An individual patient’s starling curve depends on their contractility (systolic function) and so it is meaningless to measure preload surrogates to assess fluid responsiveness (preload is end diastolic wall tension and not LV pressure or volume). It is now well established in the evidence base that static parameters like CVP, PAOP and ventricular volumes are very poor predictors of fluid responsiveness.
There are however some extremes of static parameters seen by echo which are suggestive of fluid responsiveness:
A very small (<10mm) or large (>20mm) IVC may have some relation to fluid responsiveness but has a poorer positive predictive value than other means.
A restrictive pattern in MV flow is suggestive of fluid unresponsiveness as it reflects elevated LVEDP. It should be remembered however that if LV compliance is reduced the LVED PV relationship is shifted up and left so:
- The LV can be under filled (fluid responsive) despite high filling pressures.
- The optimum filling range is narrow (it is under or overfilled easily).
Therefore a hypovolaemic LV with diastolic dysfunction may have elevated filling pressures, may respond well to fluid, but will easily be overloaded with pulmonary oedema resulting.
It should be noted that a MV inflow pattern of impaired relaxation / mild diastolic dysfunction (E/A <1, DT prolonged >200 ms) can be caused by hypovolaemia.
See diastolic dysfunction and pulmonary oedema for more info.
The size of the LV is predictive of fluid responsiveness only when it is very small (LVEDarea PSAX <10cm2 or kissing ventricles) and hyperkinetic. Tavernier et al, Feissel et al.
This appearance will be mimicked by LV hypertrophy so should be used with caution.
Alveolar pressure – pressure in the lung. It is influenced by lung and chest wall compliance. In SV it is zero with no flow (end inspiration and end expiration) and negative in inspiration. In IPPV it is positive with its max at end inspiration when flow has stopped (Pplat).
Intrathoracic pressure (ITP) – pressure in the chest. Influenced by alveolar pressure, chest wall compliance and force of respiration (if SV). Negative in SV driving inspiration and venous return. Mostly positive in IPPV especially with PEEP.
Transpulmonary pressure (TPP). Alveolar pressure minus intrathoracic pressure. The distending pressure of the lung. It acts on the pulmonary capillaries.
Intravascular pressure – directly measurable.
Transmural pressure (TMP) – distending pressure of a vessel/chamber. It is intravascular pressure minus pressure outside the vessel/chamber ie pericardial/intrathoracic.
In cardiac tamponade the intravascular and pericardial pressure will both be high meaning the transmural pressure is very low and the heart is empty. As pericardial fluid is removed both pressures fall but pericardial pressure much more so thus increasing the transmural pressure and filling of the heart.
In inspiration in IPPV, intrathoracic pressure increases more than RA pressure (both increase by transmission) thus decreasing transmural pressure and reducing venous return. In expiration transmural pressure increases with increased venous return.
The effect on SV is delayed resulting in the reverse pulsus paradoxus of IPPV. Systolic BP and SV increases in inspiration and reduce in expiration. The increase and decrease are termed deltaUp and deltaDown and are determined by changes in intrathoracic and transpulmonary pressures.
In positive pressure inspiration:
- Increased ITP reduces venous return (collapse of the SVC in hypovolaemia will exaggerate this) and
- increased TPP increases RV afterload. Increased TPP increases PVR and can cause partial or complete collapse of pulmonary capillaries (West Zones). From a certain pressure, small increases in pressure can result in large increases in PVR accounting for the sensitivity of the RV to increased plateau pressure in ARDS.
Both of these reduce RVSV. This causes reduced LV preload and reduced LVSV a few heartbeats later.
Decrease in venous return will play the biggest role in hypovolaemia (with normal or mildly impaired lung compliance).
Increased afterload will dominate if significantly reduced lung compliance (with normovolaemia).
So, while both mechanisms lead to the same result, the cause and appropriate treatment are different. Fluid is indicated if there is reduced venous return whereas it could be detrimental with increased TTP.
In RV failure the increased RV afterload will produce an exaggerated dDown giving rise to false positives for fluid responsiveness.
Increased TPP on the pulmonary capillaries in inspiration will force blood into the L heart and increase LVSV.
A decrease in afterload also occurs with increased ITP (increased pressure gradient for emptying) and also contributes to dUp.
A failing LV may produce a significant dUp from its improved performance with the reduced afterload which can contribute to a false positive for fluid responsiveness.
Biventricular failure will have an exaggerated dDown and dUp causing false positives for fluid responsiveness.
Cyclical changes in intrathoracic pressure from IPPV induce cyclical changes in preload and SV if the ventricles are on the steep ascending part of the starling curve ie if they are fluid responsive.
Aortic Vmax or LVOT VTI changes (12% and 20% respectively) predict fluid responsiveness with high sensitivity and specificity.
