Fluid-Responsiveness:

An ICU Phoenix

Darryl StewartRegional ICM Study Day

07/11/2013Altnagelvin Area Hospital

Resuscitation

• Fluid administration remains the cornerstone of immediate resuscitation

• Endless debate regarding timing, volume, duration, targets & type

Goal-directed Therapy

Fluid Balance & Mortality

• Rosenberg AL et al. Review of a large clinical series: association of cumulative fluid balance on outcome in acute lung injury: a retrospective review of the ARDSnet tidal volume study cohort. JICM 2009; 24:35-46

• Boyd JH et al. Fluid resuscitation in septic shock: a positive fluid balance and elevated central venous pressure increase mortality. CCM 2011; 39 (2): 259-61

• Bellomo R et al. An observational study fluid balance and patient outcomes in the Randomized Evaluation of Normal vs Augmented Level of Replacement Therapy trial. CCM 2012; 40 (6): 1753-60

Assessment

Often the 1st step…• Fluid administration is frequently initial response to

indicators of tissue hypoperfusion

• However, it is likely that only 50% of haemodynamically unstable ICU patients are volume responsive

• Marik PE et al. Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients. A systematic review of the literature. CCM 2009; 37: 2642-2647

Preload & SV• Fluid should only be administered to augment

preload in the belief that this will increase stroke volume & subsequently cardiac output

Haemodynamic changes with mechanical

ventilation

Vascular Waveform Analysis

Morgan BC et al. Haemodynamic effects of intermittent positive pressure ventilation. Anesthesiology 1966; 27: 584-90

Systolic Pressure Variation

•Accentuated in:• Hypovolaemia

• Tamponade

• Constrictive Pericarditis

• LV dysfunction

• Massive PE

• Bronchospasm

• Dynamic Hyperinflation

• Pneumothorax

• Increased intra-thoracic pressure

• Increased intra-abdominal pressure

Volaemic Assessment Parameters

StaticCVP

PAOP

RVEDV

LVEDA

GEDV & ITBV

DynamicSVV

PPV

SPV

IVC collapse

PLR

CVP• Kastrup et al, Acta Anaes Scand 2007

• 90% German anaesthetists & intensivists used CVP to guide fluid management (cardiac surgery & CSICU)

• McIntyre LA et al, Crit Care 2007

• 90% Canadian intensivists use CVP to guide fluid therapy in septic shock

Assumption

CVP → RAP → RV filling → LV preload → CO

• Systematic review; 213 articles screened, 24 met inclusion criteria = 803ptsOverall 56 +/- 16% (mean +/- SD) of 803pts responded to fluid challenge with pooled area under ROC curve 0.56. Pooled correlation between ΔCVP & change in stroke index/cardiac index (7 studies) was 0.11 (95% CI, 0.01 to 0.21). Baseline CVP (11 studies) was 8.7 +/- 2.3mmHg in responders, compared to 9.7 +/- 2.2mmHg in non responders (p=0.3)

Therefore the likelihood that CVP can predict responsiveness (at any given CVP) is no better than a coin toss!Pts are equally likely to be fluid responsive with low or high CVP!

CVP Fraud• The assumptions are overly tenuous

Changes in systemic & pulmonary venous capacitance

Changes in intrathoracic pressure

Changes in RV compliance & afterload

Changes in RV systolic function

Changes in LV compliance & filling

Assumption of NSR

• Systematic review, 12 studies included, looking at predictive factors of fluid responsiveness in ICU pts

• Assessed RAP, PAOP, RVEDV, LVEDA, + ΔRAP, Δdown, PPV & Δaortic blood velocity

• RAP did not predict fluid responsiveness

• PAOP investigated in 10 studies; not significantly lower in responder group in 7 studies, was significantly higher in responder group in 1 study & was significantly lower in responder group in 2 studies.

Chest 2002; 121 (6): 2008-8

• RVEDV studied in 6 studies; 4 showed RVEDV was not significantly lower in responder group; remaining 2 demonstrated RVEDV was significantly lower in responder group.

• LVEDA studied in 2 studies only. No significant association between LVEDA & fluid responsiveness.

Problems with preload indicators

• CVP & PAOP do not reflect ventricular end-diastolic volumes

• Ventricular diastolic compliance & filling is non-linear

• Don’t account for ventricular transmural filling pressures (afterload & compliance)

• It is TRANSMURAL pressures not intracavity pressures that are related to end-diastolic volumes (via compliance)

• RVEDV is strongly influenced by TR (very common in ICU)

• IMPORTANT: A patient can be fluid non-responder due to high venous capacitance, poor ventricular compliance &/or poor ventricular function

• Little wonder then that intracavity pressures & static chamber dimensions to not predict fluid responsiveness

Dynamic Parameters

• Numerous studies over last decade shown that dynamic parameters can better predict fluid responsiveness than static

• PPV - from analysis of arterial pressure waveform

• SVV - from pulse contour analysis

• Pulse oximeter plethysmography

• 40 mechanically ventilated pts in septic shock

• Demonstrated higher variations in systolic pressure (15% vs 6%) & pulse pressure (24% vs 7%) during respiration in pts who were vol responders (increase 15% CI)

• Derived that cut-off value of 13% for PPV had sensitivity of 94% & specificity 96%.

