Thank you so much for the introduction, nothing to disclose. So the objective of this talk is going to be to define the fluid responsiveness and the importance in the ICU. We're going to discuss the role of different types of fluids in the ICU. We're going to evaluate the different options to assess fluid status and how to assess fluid responsiveness. We're going to discuss the importance of fluid overload in the ICU, and we're going to discuss the most recent evidence on fluid resuscitation in septic patients. So fluid responsiveness. So most of the journal are defining these concepts as an increase in cardiac output or a stroke volume more than 10% or 15% in response to a fluid challenge. So it's a normal condition to be fluid responsive. So that doesn't mean that a fluid responsive patient always requires fluid. So the reference standard is the measurement of cardiac output by thermal dilution. But most studies have used a VTI with an ultrasound. So we have to remember that fluid responsiveness responds to the Fran Stirling curve. In this picture, you see the normal LV function that is a red line that is an appropriate response to the fluid challenge, and in the blue line, you have a case or a patient with a reduced LV function in which any amount of fluid is going to increase the end-diastolic pressure and is going to reduce the responsiveness. So there are special situations in this concept of fluid responsiveness. It's only present in 50% of the patient with hypotension. So healthy avolimic volunteers normally have 100% of their volume responsiveness. So fluid responsiveness can be found in hypervolumic patients. So there is not a clear physiologic predictor that determines the transition from being fluid responsive to not being fluid responsive. That's the reason that we have this diagram in which the fluid responsiveness is a transition that we don't know specifically at what point the patient becomes no fluid responsive. So there are multiple studies, physiologic studies, that have determined that the minimum volume that is required to perform the fluid challenge is 4 cc per kilo, and an interesting concept is that the duration of fluid challenge can also influence the fluid responsiveness. There was a study that in 51% of the patients that received a fluid challenge over 10 minutes, they were responsive. But if the same amount of fluid challenge was given over 20 minutes, that responsiveness dropped to 28% of the patients. So in regards to the duration of fluid responsiveness, this was a very important physiologic study that was done in France, that they gave to a 143 patients that were septic with lactic acid more than two, and they had low urine output. So they initially gave the 20 to 40 cc per kilo initial fluid resuscitation, and then they gave the 500 cc of crystalloids. So under the definition of fluid responsiveness, there was an increase in the VTI more than 15% or the stroke volume. And as you can see there, 53% of the patients were fluid responsive at 10 minutes, and 47% of the patients didn't respond to the fluids. So now let's look what happened with these 53% of the patients that initially responded. So 49% of the patients became fluid responsive at 30 minutes, and 51% of them remained responsive at 30 minutes. And of the 47% of the patients that initially didn't respond to the fluid, only 6% responded to the fluids after 30 minutes. So this study suggests that fluid responsiveness is a time-dependent event in significant proportion of the patient and should be evaluated in two times, at least 10 minutes and 30 minutes, and ideally at 60 minutes. So something that we have to remember is that the slope of the Frank-Starling curve depends on the LV function. So we normally use dynamic tests to evaluate the fluid responsiveness of the patient. So most of the studies, most of the tests that have been evaluated are related to the mechanical ventilation, the passive leg raising, or the mini fluid challenges. And as you can see in this curve, those patients with a reduced slope in the Frank-Starling curve are those patients that are not going to have response to the fluid challenge. So how do we do the fluid responsiveness tests? So most of the studies are using generally the 500cc that is the most frequent. So the issue is that usually it's not reversible in case it doesn't work, and it doesn't predict for the response. So also has been reported the mini fluid challenge that is performed with albumin, 50 or 100cc, and it's given over one minute. So the response is a VTI more than 10%, and it predicts a good fluid responsiveness with a sensitivity of 95% and a specificity of 78%, and it was reported back in 2011. So the limitation is that it's a very tiny amount of fluid, and the errors in the measurement are very frequent. So how will we determine the cardiac output with ultrasound? This is just a general concept that I think every intensivist has to know. So first, we get a parasternal long view, and during this assessment, we measure the aortic flow diameter during the initial opening of the valve in centimeters. So half of that value is going to be the radius that we are going to use later. Then we have to be able to get the apical four-chamber view, and in that view, we're going to get the Doppler flow through the aortic valve, and we're going to do the VTI measurement. Then we apply and we calculate the SV using the VTI number