent to perform a convalescent study following resolution of the acute illness, as patients with persisting cardiac dysfunction are likely to require further follow-up [156].

Electronic supplementary material

Below is the link to the electronic supplementary material.

Authors’ contributions

Both authors contributed to the preparation of the entire manuscript.

Funding

No funding was received.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not required.

Competing interests

The authors declare no competing interests.