Fluid responsiveness…will your echo help?! #FOAMed #FOAMcc #POCUS

The Conundrum…

Your patient is dry….or so the night docs tell you, having had it drummed into them for the past few hours into the morning by the nursing staff. How are you going to sort this, is there a bigger issue with the patient, do they need the magic pressers, just IV fluid…..or, do you reach for the good old Echo probe???

Those of us who are echo obsessed would obviously jump in with this modality immediately…the knobologists path! But are we doing this in preference to other things and what are it’s merits in assessment off a suspected volume depleted patient?

In critically ill patients at risk for organ failure, fluids can be friend or foe. We all want to increase cardiac output and oxygen delivery, but some patients just act up and you see worsened or no change in cardiac output, oedema, worsening ventilatory parameters and mortality!  This blog is based upon the fantastic review here...do have a look!

This makes reference to the spontaneously breathing patient on the ventilator…bare that in mind.


Looking at the way the clever heart works we know at constant cardiac contractility, low ventricular end-diastolic volume supplemented by a fluid bolus usually results in increased stroke volume. This immediately increases both cardiac output and oxygen delivery.

Overdo it with more IV fluids when they are at their plateau ventricular end-diastolic volume, the result; system overload!

What else are we looking for at the bedside?

  • reduction in lactate
  • Increase in urine output
  • Improvement in conscious level

All of these are signs of FLUID RESPONSIVENESS (defined as a 15% increase in overall CO).

Some fundamental physiology – Where are they on the curve?

Screen Shot 2016-12-07 at 14.43.10.png


In the original studies, preload was the term used for what was in the right atrium prior to the next contraction of the heart. Right atrial pressure was used as a surrogate of volume, also referred to as central venous pressure (CVP). Less commonly pulmonary capillary wedge pressure was also used, which in ideal situations is synonymous with left atrial pressure as a surrogate of left ventricular end-diastolic volume.

***As has been extensively reviewed and talked about – static measures of preload perform no better than chance in patients who are critically ill.

So…why don’t we use other bedside manoeuvres which rapidly change preload, these are more discriminative than static measures.


Respiratory Variation in thoracic pressure

The way the patient is breathing has massive effects on cardiac physiology.

Pulse pressure

  • In mechanically ventilated patients, the positive inspiratory pressure transfers blood from the lungs to the left heart, resulting in an increase in stroke volume.
    • At relatively low pressures, this is seen as a rapid increase in pulse pressure.
    • At high levels of positive intrathoracic pressure one can also decrease the venous return, and after transit of this reduced blood volume from the right to left heart there can be a decline in pulse pressure.
  • The larger the fluctuations in pulse pressure as a result of respiration, the greater the chance that the patient will increase cardiac output in response to fluid. This method cannot be used during arrhythmias such as atrial fibrillation or with very rapid heart rates.
    • This explains the good old ‘swing’ on the art. line we see.


Respiratory changes in intrathoracic pressure also occur in the dimensions of both the inferior and superior vena cavae. These changes also depend on the degree of intrathoracic pressure change and on the compliance of the vena cavae.

  • Positive intrathoracic pressure increases the size of the inferior vena cava (IVC), while negative intrathoracic breaths reduce its size. When the vena cava is distended, the compliance markedly reduces. The IVC diameter is easily and reproducibly measured 1–2 cm from the right atrial junction using transthoracic ultrasound see here
    • A large, nonfluctuant IVC therefore suggests that the patient is not on the steep (volume responsive) portion of the Starling curve.

You’ve given them initial IV resuscitation – to no avail!

Venous return, central venous pressure, and cardiac output are tightly co-regulated. These co-cariables stop being so synergistic in the patient who develops shock refractory to intravenous fluids.

In 2016, standardized protocols dictate that if your patient isn’t bleeding, give 20 ml/kg fluid.

