Performance of CO2 Changes to Predict Fluid Responsiveness in Spontaneously Breathing Volunteers

Predicting fluid responsiveness in critically ill patients is of paramount importance. It can help define an adequate fluid balance. Overzealous fluid administration is poorly tolerated and has been associated with poor outcomes but so has insufficient administration. Currently available predictors of fluid responsiveness rely on invasive monitors and require patients to be on mechanical ventilation. It is thus important to develop non invasive novel methods to assess fluid responsiveness to provide an accurate management for a favorable outcome. We propose a readily available non-invasive method that relies on improvement of the ventilation perfusion mismatch as recorded by end tidal CO2.

Ventilation of physiologic dead space is part of a spectrum of mismatch between ventilation and perfusion of the lungs. The extent of pulmonary dead space varies depending on factors affecting pulmonary perfusion (e.g. pulmonary capillary hydrostatic pressure) and alveolar pressure (e.g. positive pressure ventilation). Compromised pulmonary capillary perfusion can lead to ventilation-perfusion mismatch in a patient with clear conductive airway and adequate alveolar oxygen pressure. Alveolar dead space results in decreased CO2 exchange that translates into lower levels of expired CO2.

Stroke volume of the right ventricle is a major determinant of the pulmonary capillary perfusion. Right ventricular cardiac output can be increased by passive lower limb elevation maneuver, which ultimately results in improvement of the ventilation to perfusion ratio. This effect leads to a higher participation of perfused (and ventilated) alveolar units in gas exchange and narrowing of the gradient between arterial and expired CO2 concentration. Performing a passive leg raising (PLR) maneuver leads to stroke volume enhancement in both healthy patients and in those experiencing hemodynamic instability. Responsiveness to PLR can be assessed by different methods including echocardiography and pulse pressure variation. Left ventricular cardiac output (LVCO) can be easily measured by transthoracic echo and be used as a surrogate of right ventricular preload changes. LVCO can thus be used to assess the fluid responsiveness of PLR and the effects of on end tidal CO2 that ensue.

We propose this study to test the hypothesis that expired CO2 is a reliable predictor of fluid responsiveness after performance of the PLR maneuver, based on the assumption that increasing right ventricular output causes a reduction of the ventilation to perfusion ratio, leading to increased levels of expired CO2. T