IVC fluid responsiveness

The slides are building a historical argument: early small, tightly controlled studies made IVC variation look highly accurate; later larger and more pragmatic studies showed that its performance falls substantially in real ICU patients.

1. Michard et al., Critical Care 2000

“Using heart–lung interactions to assess fluid responsiveness during mechanical ventilation”

What the article contributed

This was primarily a physiology and conceptual framework paper rather than an IVC validation trial. It explained why positive-pressure ventilation creates cyclic changes in right- and left-sided stroke volume.

During mechanical inspiration:

1. Intrathoracic pressure rises.
2. Right ventricular preload decreases.
3. Right ventricular afterload may increase because alveolar pressure compresses pulmonary vessels.
4. RV stroke volume falls.
5. After pulmonary transit time, LV preload falls.
6. LV stroke volume reaches its minimum during expiration.

At the same time, positive pressure may transiently increase LV ejection by:

• reducing LV transmural afterload;
• squeezing pulmonary blood toward the left heart;
• increasing apparent LV preload early during inspiration.

The magnitude of these cyclic changes is greatest when the patient is preload dependent.

Evidence

The article synthesized studies showing that respiratory variation in:

• systolic pressure;
• pulse pressure;
• aortic blood-flow velocity;
• stroke volume;

could reveal where the patient lies on the Frank–Starling curve.

Teaching points

• Mechanical ventilation acts as a repetitive, reversible preload challenge.
• The right heart responds first; the left-heart effect appears several beats later because of pulmonary transit time.
• Dynamic variation predicts preload responsiveness, not absolute blood volume.
• Large variation does not automatically mean the patient needs fluid.
• This physiology only works predictably when ventilator breaths are regular and the patient is passive.

---

Early IVC studies: highly selected patients

2. Barbier et al., Intensive Care Medicine 2004

“Respiratory changes in inferior vena cava diameter are helpful in predicting fluid responsiveness in ventilated septic patients”

Population

• 23 patients.
• Septic circulatory failure.
• Intubated and fully passive.
• Tidal volume approximately 8.5 mL/kg.
• Sinus rhythm.
• Received approximately 7 mL/kg of gelatin.

Index

IVC distensibility was calculated approximately as:

$$dIVC = \frac{D_{\max}-D_{\min}}{D_{\min}}\times100$$

A responder was generally defined by a cardiac index increase of at least 15% after volume expansion.

Main evidence

• Best cutoff: dIVC ≥18%.
• Sensitivity: approximately 90%.
• Specificity: approximately 90%.
• Baseline dIVC strongly correlated with the increase in cardiac index after fluid, reported around $r=0.9$.
• Baseline CVP did not accurately predict responsiveness.

Why it looked so good

The study created nearly ideal conditions for a heart–lung interaction test:

• controlled ventilation;
• relatively large tidal volumes;
• no spontaneous respiratory effort;
• septic patients without major rhythm irregularity;
• a small, selected cohort.

Teaching points

• This is the classic paper supporting the 18% distensibility threshold.
• The cutoff applies to the specific formula using $D_{\min}$ as the denominator.
• Do not transfer the 18% threshold to collapsibility in spontaneously breathing patients.
• A small derivation cohort often produces better diagnostic performance than later validation cohorts.
• This paper proves that IVC variation can work under ideal conditions, not that it works universally.

---

3. Feissel et al., Intensive Care Medicine 2004

“The respiratory variation in inferior vena cava diameter as a guide to fluid therapy”

Population

• 39 patients.
• Septic shock.
• Intubated and passive.
• Tidal volume approximately 8–10 mL/kg.
• Received 8 mL/kg of 6% hydroxyethyl starch.

Index

This study used a different normalization:

$$\Delta IVC = \frac{D_{\max}-D_{\min}}{D_{\text{mean}}}\times100$$

Main evidence

Respiratory variation in IVC diameter was greater in fluid responders than nonresponders and predicted the increase in cardiac output after volume expansion. The authors concluded that ΔIVC was a simple, noninvasive method for identifying fluid responsiveness in mechanically ventilated septic-shock patients.

A commonly reported threshold from this study is approximately 12%, although thresholds vary depending on the equation and measurement technique.

Teaching points

• Barbier and Feissel are not directly interchangeable because they used different formulas.
• A reported “IVC variation of 12%” and “IVC variation of 18%” may describe essentially similar physiology normalized differently.
• Always document:
  • ventilatory status;
  • maximum and minimum diameters;
  • exact equation;
  • measurement site;
  • whether the patient was fully passive.
• The strongest early evidence was for controlled mechanical ventilation, not spontaneous breathing.

---

Studies demonstrating limitations

4. Mahjoub et al., British Journal of Anaesthesia 2014

“Evaluation of pulse pressure variation validity criteria in critically ill patients: a prospective observational multicentre point-prevalence study”

This paper evaluated PPV rather than IVC directly, but it is highly relevant because PPV, SVV, aortic-flow variation, and IVC variation all rely on predictable heart–lung interactions.