  • Fifty percent of patients do not achieve adequate perfusion with modest volume expansion and thus require vasopressors to maintain circulatory tone.
    • In these patients, intravenous volume expansion during the first 6–12 hours following admission increases dramatically to 50–70 ml/kg.
  • Norepinephrine, the vasopressor of choice in most circumstances, will not only increase arterial blood pressure through increased systemic vascular resistance but has a significant effect upon capacitance vessels, both arterial and venous, resulting in effective fluid loading to the right heart.
    • All of this IV fluid, as well as vascular capacitance changes results in an extreme change to venous return, central venous pressure, and cardiac output.

Cardiopulmonary interactions play a crucial role in establishing the equilibrium of venous return, central venous pressure, and cardiac output. 70–90 % of patients with shock require mechanical ventilation – we sedate them and plug them into the ventilator powering in mean airway pressures of up to 24 cmH2O. With all of this and the explanations above, it’s not surprising we see significant but unpredictable effects. These high thoracic pressures cause a decline in venous return and/or a functional unloading of the left ventricle through pressurization of the heart and thoracic aorta.

What can the clinician expect to find on clinical examination and echocardiography following initial volume resuscitation?

  • Central venous pressure in health is tightly governed at 0–5 mmHg
  • Correlating IVC diameter should be about 13–21 mm when supine, with a collapse of more than 50 % upon quiet inspiration.
  • Patients who have had aggressive initial volume loading will have a CVP of around 9–15 mmHg

This change in venous pressure is highly influential upon the vena caval diameter as measured by transthoracic echocardiography.

  • A study where 30 ml/kg intravenous fluid resuscitation was given looked at IVC diameter and fluctuation.
    • In 110 subjects the IVC diameter following intravenous volume expansion was increased by 35 % from normal values to 17–29 mm.
    • In nearly half (45 %) of the patients there was no variation in diameter according to respiration.
    • In a further 20 % of patients there was >0 but <15 % variability (the median cut-off point of IVC collapse which defines fluid responsiveness).

So to recap, after initial resuscitation you will probably expect to see:

  1. A central venous pressure of 9–15 mmHg
  2. Maximum IVC diameter of 17–29 mmHg
  3. According to the increase in IVC diameter upon mandated inspiration, 2/3 patients will be deemed nonresponsive to fluid.

Clinical methods at the bedside to assess volume response

1. PLR and ultrasound

Passive leg raising (PLR) is one of the most versatile techniques to assess fluid needs in ICU patients. PLR can be performed at the bedside in both mechanically ventilated patients and in spontaneously breathing patients. Within 1 minute of bilateral PLR, there is an effective increase in preload through recruitment of blood pooled in the legs. Importantly though:

  • In the patient not on vasopressors
    • an increase in blood pressure suggests that the patient will respond to a fluid bolus (more preload)
  • Whereas in those on vasoactive medications there is no detectable change in blood pressure
    • the output of interest is change in cardiac output

This technique is therefore best suited to patients who are not yet on vasopressors, with normal intrathoracic pressures (mechanically ventilated patients often fail to augment their preload as definitely), and those without significant abdominal pathology. So, you could argue that this manoeuvre is pointless most of the time on ITU as most patients don’t tick the inclusion criteria boxes!

Meta-analysis of 23 studies with a combined total of 1013 patients from a wide range of clinical settings demonstrated that the global predictive value of PLR was strong.

  • Pooled sensitivity of 86 %
  • Specificity of 92 %
  • Summary AUROC of 0.95.
  • In another meta-analysis, 21 studies were analyzed and the pooled correlation between the PLR-induced versus fluid-induced increases in cardiac output was 0.76.
    • Pooled sensitivity was 85%
    • Specificity was 91%
    • Pooled AUROC was 0.95.

How 2 PLR!

  • The patient is placed in a semirecumbent position with the head of the bed 30–45° above the horizontal.
  • The bed is rapidly moved to simultaneously elevate the lower limbs to 30–45° above the horizontal while lowering the head of the bed to supine.
    • This maneuver transfers blood from the legs and the splanchnic reservoir to the intrathoracic compartment, rapidly increasing the preload, thereby testing the preload dependency of the heart.
    • 250–350 ml of blood is effectively transferred from the legs to the heart and this method is entirely reversible.
    • It is essential that this maneuver should be done from the semirecumbent position because this increases the blood shift and accentuates the change in cardiac output compared with a supine start.
  • Cardiac output changes can be detected 1–2 minutes after the PLR maneuver using echocardiography.
  • It is useful to note that there is a close correlation between the changes in cardiac output or stroke volume induced by the PLR and that achieved through equivalent intravenous volume expansion. The greater the change you see here, the greater chance the patient will be of being fluid responsive.
  • The PLR maneuver seems inaccurate in patients with very significant intra-abdominal hypertension, but is relatively unaffected by arrhythmia.


