Thank you very much Michael, good morning. And I first disclose my collaboration with Belgian Medical Systems and with Baxor. Basically what are we speaking about, a severe critical ill patient, for instance with septic shock and pneumonia, the initial phase, with high PEEP and low PF ratio. And indeed in this patient the issue is really should I give fluid or not to this patient. And in this patient I know that, I'm sorry it's not the right presentation, they loaded the previous one, you know, anyway, I will make it shortly. When I give fluid to this patient they may respond or not to fluid administration because of this slope of the Frank-Stalin curve. And the problem is that we know that in some patients fluid have really harmful effects. And in this patient not only fluids are ineffective but also they may be harmful and basically by increasing lung water. And that's why basically not only fluid are ineffective but also they may be potentially harmful leading to the conclusion that one may assess fluid responsiveness. Basically fluids are drugs, not only they have adverse effects but also they have inconstant efficacy and the same as you would not give antibiotics without predicting their efficacy, you would not give, you should not give fluid without predicting their efficacy. In our mind we have such an algorithm, if cardiac output is low I ask myself should I give fluid or not. And then if there is no obvious fluid, if there is no obvious hypovolemia, hemorrhage, etc., then I know that I should predict fluid responsiveness. And the right way to do so is to do it dynamically. I mean to assess any change in preload, spontaneous change in preload, and to assess the response in terms of stroke volume and cardiac output. And today we have several tests and indices to assess preload responsiveness. Some of them use the ventilator, some of them mimic a fluid effect, and in fact let's review during the next minutes what are the advantages and drawbacks of this different test. The fluid challenge is definitely the easiest way to do so. Give fluid and look at the response of fluid. The problem is that if you do so you induce fluid overload. I mean that obviously you cannot remove fluid that you've given to the patient if the patient does not respond positively. It means that if you repeat this fluid challenge inherently it induces fluid overload. In addition, it requires a direct measurement of cardiac output to evidence its effect. That's why the idea of a mini fluid challenge might be interesting. You just give a small amount of the fluid bolus, for instance 100 or 150 ml, and then you expect that the response to that will predict the response to the next, to the rest of the fluid bolus. Indeed, it induces less fluid overload. It is reliable. The problem is that we have today different thresholds, and you see that the different thresholds have been reported in the literature, and definitely this may be a drawback. So we need more studies to assess the exact measurement of cardiac output. The point is that you require a precise measurement of cardiac output. Of course, if you give small amounts of fluid, it can only induce small changes in cardiac output. ECHO is likely not the best way to do because the precision of ECHO is not that good, and the smallest changes in VTI you may trust are only 10%. That's why if you keep the hand of your patient on your patient when performing the test, the precision is much better, and in such a case if you do the mini fluid challenge with ECHO, please keep your hand on the patient and the ECHO beam in the LV outflow tract, or use some more precise measurements of cardiac output. In any way, it still requires fluid infusion that you cannot remove. Passive leg raising, on the contrary, is a reversible fluid challenge. It's a preload challenge. When you transfer a patient to the passive leg raised position, indeed, you induce the transfer of roughly 300 milliliters of blood, and this is a valuable and reliable way to assess preload responsiveness. There are many studies, many positive studies, and meta-analysis, a well-defined threshold for positivity included in the surviving sepsis campaign as you know. Let's say a few words about passive leg raising. The key is that to assess the test, you cannot use the simple pulse pressure, even pulse pressure, I mean, on the bedside monitor. So you need something else that assesses throat volume. I don't insist on that. I suggest maybe pulse contour analysis, esophageal doppler, echography, just to say a few words about platysmography, and some new data about that. This plat signal we have under our eyes every day is made of two parts, and the ratio of the pulsatile and non-pulsatile portion, the perfusion index, is related to stroke volume. And we showed in this study that if cardiac index in blue increases during passive leg raising and fluid infusion, the same does with this perfusion index. This data must be confirmed, but if confirmed, you see that that could be a widely available and cheap way to assess the effects of passive leg raising in spite of some instability in some patients. You may use bioreactants. You may also use just changes in pulse pressure variation as the decreases in PPV during a passive leg raising test help assess the effects of the test, again, without any cardiac output monitor. In any way, this may be non-reliable in patients with intra-abdominal hypertension. No time to speak about that. PPV and SVV, the first ways described to assess preload responsiveness and, indeed, ventilation induces changes in preload that help you assess the response, the slope, in fact, of the Frank-Stalingraff just by looking at changes in the amplitude of pulse pressure. The main drawback, as you know, is that in many patients, it's not possible to assess the effects, arrhythmia, spontaneous breathing, and ARDS patients when you cannot use PPV and SVV. A solution may be, in case of flow tidal volume, to use this tidal volume challenge we've described years ago with some Indian colleagues in Shilamayatra. Increase tidal volume from 6 to 8 for a minute and look at the response of PPV, and if PPV increases during this short increase in tidal volume, then this may mean that the patient is preload responsive. Again, the problem is that the threshold of positivity is not that well-defined for the moment. Look, 1, 3% different thresholds have been reported, so we likely need some more validation for this tidal volume challenge. IVC and SVC diameter changes, inferior, superior, vena cava changes. My key message is that these are the least reliable indices of fluid responsiveness we have, demonstrated by many small-size studies and their meta-analysis, and by this large negative study, look, the area under the record for predicting preload responsiveness in that study was only 0.65, very, very poor. And it's likely related to the fact that these changes in IVC are not pure markers of preload responsiveness. I mean that these changes in IVC diameter during ventilation are due to the changes in the intramural pressure related to preload responsiveness, CVP changes, but also to many confounding factors. I mean the intra-abdominal pressure, the volume of the vein itself, its compliance, not a pure marker of preload responsiveness, explaining its low reliability demonstrated by many, many studies. In addition, it shares many of the contraindications with PPV and SVV, so not a good marker. Finally, the end-expiratory occlusion test demonstrated more than 10 years ago. The principle is very easy. Each change during ventilation in intrathoracic pressure is associated with decrease in preload. So if you stop mechanical ventilation for a few seconds, more than 10 seconds, preload increases, reaches a maximum. And if you wait enough, more than 10 seconds, the preload bolus crosses the pulmonary circulation. And if in response stroke volume and cardiac output increase, it's likely that your patient is preload responsive. Again, it's a way to assess the slope of the Frank-Starling curve at the bedside, this time by using the ventilator. This end-expiratory occlusion test has been validated by many studies. We recently published this meta-analysis, a well-defined threshold of positivity. Look at the area under the RAC curve compared to the IVC changes, for instance. Again, which devices can you use to assess the positivity of the test? Pulse contour analysis, but all patients do not have that. You can use echography, not very precise, and I don't have time to show that. Perhaps bioreactants, esophageal doubler, and again, the changes in this perfusion index. So likely a good way and cheap way to assess the effects of this last test we have to assess preload responsiveness. The drawback is that you cannot use the test, of course, in patients that do not tolerate these quite long interruption of mechanical ventilation. So to conclude, you see that we have different tests and indices. I tried to show you their advantages and drawbacks. I think that the three ones that can be used at the bedside in a reliable way are the mini-fluid challenge provided you have a precise measurement of cardiac output, passive leg raising can be used in many patients, not very easy to perform, and finally, the end-expiratory occlusion test. Thank you very much for your attention.

fluid administration

severely ill patients

fluid responsiveness

bedside tests

mini-fluid challenge