and apply the radius, and then we calculate the cardiac output. This is the way that in most of the study has been calculated the cardiac output when we talk about fluid responsiveness. So let's talk about fluids and ICU outcomes. So the classic question is, does the type of fluid matter? So at this point in the ICU literature, we have four big trials, so almost 35,000 patients. So all of them are randomized. Most of the three of them are randomized. One of them is an open label. All of them are multicenter studies. The main comparison was normal saline versus balanced crystalloids. Plasmolyte was a balanced crystalloid that was used in most of the studies. As you can see there in the split and in the smart trial, the primary outcome was the AKI or the major kidney event of 30 days, and in the basic and plus trial was mortality of 90 days. Secondary outcomes also in patient mortality and AKI or AKI-requiring dialysis, and as you can see there, there is no difference. So, so far today, we can say that there is no difference between giving normal saline or using balanced crystalloids. However, so there is a recent meta-analysis that included these four studies, these four randomized trials, and also included most of the prospective studies with fluids, and this meta-analysis was almost 40,000 patients. And the same, the hospital and 20 or 30-day mortality was no different, and it was especially in those randomized control trials, and the difference was only observed in these observational studies. So, we can discuss about the amount of fluids, probably if we go over 50 cc per kilo or 60 cc per kilo, probably it's associated with prolonged mechanical ventilation, more AKI. So, there are multiple questions that are still remaining, but so far the evidence suggests that there is no difference between balanced and normal saline. So, now we have to talk about albumin. So, everybody remember the SAFE study back in 2004 published in New England Journal of Medicine, almost 7,000 patients. So, they compared 4% albumin versus saline, and the outcome was mortality, organ function, AKI with renal replacement therapy, ICU, and hospital stay. So, as you can see there, there was no difference in any of the primary outcomes. But it's something that I would like to highlight, something important is that when they did a sub-analysis of only the septic patient, of course, it's a sub-analysis that was published in Intensive Care Medicine in the same year, they found that probably in the septic patient there is a probability that using albumin is associated with better outcomes. But remember, this is a sub-group analysis after the initial randomization. So, we have to take with a grain of salt this type of studies. So, right now, according to this last Surviving Sepsis Campaign, the recommendation is, of course, use 30 cc per kilo of crystalloid fluids within the first three hours. And if we are going to be more specific, so the recommendation is crystalloids as a first line fluid. The role of balanced crystalloid is also recommended. And consider the use of albumin after the patient has received a volume resuscitation with crystalloid and the patient still is able to respond to more fluids. So, assessment of the fluid responsiveness using the hemodynamic parameters. So, what are the options and indications? So, we have to recognize that there are three ways or three different options to assess the fluid responsiveness. First is using static assessments, dynamic assessment, and then we have the provocation maneuvers. So, in the static assessment, we have the CVP, the physical exam, and the pulmonary artery occlusion pressure. So, let's talk about the CVP, and I want to highlight this concept that is very important. So, according to Professor Monet from France, so we should stop using the CVP as a predictor of fluid responsiveness. There are multiple studies suggesting that this marker, this parameter doesn't work to predict fluid responsiveness. And it's the same for the other static values, the same for the end-diastolic volume, the same for the pulmonary artery occlusion pressure, it's the same situation. But we have to always remember that CVP is an excellent marker for the termination of the preload. So, no for preload responsiveness, and we have to remember that preload is an important determinant of the cardiac function. So, another concept is the pressure gradient. So, the pressure gradient that is calculated using the MAP minus the CVP is an important hemodynamic parameter that is strongly associated with acute kidney injury and also with prolonged mechanical ventilation and pulmonary edema. So, I think that CVP still has a value, but no CVP for fluid responsiveness. So, before we continue, it's important to remember the likelihood ratio, how we interpret when we're going to evaluate this static assessment. So, if we have a likelihood ratio equal to 1, the test doesn't have any effect on the impact of the disease. So, ideally, the likelihood ratio of more than 3 is that it's associated with the disease is more likely or the test is better. Or if the likelihood ratio is below 0.1 is when the test is less useful for any condition. So, this is how we're going to be interpreting the test. Ideally, any of these tests, we are going to be targeting a likelihood ratio more than 3. So, we calculated, using multiple studies, the performance of these static tests. And as you