Can We Accurately Predict Fluid Responsiveness?

Abstract & Commentary

By Andrew M. Luks, MD, Pulmonary and Critical Care Medicine, University of Washington, Seattle Dr. Luks reports no financial relationship to this field of study. This article originally appeared in the April 2010 issue of Critical Care Alert. It was edited by David J. Pierson, MD, and peer reviewed by William Thompson, MD. Dr. Pierson is Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, and Dr. Thompson is Staff Pulmonologist, VA Medical Center; Associate Professor of Medicine, University of Washington; they both report no financial relationships relevant to this field of study.

Synopsis: This non-randomized, prospective trial demonstrated that passive leg raising can be used to predict fluid responsiveness in non-intubated, spontaneously breathing patients with severe sepsis or acute pancreatitis.

Source: Préau S, et al. Passive leg raising is predictive of fluid responsiveness in spontaneously breathing patients with severe sepsis or acute pancreatitis. Crit Care Med. 2010;38:819-825.

Passive leg raising (PLR), a rapidly reversible maneuver that simulates rapid volume expansion by putting several hundred milliliters of fluid back into the circulation, has been shown to predict fluid responsiveness in mechanically ventilated patients. Préau and colleagues sought to determine whether the same maneuver could be used in spontaneously breathing patients and whether there were any differences between three potential means of assessing the hemodynamic response to the PLR: stroke volume changes measured by echocardiography, pulse pressure variation on arterial pressure monitoring, and Doppler flow measurements in the femoral artery.

The authors enrolled consecutive, non-intubated patients with sepsis or acute pancreatitis at a single center in France. Patients were eligible for participation if the attending physician decided to perform a fluid challenge based on the presence of signs of inadequate tissue perfusion, including hypotension, decreased urine output, mottled skin, or tachycardia. Patients on non-invasive ventilation, or those with poor echocardiographic images or high-grade aortic insufficiency, were excluded. Enrolled patients were placed in the supine position. Systolic, diastolic, and mean arterial pressure (SAP, DAP, MAP), as well as pulse pressure (PP) were measured using an arterial catheter while blood flow velocity in the femoral artery (VF) was assessed with continuous Doppler and stroke volume (SV) was measured with transthoracic echocardiography. The measurements were then repeated within five minutes of raising the patients' legs to 30°-45° relative to the trunk. Following a five-minute period with legs back in the flat position, patients then received a 500 mL infusion of 6% hydroxyethylstarch over < 30 minutes and a final set of measurements was completed. Individuals whose SV rose > 15% following volume expansion (VE) were labeled as responders, while those with < 15% change in SV were labeled non-responders. Changes in each variable in response to PLR were then compared to changes observed following volume expansion. Observed changes were also compared between responders and non-responders.

Of 890 patients admitted to the ICU in the two-year study period, only 34 met inclusion criteria and had suitable echocardiographic windows to permit SV measurements, including 28 patients (82%) with severe sepsis and six (18%) with acute pancreatitis. In the entire group, PLR increased SV from 47 ± 14 to 50 ± 14 mL (p < 0.01), while VE increased SV from 47 ± 14 to 53 ± 15 mL. SV positively correlated with PP (r2 = 0.4) and VF (r2 = 0.62). Among the 14 patients (41%) determined to be VE-responders, SV, PP, and VF were significantly higher than in the non-responders, and also showed positive and statistically significant correlations with SV changes seen following VE. A SV greater than or equal to 10% predicted volume responsiveness with a sensitivity of 86% and specificity of 90%, while PP greater than or equal to 9% (sensitivity 79%, specificity 85%) and VF greater than or equal to 8% (sensitivity 86%, specificity 80%) were also found to be useful in this regard.

Commentary

Just about every critical care physician has been at the bedside of a hypotensive patient and wondered whether to give fluids or start a vasopressor. A reliable tool for answering this question and predicting volume responsiveness remains one of the "holy grails" of critical care medicine. The clinician can always stand there and give more fluid to see if the patient is volume-responsive but in non-intubated patients this simple strategy may provoke worsening respiratory failure and create a need for intubation that may not have been there otherwise. Central venous pressure measurements are often relied upon to guide decisions, but studies have consistently shown that such static measures of preload are poor predictors of volume responsiveness. In the face of such issues, PLR and other "dynamic" measures have been proposed as an alternative means for guiding these decisions.1

At first glance, the technique sounds exceedingly simple. Place the patient in the supine position, raise his/her legs to 30°-45°, during which time about 300-500 mL of fluid reenters the circulation, and observe the response. The devil is in the details, however, as the response to PLR and, therefore the determination of volume responsiveness, is not made by simply looking at whether the patient's blood pressure improves, but rather requires the use of techniques such as echocardiography or esophageal Doppler monitoring to assess the hemodynamic response to the maneuver. These requirements limit the utility of PLR a great deal as clinicians may either lack the skills to do the measurements themselves or may not have access to the tools in the short time they have to make a decision about treating the hypotensive patient.

The study by Préau et al moves PLR a bit closer to being a useful bedside technique because it demonstrates that PP changes on an arterial catheter correlate to a reasonable degree with echocardiographic measurements of the change in SV and, as a result, can be used as a surrogate measure of the hemodynamic response to PLR. However, while the study does move us closer, it does not get us all the way there. The authors only looked at non-intubated patients in their study, a significant limitation when one considers the high percentage of patients with severe sepsis and acute pancreatitis who end up on mechanical ventilation. Further study will be needed in these patient groups using a PLR protocol similar to that used by Préau and colleagues before we can rely on this technique to guide our assessments of volume responsiveness.

Reference

1. Pinsky MR. Hemodynamic evaluation and monitoring in the ICU. Chest. 2007;132:2020-2029.