Echo should be mandatory for assessment of circulatory failure.
Need to firstly identify clear reversible causes e.g. tamponade, severe valvular pathology, severe LVSD.
Secondly more intricate assessment e.g. fluid responsiveness.
Haemorrhagic, cardiogenic, distributive and obstructive shock all demonstrate individual cardiac physiology with variable risk of inadequate or excessive fluid resuscitation.
End of the bed clinical suspicion that volume loading will improve cardiac output is correct only half of the time in critically unwell.
Echo can be used to accurately describe a patient’s position on the Frank-Starling curve.
Fluid responsiveness is defined as:
…an increase in cardiac output or stroke volume of 10-15% following delivery of 500ml of fluid over 15 minutes.
Fluid responsiveness is not the same as being intravascularly deplete as patients with low central blood volume may not be responsive if the ventricle is failing.
During inflammatory response the damage to the endothelium increased permeability and leak of water. This manifests as:
- Pulmonary capillary endothelium causing increased extravascular lung water in relation to delivered volume. Increases duration of ventilatory support.
- Capillary endothelium glycocalyx layer is damaged by inflammatory response and further fluid administration strips the glycocalyx leading to widespread endothelial dysfunction and formation of oedema.
- Neurological, renal and gastrointestinal functions impaired by over-resuscitation.
Static parameters of fluid responsiveness:
Measurement taken without disturbing the system in one point of time.
Position on Starling curve determined by contractility and so use of preload surrogate markers is not useful.
Historically static makers such as CVP, RAP, pulmonary artery occlusion pressure and measurement of left or right end diastolic dimensions are poor predictors of fluid responsiveness.
Very small (<10mm) or large (>20mm) are commonly used measures but are not reliable. Extremes may be suggestive of fluid responsiveness.
Intra-abdominal pressures often have independent effect off IVC calibre which is not related to fluid status.
Exaggerated respiratory effort can also have more profound effect on IVC respiratory variability.
Evidence of high left ventricular diastolic/atrial pressures may support lack of benefit from fluid loading.
Measured with MV inflow doppler, TDI & pulmonary vein flow.
Though several issues with this – not substantiated in literature.
If LV compliance reduced then LVEDPV loop is shifted upwards and left:
- LV may be underfilled and fluid responsive despite high filling pressures.
- Narrow filling range – under or over filled easily.
Additionally – MV inflow pattern suggesting mild diastolic dysfunction (E/A reversal, prolonged DT) may be caused by hypovolaemia.
Simple LV dimensions (PSAX EDA, LVIDd) may be reasonable markers at very low circulating volumes.
PSAX EDA <10cm^2 or kissing ventricles alongside hyperkinetic LV.
Key to exclude obstructive pathology in this setting.
LV end diastolic dimension may be small in hypovolaemia but this may be mimicked by LV hypertrophy and knowledge of LV size in health is required for this to be accurate.
LVED area may be more useful than simple dimension but prone to similar inaccuracies for above reasons. Also contributed to by myocardial stretch and capacitance and so is influenced by diastolic function.
May add supporting information but causes of dilated right heart are diverse and usually pressure not volume related.
Should not be used in isolation.
Presence of septal flattening suggests fluid would be poorly tolerated.
Suspected that presence of elevated LVOT velocities sometimes represent impaired flow that may be impoved by fluid loading.
Those with asymmetric ventricular hypertrophy are at risk of dynamic outflow tract obstruction including encroachment of aMVL into LVOT.
Those with lesser degrees of asymmetry may also be at risk in low-volume or hyperdynamic states – including borderline hypertrophy, proximal septal hypertrophy or an apparently normal ventricle that behaves abnormally under hypovolaemic or distributive shock conditions.
These patients may have a disproportionate benefit from fluid loading due to improved outflow tract dynamics.
Dynamic parameters are measurements taken during ‘provocation’ of circulation via:
- Interaction of the respiratory cycle.
- Change in posture to provide surrrogate fluid challenge.
Stroke volume variation with respiration:
Preload changes during respiratory cause cyclical changes in SV due to changes in LVEDV.
Heart-lung interaction and venous return:
Spontaneous inspiration – negative intrathoracic pressure increases RA filling, RV SV and 1-4s later (depending on CO) LV SV…
This causes breath to breath SVV of around 20-30% in health whilst breathing at rest.
During full mechanical ventilation (without any intrinsic effort) the effect on venous return are inverted and blunted.
Positive intrathoracic pressure during inspiration compresses heart and great vessels to a degree related to inflation pressures and vT.
Positive pressure therefore reduces RV filling due to:
- reduction in venous return.
- pulmonary vascular bed compression.
- compression of the RV.
The inverted variation in SV with respiration is usually 15-20% during mechanical ventilation in euvolaemic patient without cardiac compromise.
Therefore, in full mechanical ventilation, a low SVV of <10% suggests patient is near apex of Starling curve and therefore unlikely to be fluid responsive.
High SVV of >15% suggests great fluid responsiveness as on steeper part of curve.
10-15% is grey area and other features should be sought.