can see there, the CVP of less than 8 has a likelihood ratio of 2.6. And the post-death probability of fluid responsiveness is only of 72%. So, the same with the decreased skin tumor or the prolonged capillary refill time. So, as you can see there, the likelihood ratio is only of 0.9. And the post-death probability is 48%. So, it's important to remember that probably the capillary refill time is like flipping a coin. So, when we talk about the dynamic assessment, so first we have to consider the pulse contour analysis. And in that area, we have the pulse pressure variation and the stroke volume variation. And we have the use of the ultrasound. So, first we're going to talk about the pulmonary artery catheter. Most of the studies with pulmonary artery catheter are in patients with cardiogenic shock or advanced heart failure. So, we don't have many studies, especially randomized studies, testing the role of a pulmonary artery catheter in hypovolemic patients or patients that are septic. So, it's an invasive hemodynamic monitoring. So, requires the placement of an introducer, ideally in the right IJ. So, it still has a role in the management of pulmonary hypertension, advanced heart failure, and usually in the post-cardiothoracic surgery population. So, usually the risks of the migration, the infection, the arrhythmias, and the PA rupture. So, it's limited by the valvular regurgitation and also needs a good training of the provider interpreting the information that is given by the pulmonary artery catheter. Something that I want you to remember is that there is a recent study evaluating the pulmonary artery catheter in cardiogenic shock patients. This was a retrospective study of almost 5,000 patients. And when you consider all the values that the pulmonary artery catheter can give you, so the complete assessment that includes the pulmonary arteries, systolic, diastolic, the wedge, and the pulmonary artery saturation, if we consider all these parameters, there is an improvement in the mortality of patients with cardiogenic shock, especially in those who has stage D or E of the SKY classification. So, let's talk about the pulse pressure and the stroke volume variation. So, normally, we evaluate that using a certain device. The classic is a flow track. This technology requires a normal sinus rhythm. So, it was validated in patients that were on mechanical ventilation, usually with a tidal volume 7 to 10 cc per kilo. It requires arterial line and is not validated for patients who has open abdomen or open chest. So a patient has to have a closed thoracic cavity. So the basis is that with hypovolemia, mechanical ventilation significantly alters the preload. So usually the threshold, the number that we have to remember is 10 to 15% to predict flu responsiveness. So on this technology has several limitations. Usually we have the mnemonic of limits. So usually low heart rate or increased respiratory rate give you false positive and false negative. Mechanical ventilation give you also some false negative. The increased intra-abdominal pressure give a false positive value. If we consider patient with the open thoracic cavity also is going to give us false negative. And the fact that the patient is breathing spontaneously can give false positive or false negative. So the point of care ultrasound, this is the most common use of the ultrasound in the ICU currently. It can provide a visual estimation of the ventricular size and is very important to determine the IVC or the SVC diameter. So usually the number that we consider is in the IVC diameter less than 1.5 centimeters or a collapsibility more than 40% has a very strong positive predictive value and very good sensitivity to predict a volume responsiveness of the patient or that the patient is hypovolemic. So the positive pressure ventilation alter these parameters because it can increase intra-thoracic pressure. So when we talk about the provocation maneuvers, we have to remember two of them, the passive leg raising and the respiratory occlusion test. So first we're going to talk about the physiology of the fluid response in the respiratory tract. So when we use positive pressure ventilation, we are increasing the intra-thoracic pressure that is going to reduce the preload to the right ventricle and is going to increase the preload to the left ventricle. So during expiration is going to be the opposite. So it's going to increase the preload to the right ventricle and at the same time it's going to increase the preload to the left ventricle. So initially, nothing or there is a small increase in the pulse pressure variation, but then it's going to decrease. So in order to evaluate fluid responsiveness, we need to make a pause and evaluate the change in the pulse pressure variation and that is going to determine if the patient is fluid responsive or not. So usually the recommendation is to do this maneuver for 15 to 30 seconds. The other maneuver that is more simple is the tidal volume challenge. So it's basically to change the tidal volume to ACC per kilo for one minute. We need to observe the change in the pulse pressure variation and the change of 3.5 has a very excellent sensitivity and a specificity of 94 and 100% respectively to predict fluid responsiveness. So I think that this is the easiest way to predict fluid responsiveness when the patient is on the ventilator. And then we have the passive leg raising. So