Inferior vena cava distensibility as a predictor of fluid responsiveness in patients with subarachnoid hemorrhage

Moretti. Neurocrit Care. 2010 Aug;13(1):3-9. doi: 10.1007/s12028-010-9356-z

Clinical Question

• In mechanically ventilated patients with subarachnoid hemorrhage (SAH) and a low cardiac index, does the distensibility of IVC on ultrasound, predict fluid responsiveness?

Design

• Prospective observational study
• Consecutive patients
• Statistics:
  • No sample size calculation
  • Comparisons of:
    • Quantitative variables between groups: independent, two-tailed t test
    • Categorical data: Fisher’s test
    • Repeated measurements: paired t test
    • Inter-rater reliability: Pearson’s correlation coefficient
• Blinding: ultrasonographers blinded to haemodynamic parameters and not involved in patient care
• Intra and inter-observer variability of test of interest determined with repeated measurements in 10 patients

Setting

• Single 10 bedded intensive care unit, Italy
• August 2008 and July 2009

Population

• Inclusion criteria:
  • Adult patients with SAH (Fisher Grade 3/4)
  • Requiring advanced haemodynamic monitoring, sedation, and mechanical ventilation
  • Haemodynamic instability: Cardiac index (CI) <2.5 l/min/m2 or a cerebral perfusion pressure (CPP) <60 mmHg
• Exclusion criteria:
  • Pre-existing heart failure, cardiac arrhythmias, ARDS, extravascular lung water (ELWI) >14 ml/kg, inability to perform femoral artery cannulation or ultrasonography
• 31 patients recruited. 2 patients excluded, due to suboptimal visualisation of the IVC (n=1), ELWI >14 ml/kg (n=1)

All patients

• Sedated with propofol and remifentanil to achieve bispectral index <60
• Paralysed with cisatracurium. Confirmed with peripheral nerve stimulator
• Given fluid challenge of 7 ml/kg infusion of 6% hydroxyethyl starch over 30 minutes
• Test of interest and gold standard investigation performed at baseline and immediately post fluid challenge

Test of Interest

• IVC sonography
  • Performed by 2 intensivists with at least 3 years (60 scans+) ultrasound experience
  • Use of 3.5-MHz ultrasound probe
  • Performed supine
  • During ultrasonography patients ventilated with standardised settings: volume-controlled ventilation, inspiratory:expiratory ratio of 1:2, respiratory rate of 12/min, tidal volume of 8 ml/kg and a positive end-expiratory pressure (PEEP) of 0 cm H2O
  • IVC diameter measured in M-mode, 2cm upstream of the origin of the suprahepatic vein. M-mode tracing perpendicular to the IVC
  • Maximum IVCD on inspiration (IVCDi) and minimum IVCD on expiration (IVCDe) measured
  • IVC distensibility index (dIVC) = (IVCDi-IVCDe)/IVCDe

Gold Standard Investigation

• Thermodilution derived cardiac index
  • Measured with femoral PICCO line and internal jugular or subclavian central line
  • PEEP set at 5 cm H20 for measurements of stroke volume variation (SVV)
  • Patients classified as fluid responsive if cardiac index increased ≥15% with fluid challenge

Outcome

• Primary outcome: the reliability of dIVC as a predictor of fluid responsiveness
  • Area under the ROC curve
    • dIVC: 0.902 (95% C.I. 0.73-0.98)

• Secondary outcomes:
  • 17 patients classified as fluid responders (FR) and 12 patients classified as non-responders (NR)
  • At baseline
    • dIVC (%) significantly higher in FR than NR (16 vs. 10, p < 0.0001)
    • SVV (%) significantly higher in FR than NR: (11.8 vs. 4.4, p=0.03)
  • Post fluid bolus
    • In fluid responders significant decrease in
      • dIVC (%) 16 ± 3.3 vs. 11.9 ± 2.3, p < 0.001
      • SVV (%) 11.8 ± 4.4 vs. 7 ± 2.4, p < 0.001
    • In non responders no significant change in
      • dIVC (%) 10 ± 3.4 vs. 8.7 ± 2.2, p = 0.42
      • SVV (%) 7.5 ± 2.9 vs. 8 ± 4.1, p=0.68
    • ROC curves to determine the best predictor of fluid responsiveness: Area under the curve (AUC)
      • SVV: 0.779 (95% C.I. 0.587-0.911)
      • dIVC: 0.902 (95% C.I. 0.733-0.979)
      • CVP: 0.667 (95% C.I. 0.468-0.829)
  • Intra- and inter-observer variabilities in the measurement of IVCDe were 4 ± 4% and 6 ± 4%, respectively.

Authors’ Conclusions

• IVC distensibility was a reliable measure for predicting fluid responsiveness in critically ill patients with SAH

Strengths

• Blinding of ultrasonographers
• Appropriate gold standard investigation
• Clear description of how test of interest was performed
• Tested for inter and intra-observer variablitiy

Weaknesses

• No sample size calculation, with small numbers of patients included
• External validity may be limited as:
  • Restrictive inclusion/exclusion criteria
  • PEEP changed to zero for measurement of dIVC. This will not be possible for all patients
  • Single centre

The Bottom Line

• In patients with a grade 3/4 SAH and a low cardiac index, who are mechanically ventilated with tidal volumes of 8ml/kg, and are sedated and paralysed, dIVC was a useful predictor of fluid responsiveness. Further trials will need to determine if these results can apply to other populations.

External Links

• [article] Inferior vena cava distensibility as a predictor of fluid responsiveness in patients with subarachnoid hemorrhage
• [further listening] EMCrit – assessing fluid responsiveness
• [further listening] SMACC Debate – Rob MacSweeney vs Paul Marik – Predicting Fluid Responsiveness is a Waste of Time
• [further reading] LITFL fluid responsiveness
• [further listening] EMCrit – IVC Ultrasound for Fluid Tolerance in Spontaneously Breathing Patients
• [further listening] SMACC: The Dark Art of IVC Ultrasound
• [further watching] Ultrasound Podcast: IVC

Metadata

Summary author: David Slessor
Summary date: 25th October 2016
Peer-review editor: Duncan Chambler