This is thought to have a sensitivity and specificity of 80%.
- Movement of heart during respiration may move the LVOT during respiratory cycle which artificially increases flow variation. Can temporarily alter ventilation settings to optimise if required.
- Arrhythmias cause large stroke volume variation and longer sample times are often not practical.
- Low or high tidal volumes cause changes in PVR and SV. Volumes <6ml/kg or >8ml/hr induce smaller or larger degrees os SVV. Briefly altering ventilation settings to within these parameters may be useful.
- Use of negatively chronotropic agents affects validity of SVV by increasing diastolic filling time.
Posture induced capacitance volume changes:
Passive leg raising introduces between 150-300ml of blood into the central circulation which is instantly reversible and without harm unlike fluid administration.
Performed either by lifting legs to 45 degrees or, more accurately, by sitting the patient up at 45 degrees then lying them down and raising legs to 45 degrees using bed.
Measure change in SV/CO using LVOT/aortic velocities.
Simply lifting legs may compress femoral veins which introduces inaccuracy.
Measurements should be taken 60-90 seconds after PLR (effect peaks at 90s).
12% increase in SV predicts responders with a sensitivity of 77% and specificity of 100% (Lamia et al).
A 5-10% rise in CO/SV is a reasonable indicator for fluid responsiveness.
10% strongly suggests fluid responder.
<5% are very unlikely to be responders.
If repositioning leads to pain, sympathetic stimulation may cause false positive. Additionally respiratory compromise due to position changes will compromise accuracy.
Central vein diameter variability:
Normal inspiration – degree of collapse of abdominal IVC and distension of thoracic SVC. These changes are reversed in PPV.
IVC size and respiratory variation cannot be reliably used for fluid responsiveness in presence of spontaneous respiratory effort.
If completely spontaneously breathing, size and collapsibility of IVC reflects only the RAP which is not useful for gauging fluid responsiveness (with the exception of a very small collapsed vessel suggesting fluid tolerance).
Assisted ventilation (triggering breaths), interpretation of IVC variability is not accurate.
Fully ventilated patients – degree of IVC diameter variation may be useful. The less the variation, the less likely the patient will respond well to filling.
- IVC diameter <1.5cm and varies >50% the RAP is usually <5mmHg.
- 2cm with <50% variability, RAP is likely to be >15mmHg which is not helpful for fluid responsiveness.
- Lack of resp variation means 90% will not be fluid responsive.
- 12% variation (sinus rhythm and vT >8ml/kg) identifies responders. Value is >18% for collapsibility index calculations.
Collapsibility index (or distensibility index if mechanically ventilated) may be calculated with:
IVC CI = max diameter – minimum diameter / mean diameter x 100 (for percentage)
Other limitations include high intra-abdominal pressures, high work of breathing.
Variation in SVC size of 36% may be used with TOE.
Practicalities of assessing fluid responsiveness:
- Is there frank hypovolaemia?
- Will CO improve with fluid – is there evidence of fluid responsiveness?
- Will fluid removal be tolerated?
Assessment of frank hypovolaemia:
- Hyperdynamic LV with low LVESV/papillary apposition. Correct for BSA. Use PLAX, PSAX & subcostal long axis.
- PSAX – CSA at mid LV – <10cm^2 or <5.5cm^2/m^2 corrected for BSA.
- Very small IVC – <1cm with insp collapse in non-ventilated pt.
1.5cm with exp collapse in ventilated pt.
- IVC that collapses by 50% or more, low EDV and lack of lung B-profile can be assumed to be at least volume tolerant.
Assessment of fluid responsiveness:
Assess either LVOT in A5C or A3C or RVOT in RVO or PSAX for best location to perform SVV or flow variation.
If mandatory ventilation – perform SVV/VTI variation – in reality VTI variation may be sufficient without measurement of LVOT to obtain SV.
If required – peak velocity of PW doppler can be used in place of VTI although will be slightly less accurate.
VTI variation or SVV (%) = [100 x (VTI max – VTI min)] / [(VTI max + VTI min) x 0.5]
PLR – prepare as above and optimise image as possible.
Perform baseline VTI envelopes averaging across a minimum of 3 waveforms.
Perform PLR or position change carefully but efficiently to mimic fluid bolus. Position should be held for 90-120s. Repeat measurement after 60-90 seconds.
Increment of >10% predicts fluid responsiveness.
Measure as above.
Distensibility index of ~18% of more predicts fluid responsiveness in ventilated patient without respiratory effort.
Direct measurement is recommended.
If all above measures result in uncertainty – small rapid fluid bolus (e.g. 100ml) may be given and repeat assessment performed.
Assessment for fluid removal:
High blood volume suspected e.g. pulmonary oedema, renal impairment and over-resuscitation.
The usual fluid responsive assessments as listed above may be used to cap fluid removal as suggest circulation has once again become fluid responsive.
Additionally using lung US to identify a B-profile is useful in identifying excessive extravascular lung water.
Not yet clear how to apply echo to this discipline, although identification of SVV of IVC variability through serial assessments may be useful.