Predicting fluid response in the ICU

Predicting fluid response in the critically ill Dr. Andrew Ferguson Consultant in Anaesthesia & Intensive Care Medicine Craigavon Area Hospital

Approach to shock Fluid challenge central to therapy +/- CVP (and/or PA) monitoring Repeat if CVP/PAWP still low Stop if CVP/PAWP goes high Surrogate markers for CO Lactate SvO 2

So what’s the problem? ? validity of CVP as end-point ? validity of PAWP as end-point Preload-SV relationship unknown Only 50% of patients fluid-responsive Excess fluid problems Interstitial fluid excess Worsened gas exchange Limitation of oxygen diffusion

Variability of fluid response rates Michard (Chest 2002; 121: 2000-2008)

Preload does not guarantee response

To be a fluid responder, both ventricles must be on ascending portion of Frank-Starling curve Response depends on contractility and diastolic function as well as load

Common measures used to indicate likelihood of response CVP PAWP RVEDV (thermodilution) LVEDA (echo)

R 2 = 0.2 In spontaneous resp. a fall > 1 mmHg in RAP has positive predictive value of 77-84% and negative predictive value of 81-93% for response

They don’t work --- what next??

BP change relates to SV change

Cardio-pulmonary interactions Changes in SV, PP, SBP with positive pressure ventilation

Increased pleural pressure RV preload falls LV afterload falls Increased transpulmonary pressure RV afterload increases LV preload increased by alveolar vessel squeeze Decreased RVSV Increased LVSV

Inspiratory decrease in RVSV Expiratory decrease in LVSV Expiratory decrease in LV preload Pulmonary transit time

Stroke volume variation and LVEDP

Potential tools Stroke volume variation Systolic pressure variation Pulse pressure variation Peak aortic blood flow velocity variation

Systolic Pressure Variation  down is the important one for fluid response

 SP as indicator of fluid response

 PP as indicator of fluid response

Measures of response to volume

Predictive values Study No. of patients Measure Threshold Positive pred. val. Negative pred. val. Magder 33  RAP (SPONT) 1 mmHg 84 93 Tavernier 35  Down 5 mmHg 95 93 Magder & Lagonidis 29  RAP (SPONT) 1 mmHg 77 81 Michard 40  PP 13% 94 96 Feissel 19  VPeak 12% 91 100

Problems with  PP and  SV Equipment not universal Need sinus rhythm False positive in severe abdominal distension

Normal values  PP 13% SPV  down 5%  Vpeak (aortic blood flow velocity) 12%  SV 10%

Conclusions Conventional measures often not valid New and accurate measures available Consider passive leg raising! Know cardio-pulmonary interactions

Predicting fluid response in the ICU

More Related Content

What's hot (20)

Similar to Predicting fluid response in the ICU (20)

More from Andrew Ferguson (20)

Recently uploaded (20)

Predicting fluid response in the ICU