Fluid Responsiveness in HFNC Patients

STUDY BACKGROUND The plethysmographic variation index (PVi) is a measure of the respiratory-induced variations in the plethysmographic waveform. Interestingly, in mechanically ventilated patients and under certain conditions, PVi may reflect fluid responsiveness (FR) (Loupec T Crit Care Med 2011; Sandroni C Intensive Care Med 2012).

High flow nasal cannula (HFNC) has been described as a useful supportive therapy in spontaneously breathing patients with respiratory failure (Roca Crit Care 2016), by increasing lung volume and generating a certain level of airway positive pressure. It has been shown that patients treated with HFNC may present the same hemodynamic changes, measured by transthoracic echocardiography, as those patients who are mechanically ventilated (MV) (Roca J Crit Care 2013). Therefore, PVi may also be useful in HFNC patients.

HYPOTHESIS AND OBJECTIVE As the hemodynamic changes associated with HFNC use are similar to those observed in MV patients, the hypothesis of the present study is that PVi may predict FR in HFNC patients.

The objective is to investigate whether the PVi can predict FR in patients treated with HFNC.

METHODS

1. Subjects: Patients treated with HFNC presenting with circulatory insufficiency, in whom volume expansion was planned by the attending physician. Exclusion criteria included spontaneous respiratory activity, cardiac arrhythmia, known intracardiac shunt, severe hypoxemia (PaO2/FIO2 <60 mmHg), left ventricular ejection fraction of <50%, hemodynamic instability during the procedure (defined as the variation in heart rate or blood pressure of >10% over the 15-min period before starting the protocol) and those with variations in arterial pulse pressure <13%.
2. Setting: Medico-surgical intensive care unit of a university hospital
3. Design: Pilot prospective, observational study
4. Procedures:
5. At baseline, all patients will be in semirecumbent position with the head at an angle of 45 degrees. ii. Measurements of CO and PVI will be performed in duplicate, before and after a fluid challenge (8ml/kg of colloids in 10 minutes). iii. Cardiac output (CO) will be estimated by averaging three consecutive measures obtained with transthoracic echocardiography (Porter TR. Journal of the American Society of Echocardiography 2015). The PVi will be measured using Masimo SET® pulse oximetry technology. The PVi is an automatic measure of the dynamic change in the perfusion index (Pi) that occurs during the respiratory cycle. The perfusion index (Pi) is the ratio of nonpulsatile to pulsatile blood flow through the peripheral capillary bed. iv. The decision to administer fluid was based on the presence of at least one clinical sign of inadequate tissue perfusion. Clinical signs of inadequate tissue perfusion were suspected at the bedside by observing hypotension (systolic blood pressure <90 mm Hg or the need for norepinephrine), oliguria (urine output <0.5 mL/kg/hr), and cool, mottled extremities.
6. Total flow delivered by HFNC and dosages of vasopressive drugs will be kept unchanged during the entire procedure. vi. The HFNC device (Optiflow™ system, MR850 heated humidified RT202 delivery tubing, and RT050/051 nasal cannula; Fisher \& Paykel Healthcare Ltd, Auckland, New Zealand) consists of a low resistance nasal cannula that can deliver up to 60L/min of totally conditioned (37ºC and 100% of relative humidity) gas admixture. It is initiated with a minimum flow of 40L/min and a fraction of inspired oxygen (FIO2) of 1. Then, FIO2 is set to maintain a pulse oximetry (SpO2) above 92% and flow rate is set according to the physician's judgment.
7. Endpoints: Diagnostic accuracy of the PVi (AUROC) for correctly classifying HFNC patients who would respond to fluid challenge (ΔCO >15%).
8. Data analysis: Changes in hemodynamic measurements induced by fluid challenge will be assessed by using the nonparametric Wilcoxon's rank sum test for paired data. Patients will be divided into two groups according to changes ΔCO after fluid challenge. Those presenting a ΔCO >15% will be considered as responders and the others as nonresponders. Hemodynamic measurements will be compared using the nonparametric Mann-Whitney test for unpaired data. To assess the accuracy of the PVi for correctly classifying patients who would respond to fluid challenge, receiver operating characteristic curves (ROC) will be performed and the area under the curves (AUROC) will be calculated. The optimal threshold of continuous variables will be chosen to maximize the sum of sensitivity and specificity. This is a pilot observational study that will involve 20 patients.