Population and design

• Multicenter point-prevalence study.
• 26 ICUs in 22 French hospitals.
• Investigators determined how many ICU patients met the prerequisites for respiratory dynamic indices.

Validity conditions

Typical requirements included:

• controlled mechanical ventilation;
• no spontaneous breathing;
• regular rhythm;
• tidal volume ≥8 mL/kg predicted body weight;
• adequate lung compliance;
• sufficiently low respiratory rate relative to heart rate.

Main evidence

Only approximately 2% of all ICU patients met all validity criteria. Even among patients with an arterial catheter, only about 3% met all criteria.

Teaching points

• A test can be physiologically valid but clinically applicable to very few patients.
• Modern lung-protective ventilation at 6 mL/kg reduces the transmitted preload challenge and may yield falsely low respiratory variation.
• Spontaneous effort, atrial fibrillation, low compliance, RV dysfunction, and elevated intra-abdominal pressure undermine interpretation.
• The slide’s “only 2%” statement comes from PPV applicability, but the same restrictions undermine IVC distensibility.
• Before interpreting respiratory variation, first ask: Is this patient eligible for the test?

---

5. Charbonneau et al., Critical Care 2014

“Predicting preload responsiveness using simultaneous recordings of inferior and superior vena cava diameters and aortic velocity variations in ventilated patients”

Question

Are IVC variation, SVC variation, and aortic-flow variation equally accurate when measured simultaneously?

Design

• Mechanically ventilated ICU patients.
• Fluid responsiveness assessed using a reference change in stroke volume or cardiac output.
• Simultaneous echocardiographic measurements allowed direct comparison.

Main evidence

• ΔSVC performed better than ΔIVC for predicting fluid responsiveness.
• On the slide, a threshold of ΔIVC ≥21% had:
  • sensitivity approximately 38%;
  • specificity approximately 61%.
• The study demonstrated substantial overlap between responders and nonresponders.
• SVC variation was more discriminating, although it requires transesophageal echocardiography.

Why IVC performance fell

The IVC is influenced by more than circulating volume:

• right atrial pressure;
• RV function;
• abdominal pressure;
• diaphragmatic movement;
• measurement site;
• venous compliance;
• ventilator pressure transmission.

The SVC is intrathoracic and may more directly reflect changes in venous return produced by positive-pressure ventilation.

Teaching points

• A plethoric or variable IVC is not a direct measurement of stroke volume.
• IVC diameter reflects the interaction of thoracic pressure, abdominal pressure, venous compliance, and right-heart function.
• SVC collapsibility is often physiologically stronger in ventilated patients but is less practical because it requires TEE.
• The low sensitivity means a negative IVC test cannot confidently exclude fluid responsiveness.
• Later studies did not reproduce the spectacular 90%/90% performance of the early trials.

---

6. Vignon et al., American Journal of Respiratory and Critical Care Medicine 2017

“Comparison of echocardiographic indices used to predict fluid responsiveness in ventilated patients”

This is the most important large pragmatic comparison shown on the slides.

Population

• 540 mechanically ventilated patients.
• Multicenter prospective study.
• Acute circulatory failure of any cause.
• 42% were fluid responsive.

Reference standard

A passive leg raise was performed. An increase in aortic VTI of at least 10% defined preload responsiveness.

Indices compared

• Pulse-pressure variation.
• Aortic maximum velocity variation.
• SVC diameter variation.
• IVC diameter variation.

Feasibility

• ΔSVC: measurable in 99.6%, using TEE.
• ΔVmaxAo: 78.0%.
• ΔIVC: 78.1%.
• ΔPP: 78.5%.

Main evidence

For ΔIVC ≥8%:

• sensitivity: 55%;
• specificity: 70%.

For ΔSVC ≥21%:

• sensitivity: 61%;
• specificity: 84%.

For ΔVmaxAo ≥10%:

• sensitivity: 79%;
• specificity: 64%.

SVC variation had significantly greater overall diagnostic accuracy than IVC variation and PPV. Aortic-flow variation had the best sensitivity, whereas SVC variation had the best specificity.

Interpretation

This large study is much closer to real ICU practice than the 23- and 39-patient studies.

At ΔIVC ≥8%:

• nearly half of responders would be missed;
• approximately 30% of nonresponders would test positive.

Therefore, IVC was not accurate enough to function as a stand-alone yes/no test.

Teaching points

• In heterogeneous ICU patients, IVC variation has only moderate diagnostic performance.
• A cutoff is not a magic biological boundary.
• Values close to the threshold belong in a diagnostic gray zone.
• Measuring actual flow—especially LVOT VTI response to PLR—is generally more informative than measuring venous diameter alone.
• The test should modify probability, not dictate treatment.

