Fluid administration is one of the first-line therapies for most patients with acute circulatory failure. And it was supposed to increase cardiac preload and thus output significantly. However, it only has this effect if cardiac output is dependent on cardiac preload, that is if both ventricles work on the ascending part of the Frank-Starling curve. Our group had already shown that, in half of the critically ill patients admitted in ICU, fluid administration is likely to exert only deleterious effects without any hemodynamic benefit. To predict whether it will exert beneficial effects or not before administering the fluid therapy, a "dynamic approach" has been developed. It consists of observing the effects of changes of preload induced by various tests on cardiac output. 1) Pulse pressure variation: mechanical ventilation induces cyclical changes in cyclical changes in cardiac preload and right ventricular afterload due to cardiopulmonary interactions. If both ventricles are in a preload-dependent state, these variations will induce a cyclical variation in stroke volume. The latter being physiologically related to the pulsed arterial pressure (systolic - diastolic), the respiratory variation of the pulsed arterial pressure (PPV) indicates the existence of a preload-dependence of the two ventricles. 2) The End-expiratory occlusion test (EEOT): this is another method that takes advantage of heart-lung interactions to predict fluid responsiveness in ventilated patients. During mechanical ventilation, each insufflation increases intrathoracic pressure, which hinders systemic venous return. Thus, interrupting the respiratory cycle at the end of expiration inhibits this cyclic hindrance to venous return, increases cardiac preload and cardiac output if both ventricles are preload dependent. The duration of the EEOT must be at least 15 seconds. 3): Passive leg raising (PLR): when a patient is in a recumbent position, the elevation of the lower extremities and the horizontalization of the trunk passively transfers a significant volume of blood from the lower part of the body to the heart chambers and mimics volume expansion.
Numerous studies have reported that the increased cardiac output induced by PLR predicts fluid responsiveness. There is always the question of the prevalence of cases in which the different predictive indices of fluid responsiveness are not applicable and data on this issue are scarce, incomplete, and unsatisfactory.
Few studies have systematically investigated the number of patients in whom PPV cannot be used in the ICU settings. Some studies have reported a very low prevalence of cases where PPV was usable but they included the entire ICU population on a given day, including many patients who did not have an acute circulatory failure, which had no sense since PPV is only used in patients in whom the question of fluid therapy arises. Other studies have reported a higher prevalence of cases where PPV is usable, but they have only looked at the first 24 hours of hospitalization or have focused on patients with an unstable hemodynamic event. Finally, no study has ever studied the prevalence of cases where the respiratory variation of the PLR, or the EEOT are not applicable.
