A Prospective Investigation of Pleth Variability Index (PVI) as a Dynamic Parameter of Fluid Responsiveness in Children

Background: Hypovolemia is a frequent cause of circulatory failure in anesthetized and critically ill children. Fluid boluses are the first line of treatment aimed at optimizing stroke volume and ultimately improving oxygen delivery. A patient whose stroke volume significantly increases as a result of a fluid bolus is deemed 'fluid responsive'. Traditional parameters of circulatory status (e.g. blood pressure, heart rate, central venous pressure) have been shown to be ineffective at predicting fluid responsiveness. Dynamic parameters of fluid responsiveness, such as pulse pressure variation (PPV) and stroke volume variation (SVV), reflect hemodynamic changes secondary to intermittent positive pressure ventilation. These indices have proven to be superior to the traditional static measures of circulatory status in predicting fluid responsiveness. Furthermore, the use of these dynamic parameters in a goal-directed fashion has been shown to improve outcome in high risk, adult surgical patients. As such, there is increasing interest in using dynamic parameters to guide fluid resuscitation in mechanically ventilated infants and children. Unfortunately, there is a paucity of literature examining the validity of these dynamic parameters in children. The need for arterial vascular access, which is less common in small children, has also limited the ability to investigate these parameters. Recently, a non-invasive dynamic parameter has been developed examining the plethysmographic waveform from a pulse oximeter. The pleth variability index (PVI) represents the dynamic changes in hemodynamic performance that occur during the respiratory cycle. Studies of PVI in adults have shown it to be a reliable predictor of fluid responsiveness. However, conclusive data in children are lacking. Hypothesis: This study is designed to investigate the ability of PVI to non-invasively predict fluid responsiveness in mechanically ventilated children. The hypothesis is that a higher PVI will predict significant stroke volume increases in response to a fluid bolus in mechanically ventilated children. Methods: The study population will consist of 20 children undergoing cardiac catheterization who will have normal (non-shunted) physiology at the conclusion of the procedure, ages 1-12 years. All subjects will receive a PVI measurement as well as a stroke volume measurement via transthoracic echocardiography. Subsequent volume expansion will be achieved with an intravenous bolus of 10 mL/kg of isotonic crystalloid (normal saline) administered over 10 minutes. After completion of the bolus, an additional PVI measurement and stroke volume measurement will be obtained. Significance: Hypovolemia is a significant cause of harm in anesthetized and critically ill infants and children. In order to successfully resuscitate these patients, we must gain a better understanding of how to optimize cardiac output and oxygen delivery. A non-invasive monitor which indicates stroke volume responsiveness has the potential to better guide fluid resuscitation and improve outcomes.