---

Spontaneously breathing patients

7. Muller et al., Critical Care 2012

“Respiratory variations of inferior vena cava diameter to predict fluid responsiveness in spontaneously breathing patients with acute circulatory failure”

Population

Spontaneously breathing patients with acute circulatory failure.

Main evidence

• A high collapsibility index, generally >40%, was associated with fluid responsiveness.
• A low value did not exclude fluid responsiveness.
• Thus, marked collapse was more useful for ruling in than modest collapse was for ruling out.

Teaching points

• Spontaneous breathing is not simply the opposite of controlled ventilation.
• The magnitude of IVC collapse depends heavily on how forcefully the patient inspires.
• A vigorous inspiratory effort can collapse the IVC even without true preload responsiveness.
• A weak, exhausted patient may show little collapse despite being responsive.
• In spontaneous breathing, extreme values are more useful than intermediate values.

---

8. Airapetian et al., Annals of Intensive Care 2015

“Does inferior vena cava respiratory variability predict fluid responsiveness in spontaneously breathing patients?”

Main evidence

• Resting IVC diameter and ordinary respiratory variability did not reliably predict fluid responsiveness.
• A very high collapsibility index, approximately >42%, could identify some responders.
• Lower values could not exclude responsiveness.

Teaching points

• The main problem is poor sensitivity.
• “Not very collapsible” does not mean “do not give fluid.”
• Deep inspiratory effort produces more useful discrimination than uncontrolled tidal breathing, but also makes the test effort dependent.
• PLR coupled with a direct stroke-volume measurement is preferable.

---

9. Préau et al., Critical Care Medicine 2017

“Diagnostic accuracy of the inferior vena cava collapsibility to predict fluid responsiveness in spontaneously breathing patients with sepsis and acute circulatory failure”

Technique

Instead of observing uncontrolled breathing, patients performed a standardized deep inspiration, sometimes with monitoring of mouth pressure to make inspiratory effort more reproducible.

Main evidence

A marked IVC collapse during standardized inspiration predicted fluid responsiveness more accurately than ordinary quiet breathing. The authors presented it as a simple bedside test, but in a selected population.

Teaching points

• Standardization improves a physiologic test.
• The result does not validate casual visual estimation during random breathing.
• The patient must be able to follow commands and generate the prescribed inspiratory effort.
• This approach is not applicable to encephalopathy, severe distress, or exhaustion.
• Positive extreme values may be helpful; intermediate and negative values remain less decisive.

---

Synthesis of the evidence

What supports IVC use

1. Strong physiological rationale
   Ventilation cyclically changes venous return, which can reveal preload dependence.
2. Good performance in ideal patients
   Barbier and Feissel showed high accuracy in deeply controlled, mechanically ventilated septic patients with large tidal volumes.
3. Noninvasive and rapidly repeatable
   It can be incorporated into a broader cardiac and lung POCUS examination.
4. Extreme findings can be useful
   Marked distensibility in a passive ventilated patient or marked collapse during a standardized inspiration increases the probability of responsiveness.
5. Better than static CVP in the original studies
   Dynamic variation contains more information than a single filling-pressure value.

What argues against relying on IVC

1. Small derivation studies overestimated accuracy.
2. Large validation studies found sensitivity only around 38–55% and specificity around 61–70%.
3. Very few ICU patients satisfy all heart–lung interaction assumptions.
4. The formulas and cutoff values are inconsistent.
5. IVC is affected by RV function, intra-abdominal pressure, ventilator settings, inspiratory effort, and measurement technique.
6. Fluid responsiveness is not the same as fluid tolerance or a need for fluid.

---

Practical fellow-level interpretation

Controlled mechanical ventilation

Consider IVC variation only when:

• no spontaneous effort;
• regular rhythm;
• adequate tidal volume or a validated tidal-volume challenge;
• no major RV failure;
• no significant intra-abdominal hypertension;
• no major tricuspid regurgitation;
• technically adequate view.

A clearly high value may support preload responsiveness, but confirm with:

• PLR and LVOT VTI;
• end-expiratory occlusion;
• stroke-volume change;
• a mini-fluid challenge.

Spontaneous breathing

Do not use ordinary IVC collapsibility as a binary test.

A very high collapse during a standardized inspiration may raise the probability of responsiveness, but a low or intermediate value is nondiagnostic.

Before giving fluid, ask three separate questions

1. Is the patient preload responsive?
   Would stroke volume rise?
2. Does the patient need increased flow?
   Is there hypoperfusion that greater cardiac output could correct?
3. Can the patient tolerate fluid?
   Look for pulmonary edema, elevated left-sided filling pressure, RV failure, venous congestion, and renal congestion.

One-sentence teaching summary

IVC variation is a physiologically valid but condition-dependent marker of preload responsiveness; it performs best in passive, regularly ventilated patients, performs poorly as a stand-alone test in general ICU populations, and should be integrated with direct flow measurements, cardiac function, lung ultrasound, and the clinical need for fluid.