Note that patients must be making no respiratory effort, have a TV of >7mls/kg and must be in sinus rhythm. RV or LV failure can also result in false positives due to afterload changes.
Passive leg raising (which transfers about 300mls of blood into the central circulation) can be used in spontaneous ventilation and arrythmias. >12% increase in SV predicts responders with a sensitivity of 77% and specificity of 100% (Lamia et al). Another study (Maizel et al) showed sensitivity and specificity figures of 69% and 89%.Measurements should be taken before and 1 minute after the PLR.
PLR should be done by tipping the whole bed from 45 up to 45 down and not lifting the legs which can compress the femoral veins.
PLR can of course be used in IPPV too. Its elegance is that it is instantly reversible so causes no harm from fluid administration if the patient is not fluid responsive.
PLR can be falsely negative in severe hypovolaemia (reduced blood volume in the legs).
(max – min / mean) x 100
IVC size is correlated to CVP/RAP. In IPPV an RAP of <10 can be assumed if the ICV is <12mm.
In controlled ventilation IVC will expand in inspiration (as venous return is reduced) and reduce in expiration (opposite of spont).
Absence of respiratory variation means approx 90% chance will not be fluid responsive.
>12% variation identifies responders (with SR and TV >8mls/kg). >18% if max-min / min used
IVC collapsibility index = max diameter – minimum diameter / mean diameter
x100 to get percentage
Variation in SVC size (36%) can be used with TOE.
During surgery changes in LV size (a baseline must be measured) reflects changes in preload (this then becomes a dynamic parameter).
Variation in LV stroke area (trace area of LV in PSAX papillary level in systole and diastole and minus systole from diastole – see how this changes with respiration) with respiration has been shown to predict fluid responsiveness (change >16%). Impractical without appropriate software in machine.
Could perhaps look at variation in LVEDarea with respiration (not sure if this has ever been done).
Echo measurement of stroke volume
Flow is often not used to guide resuscitation. It is more time consuming to measure than standard parameters, requires more technical resources and based on numbers alone it is not possible to know what value will achieve adequate perfusion for an individual patient. ScVO2 and lactate are useful but only tell you about a large mismatch between supply and demand and not more subtle shock. MAP is of course proportional to flow with a fixed SVR but flow can be significantly compromised if SVR is high (too much noradrenaline). Many critically ill patients have received resuscitation fluid and are on vasopressors. The question of whether they need more fluid or more vasopressor can be difficult to answer.
As seen above, dynamic parameters are extremely useful for predicting fluid responsiveness. The gold standard however must be whether SV actually increases with a fluid challenge.
Assessment of the response in flow to a fluid challenge is easy with echo. It is really only necessary to measure LVOT VTI rather than adding in LVOT measurement to get SV.
Measurement of VTI should occur before and after a fluid challenge or passive leg raise. This method overcomes the limitations of assessment of fluid responsiveness.
Measurements should ideally be end expiratory or averaged over several consecutive beats.
SV should also be assessed if pulmonary hypertension is causing RV failure. Reducing PVR should result in an increase in RV SV which will in turn increase LV SV.
It should be noted that significant improvements in SV may not be reflected by changes in MAP but systemic blood flow and oxygen delivery will be improved.
When to stop giving fluid
If aortic VTI increases with a fluid challenge then the patient is fluid responsive (this may or may not mean further fluid is indicated depending on the clinical condition of the patient).
When VTI no longer increases with fluid then the patient is unresponsive and fluid administration should stop.
LV filling pressures
When to stop a fluid challenge can be guided by changes in PAOP.
MV doppler indices such as E velocity and E/A ratio are preload dependent.
Hence an increase in the E velocity or E/A ratio without an increase in aortic VTI or SV suggests that further fluid will be detrimental.
A dilated RV should prompt extreme caution in fluid administration.
Some patients with mild to moderate RV dilatation and failure may tolerate fluid loading. The more severe it is the less likely they are to tolerate fluid.
Fluid loading is contraindicated in the presence of ACP with septal dyskinesia.
Remember that RV failure can result in false positives in signs of fluid responsiveness (see above).
Failure to increase aortic VTI or SV with fluid or PLR with RV dilatation means no more fluid should be given.
A small hyperdynamic LV with a small IVC suggests significant hypovolaemia.
More subtle hypovolaemia should be sought by signs of fluid responsiveness.
If minimal training use IVC size variation from the subcostal view.
If more experienced use LVOT VTI variation with a PLR.
An A-line trace will of course provide dyanamic indices of fluid responsiveness (systolic pressure, pulse pressure and stroke volume variations). Echo will give additional information such as ventricular function (which can lead to false positives for dynamic indices) and can be used to accurately measure flow (VTI, SV or CO). It will also reveal other pathology such as valve disease, tamponade, signs of PE etc.
Echocardiography should therefore be mandatory for the assessment of shock.