  1. PLR can be performed regardless of arrhythmia or mode of ventilation.
  2. The PLR is not simply “positive” versus “negative”; the degree of increase in stroke volume to PLR predicts the increase in these parameters to fluids.


  1. The interobserver and intraobserver reliability of measurements in cardiac output is highly operator dependent. A skilled operator is required to achieve high-quality measures of aortic blood flow.

Also see here on how to assess fluid responsiveness via PLR and Echo in parts 1 and 2 on the page.

2. Aortic Blood Flow

Aortic flow variations during mechanical ventilation may be a superior measure of what is observed clinically as stroke volume variation (SVV), a parameter correlated with fluid responsiveness. Feissel et al. assessed the variation of the maximal velocity during the respiratory cycle and found that variation greater than 12 % accurately predicted fluid responsiveness of ICU patients.

Get your A5C view 

The aortic blood flow is recorded from an apical five-chamber view using pulsed Doppler imaging. If you want to go further and calculate full CO, see here.

Aortic blood flow variation shares the same limitations as pulse pressure variation. These two parameters may be used only in patients without arrhythmia and seem invalid in patients with right ventricular dilation or dysfunction. The pathophysiology of these parameters is based on the effects of mechanical ventilation, which induces transpulmonary and intrathoracic pressure change. The magnitude of these effects depends mainly on the transmission of airway pressure variations to the heart. Open chest conditions therefore make all of these parameters invalid to assess fluid need, as does protective mechanical ventilation (in which a low tidal volume is used to decrease the plateau pressure and driving pressure) is now widely used for ARDS patients. In ARDS patients in particular,  the low tidal volume decreases airway pressure variations and may dramatically decrease the hemodynamic effects of mechanical ventilation.

  • De Baker and Scolletta demonstrated that low tidal volume (<8 ml/kg) invalidates the cutoff value of 12 % for pulsed pressure variation (PPV) (a surrogate of stroke volume variation).
  • In an attempt to solve this problem, Liu et al. suggested estimating pleural pressure variations as a surrogate of thoracic pressure variations in ARDS patients and then adjusting the PPV accordingly in order to improve prediction and prevent false negatives for fluid responsiveness.
    • This approach, however, requires measurement of esophageal pressure using a balloon catheter, increasing the complexity of care, and is therefore used clinically in a small number of centers.


  1. No additional maneuvers are required; standard mechanical ventilation provides the dynamic changes in preload.


  1. The interobserver and intraobserver reliability of measurements in cardiac output is highly operator dependent. A skilled operator is required to achieve high-quality measures of aortic blood flow.
  2. Not accurate during arrhythmia.
  3. Of limited utility with “open lung” ventilator strategies which reduce pleural pressure swings.

3. IVC Variation

Controlled ventilation

Positive pressure inspiration = IVC collapse

Under controlled mechanical ventilation, positive pressure is applied into the thorax. The superior vena cava (SVC) is therefore subjected to this positive pressure during mechanical insufflation. Vieillard-Baron et al. demonstrated that respiratory variation of the superior vena cava analyzed using transesophageal echocardiography accurately predicts fluid responsiveness of ICU patients with a cutoff value of 36 %.

Following this study there was a revolution in ultrasound technology which facilitated a less invasive approach, and in 2016 most clinicians prefer transthoracic echocardiography rather than the transesophageal approach to assess the IVC.