it was validated in many studies. The study is described that it can mobilize up to 500 cc of intravascular volume. The increase in more than 10% in the stroke volume within 30 to 90 seconds is considered that the patient is responsive. So and we need to remember this technique because here are the when the errors happen. So we need to start the procedure from the semi recumbent position that is 45 degrees and we have to maintain the legs 3 to 5 minutes elevated. So usually what happens is the legs, the physician or the person who is performing the test is not maintaining the leg for 3 to 5 minutes and we have to measure the cardiac output. So here the change in the blood pressure doesn't have any significance. So what is important to measure is the stroke volume or the cardiac output. So on this measure, the measurement of the cardiac output should be do in real time within 3 to 5 minutes and that measurement has to be repeated when the patient is back to the original position. So in the table is explained the physiology of this technique. Here again, we are calculating the likelihood ratio of the dynamic assessment of the fluid status. So in this table, I put all the variables and as you can see here, I'm highlighting the respiratory occlusive pressure that are used using as a target 13% has the best likely ratio to predict fluid responsiveness and the next one is the IVC variation of more than 15%. So I didn't include the pulse pressure variation of the stroke volume variation. Why? Because those parameters needs a high tidal volume. This is another table that also is highlighting that the variation in the IVC more than 40% has an excellent likelihood ratio between 3.5 and 9.3 to predict fluid responsiveness. So in this another table, we are just summarizing the limitation, the threshold and the preferred population to perform these tests. As you can see there, the passive leg raising and the respiratory occlusion test you can perform in patient on mechanical ventilation and ideally this patient shouldn't be breathing spontaneously and the same with the IVC. So the collapsibility index of more than 40% can be implemented in patients that are on mechanical ventilation. But here I'm highlighting in red the fact that using the stroke volume variation of the positive pulse pressure variation in patients that are on mechanical ventilation should be interpreted with caution if you are not using a tidal volume of 8 to 10 cc per kilo. So according to the Surviving Sepsis Campaign, so the recommendation is to always use dynamic measurements to guide fluid resuscitation and this is also supported by the Society of Critical Care Medicine here in the United States. So let's talk about fluid causing venous congestion. So first we have to recognize the important implication of volume overload in the ICU. So it has been associated with prolonged mechanical ventilation, higher risk for ventilator-associated pneumonia. So we all know that it can decrease the LV function. There are more risk of arrhythmias, especially atrial fibrillation. We know the impact of volume overload with acute kidney injury and there are new data suggesting that fluid overload can trigger ileus, poor GI absorption and intra-abdominal hypertension, liver congestion, and also skin infections. So this is an analysis of the patient with reduced ejection fraction that is showing that when we have an increased venous congestion and we have higher CVP, it's associated with lower GFR. And the magic number that triggers lower GFR is any CVP more than 12 is when we have the inflection in the core. So when we evaluate with ultrasound venous congestion, there are two markers, two parameters that we have to consider. First is the IVC collapsibility index, less than 40%. As I'm showing there, the sensitivity is 88% and specificity of 93%. And then we have the other one that is the IVC diameter index that is also considering the body surface area. So the number is 1.5. Those two parameters are very strong predictor of the patient's volume overload. And more recently is coming the new concept that is the VEXUS. So this was initially validated by a group in France, but is also now reported by Dr. Bronco and his group in Italy. So on the initial trigger is evaluating the IVC, if the IVC is more than two centimeters, the next step should be to evaluate the portal vein and then evaluate the renal veins. And based on that, we do an scoring system. So a grade three is considered a severe congestion in which you have abnormalities in the portal veins and also in the renal veins. So what happened with the VEXUS is that we are still waiting in more studies associating this tool with clinical outcomes. So the main outcome that has been evaluated with VEXUS was duration of mechanical ventilation. So we are still waiting to evaluate VEXUS with acute kidney injury or with mortality. So now the concept that is coming more recently is we give more fluids or we start early the vasopressor. So there are a lot of controversies on fluid resuscitation based on the recent studies that have been published. So let's start with a fresh trial. This was published in CHESS. So it was a multicenter randomized study. The issue was that it was only 150 patients and they determined the fluid resuscitation using the passive leg raising assessment. So this is the algorithm how they did the trial. So one of the negative points about the