Prediction of fluid responsiveness in ventilated patients

Fluid administration is the first-line therapy in patients with acute circulatory failure. The main goal of fluid administration is to increase the cardiac output and ultimately the oxygen delivery. Nevertheless, the decision to administer fluids or not should be carefully considered, since half of critically ill patients are fluid unresponsive, and the deleterious effects of fluid overload clearly documented. Thus, except at the initial phase of hypovolemic or septic shock, where hypovolemia is constant and most of the patients responsive to the initial fluid resuscitation, it is of importance to test fluid responsiveness before administering fluids in critically ill patients. The static markers of cardiac preload cannot reliably predict fluid responsiveness, although they have been used for decades. To address this issue, some dynamic tests have been developed over the past years. All these tests consist in measuring the changes in cardiac output in response to the transient changes in cardiac preload that they induced. Most of these tests are based on the heart-lung interactions. The pulse pressure or stroke volume respiratory variations were first described, following by the respiratory variations of the vena cava diameter or of the internal jugular vein diameter. Nevertheless, all these tests are reliable only under strict conditions limiting their use in many clinical situations. Other tests such as passive leg raising or end-expiratory occlusion act as an internal volume challenge. To reliably predict fluid responsiveness, physicians must choose among these different dynamic tests, depending on their respective limitations and on the cardiac output monitoring technique which is used. In this review, we will summarize the most recent findings regarding the prediction of fluid responsiveness in ventilated patients.