  • Multiple studies analyzed the IVC in ICU patients under controlled mechanical ventilation.
    • Together these studies demonstrated that the absolute size of the IVC (static value), was not able to accurately predict the effect of fluid infusion on cardiac output.
  • In contrast, the change in IVC diameter induced by intrathoracic pressure swings during mechanical ventilation is useful.
    • Using the ratio between maximal size minus minimum size to the average of these two values, Feissel et al. found that a variation higher than 12 % was associated with an increase of cardiac output after fluid infusion.
    • Barbier et al. found that 18 % was the cutoff value by using the ratio of the maximal size minus the minimum size to the minimum size.

All of these measurements were made on M-mode images of a longitudinal view of the IVC obtained from a subcostal window. Intra-abdominal hypertension, the tidal volume, and the patient’s inspiratory efforts in spontaneous breathing may be possible limitations of this approach.

Spontaneously breathing patients

Negative pressure inspiration (sniff) = IVC collapse

Following the initial resuscitation, most patients in the modern era are nursed while awake and are encouraged to breathe in collaboration with the ventilator. This means that in awake, spontaneously breathing patients the swings in pleural pressures during inspiration which are transmitted to the IVC can vary from deeply negative (in those ventilated on CPAP only, as in a spontaneous breathing trial) to neutral/positive in cases with high levels of pressure support or neuromuscular weakness.

It has recently been shown in healthy volunteers that the change in IVC diameter is highly correlated with respiratory effort. Clearly both the IVC diameter and fluctuation are highly influenced even at modest levels of pressure support. This intuitive but under-recognized fact has important implications when interpreting the results of an ultrasound examination in awake patients with the intent of guiding fluid therapy.

  • Reports of changes in IVC diameter in spontaneously breathing patients have found that these failed to accurately predict fluid responsiveness.
    • For instance, IVC respiratory variations >42 % in spontaneously breathing patients demonstrated a high specificity (97 %) and a positive predictive value (90 %) to predict an increase in CO after fluid infusion with a cutoff value >42 % [47] but a low sensitivity and negative predictive value.
    • A recently published physiology-based opinion suggests that IVC respiratory variations are in fact prone to both false negatives and positives due to five major categories: ventilator settings, patient’s inspiratory efforts, lung hyperinflation, cardiac conditions impeding venous return, and high intra-abdominal pressures.

Advantages of measuring vena caval diameter during ventilation

  1. No additional maneuvers are required; the standard mechanical ventilation provides the dynamic changes in preload.

Disadvantages of measuring vena caval diameter during ventilation

  1. Of limited utility in controlled modes of ventilation using “open lung” strategies which reduce pleural pressure swings.
  2. In awake patients, the IVC diameter and collapse are highly dependent upon the patient’s respiratory effort and levels of ventilatory support.
  3. While echocardiography is of great value in the diagnosis of right ventricular failure, more direct measures of left ventricular performance such as descending aortic Doppler flow are required in this population.


Future developments

Left ventricular outflow tract obstruction

In a recent study, Chauvet et al. found that 22 % patients in the early phase of septic shock presented with functional left ventricular outflow tract obstruction (LVOTO). In most of these patients, fluid infusion decreased this obstruction, increased cardiac output, and clinically improved the patient. Pending confirmation and prospective validation, LVOTO may be considered a new fluid-responsiveness parameter.

Have we sold this to you?

IVC views and collapsibility indices are easy to do by the bedside amongst your more standard mechanically ventilated patients. But, bare in mind after initial resuscitation, two-thirds of patients will not be fluid responsive.

In those breathing spontaneously on ventilators, we run into more dangerous territory on assessment of the IVC. It depends on respiratory effort and the pressure applied to assist ventilation, and without standardized ventilator settings it has not been proven a reliable indicator of fluid responsiveness.

So…if you are skilled enough to look at aortic blood flow and can do CO calculations reliably, combine this with IVC measurement to get a good idea. If you aren’t, just the IVC alone, in patients weaning with a bit of ASB still on the vent, you are perhaps in dicey territory to use this alone to govern whether you will drown them or help drive them! In which case….it’s back to static CVP, arterial BP, Urine output and lactate measurements with good old clinical acumen!


2 thoughts on “Fluid responsiveness…will your echo help?! #FOAMed #FOAMcc #POCUS

Add yours

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Powered by WordPress.com.

Up ↑