trial is that it wasn't blinded. So the primary outcome that it was a composite outcome show a decrease in the ICU duration of a state or fluid balance of 72 hours, and it was positive. So secondary outcomes also show a decrease in mechanical ventilation with a difference of 16% and also decrease in renal replacement therapy of about 12%. But the limitation is, as I mentioned, the lack of blinding, the modified intention to treat. So the issue that the primary composite outcome was driven statistically significant by a surrogate outcome that was the amount of fluid. So what triggered that the composite outcome was significant was that the amount of fluid was lower in the population, in the group of patient that was receiving fluid based on the passive leg raising. And also this trial was stopped early after the interim analysis, and there was a significant inequality between male and females in this trial. So now the newest trial is a classic trial. It was a randomized control trial of 1,500 patients, all of them with septic shock. So the mortality in this population was 42%, 43%. So this population was assigned to a standard or restrictive fluid within 12 hours of septic shock. And all of these patients received one liter before the randomization. So both groups have very high mortality. So the median difference in fluid was about one liter in the first day, and approximately two liters from day five and during the follow-up. So from day one and throughout the study, the cumulative fluid balance in the restrictive group was consistently about 700 cc less in the standard group. So there was no difference. As you can see in the curve, there was no difference between the standard and restrictive group. The mortality was similar, and also there was no difference in the incidence of AKI or ischemic events. Days alive without life support also was similar, and days alive out of the hospital also was similar. So on the most recent trial was the Clovers. This study was also a randomized study, 1,500 patients. The only difference if we compare the classic with the Clovers is the Clovers trial has a mortality of only 15%. So in this study also they randomized to restrictive fluid or liberal fluid, but also they evaluated the initiation of vasopressors, if the vasopressor was initiated early or delayed based on the amount of fluids. And the results is suggesting that the group with the restrictive fluid, there was a difference of almost two liters with the group that received liberal amount of fluids. The group that was assigned to the restrictive group received early initiation of vasopressors, and it has very high doses of vasopressor if we compare with the standard with the liberal group. So and as you see there, the results was no significant. There was no difference in mortality. There was no difference in organ support. No difference in the duration of mechanical ventilation, and no difference in AKI. So this trial is suggesting that probably early initiation of vasopressors or probably using a restrictive amount of fluid is not impacting significantly the mortality of this patient. But we have to consider that the mortality was only 50% in this group. Just to finish, this is the proposed algorithm for fluid administration and assessment of fluid responsiveness. So the first question that we always have to ask is, is the patient hypovolemic? Is there an obvious fluid loss? If the answer is yes, of course, we have to start giving fluids. If the answer is no, we can start implementing this fluid responsiveness assessment, and we have to consider three specific situations. Is the patient spontaneously breathing? Is the patient on the ventilator? Is the patient having ARDS and receiving low tidal volume? Because if the answer is yes, there are very limited amount of tests that we can use. If the patient is on the ventilator, so the only test that is validated is using the passive leg raising, the respiratory occlusion test, or the mini fluid challenge. And with regards to this algorithm, important consideration is no specific algorithm has shown to improve ICU outcomes, and always each institution must evaluate their performance and ability to implement any of these tests. So in conclusion, so fluid responsiveness is variable and is time dependent and is present in around 50% of the hypovolemic patient. So the preferred fluid in septic patient today, based on the evidence, is balanced crystalloids. The static variables, such as the CVP, has a very poor role in the assessment of fluid responsiveness. Dynamic variables are the preferred and are recommended based on guidelines, and we should consider always when we are assessing the patient. The VEXUS is a new tool, and we have to remember that so far it's only associated with VEXUS of three. It's only associated with prolonged mechanical ventilation. So far we don't have a data of VEXUS with mortality, but it's a good predictor of volume congestion. And based on the Clover's trial, probably the use of restrictive amount of fluids and early initiation of vasopressor have no negative impact on clinical outcomes in patient with septic shock. Thank you so much, and I want to thank my collaborator in this project, that is Dr. Sierras. And if you have any questions, please, this is my email. Thank you so much.

fluid responsiveness

ICU

types of fluids

fluid overload

fluid resuscitation

septic patients

VEXUS tool