• Also, variability in pulse pressure was superior to systolic pressure variation in discriminating fluid responders.

• Meta-analysis of 29 clinical studies, 685 pts to evaluate utility of PPV & SVV in predicting fluid responsiveness

• Correlation coefficient for baseline PPV & SVV, and changes in CI or SV in response to volume expansion were 0.78 & 0.72 respectively.

• Area under ROC curve for PPV was 0.94, for SVV 0.86

• Included studies had remarkably consistent threshold PPV/SVV of 12-13% for defining fluid responsiveness

• PPV found to be more reliable predictor than SVV (directly measured, no calculated assumptions)

• NOTE: Appears to be “grey area” of PPV values (9-13%) were fluid responsiveness cannot be reliably predicted

• Cannesson et al. Anesthesiol 2011; 115: 231-241

• This “grey zone” may affect up to 25% pts under GA

Crit Care Med. 2009 Sep;37(9):2642-7. doi: 10.1097/CCM.0b013e3181a590da.

Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature.Marik PE, Cavallazzi R, Vasu T, Hirani A.

Pulse Oximeter Plethysmography

• Less invasive alternative to SVV & PPV

• Uses form of pulse pressure analysis to consider changes in peak & amplitude of pleth waveform (assesses vol changes)

• Numerous studies demonstrating significant correlation between changes in pulse oximeter waveform & PPV & hence may predict fluid responders

• “Pleth Variability Index” (Masimo, Irvine)

• Automatic measure of dynamic change in perfusion index (pulsatile:nonpulsatile blood flow) during ventilation.

• Pleth variability index >14% is predictive that pt will be fluid responder with sensitivity 81%

Dynamic Limitations• Arrhythmias & spontaneous respiration may influence change in

PPV / SVV response to volume loading

• PPV / SVV varies with tidal volume, PEEP, chest wall compliance etc

• De Backer et al, ICM 2005, demonstrated tidal volume must be > 8 ml/kg to ensure accuracy

• Lansdorp et al, BJA 2012, demonstrated predictive value of SPV, SVV & PPV were optimal when Vt >7ml/kg & patient in SR

• Requires optimum arterial trace

• PVI influenced by acute vasomotor changes

Oesophageal Doppler

Oesophageal Doppler

• Respiratory changes in aortic peak velocity may used to estimate fluid responsiveness (Monnet X et al, ICM 2005)

• Feissel et al, Chest 2001, demonstrated that a change in aortic peak velocity with respiration >12% allowed discrimination between fluid responders & non-responders with sensitivity of 100% & specificity 89%.

IVC Collapse • Intrathoracic pressure changes affect VR &

therefore diameter IVC

• Absolute diameter or Δdiameter have both been used to assess col status

• Sefidbakht et al, Emer Radiol 2007, developed vena cava collapsibility index (= (end-expiratory diameter - end-inspiratory diameter) / end-expiratory diameter).

• Those who responded had greater collapsibility at baseline (25% vs 6%)

• Studied 66 pts in medical ICU with “ALI”

• Used TOE Doppler to measure SVC collapsibility at baseline and following 10ml/kg fluid (HES).

• Threshold SVC collapsibility of 36% allowed discrimination between responders and non-responders (defined by increase CI of at least 11%) with a sensitivity of 90% & specificity of 100%

Intensive Care MedicineSeptember 2004, Volume 30, Issue 9, pp 1734-1739

Superior vena caval collapsibility as a gauge of volume status in ventilated septic patients

Antoine Vieillard-Baron, Karim Chergui, Anne Rabiller, Olivier Peyrouset, Bernard Page, Alain Beauchet, François Jardin

Passive Leg Raise• All previous techniques required sedated, ventilated pts

• PLR can be used reliably in awake pts with cardiac dysrhythmias

• Numerous studies demonstrating ability of PLR to predict fluid responsiveness

• Cavallaro et al, ICM 2010, meta-analysis determined AUC for PLR in determining fluid responsiveness was 0.95

• Should use CO monitor - FloTrac-Vigleo system can be used to measure SVV with PLR where a 10% or greater increase in CO predicts fluid responsiveness (Biais M et al, CC 2009)

Summary