OACCM Advisor's Blog # 3 - September 2011, Dr. McGee on Hemodynamics - William T. McGee, M.D., M.H.A.

For this month as we approach the end of the summer, I would like to wish all of my readers a good close to whatever leisure time they were able to enjoy during a well deserved summer break. Hopefully, everyone was able to get some time away, although recognizing that the present environment in medicine and especially critical care medicine is seemingly making this more difficult as time goes forward. I also would like to congratulate everyone for what you do on a daily basis. This is a difficult job that is not getting any easier and only through the dedication of all the practitioners within an intensive care unit can this important job be accomplished. Our patients and their families also appreciate your commitment. By way of introducing the Physiologic Optimization Program clinically, I think I will start with a case and simply illustrate the steps and how we use physiologic optimization using dynamic variables of fluid responsiveness along with cardiac performance to treat patients with a variety of critical clinical illness in the intensive care unit.

slide 1

● The first slide represents the Physiologic Optimization Program, operationally: 1. A clinical question, often oliguria or hypotension, is always a necessary first step before attempts to manipulate physiology. Our strong contention is that if the patient is doing well; leave them alone. Other the other hand, when questions come up regarding blood pressure and urine output we prefer a physiology-based approach to management in the intensive care unit. a. Physiology is interrogated at the level of volume responsiveness, SVV (please recall from the prior installment of my blog the conditions where the use of stroke volume variation as a measure of volume responsiveness has validity and that the greater the SVV the more volume responsive the patient will be). For the ensuing discussion as was the case in the patient that will be presented, the patient is on controlled mechanical ventilation and has a regular rhythm. b. The first step looks at fluid responsiveness as assessed by the stroke volume variation (SVV) if stroke volume variation is > 13%. We use the left hand side of the algorithm which assesses this patient to be volume responsive and the patient is given a fluid bolus. Typically 1 to 2 L of crystalloid but depending on the clinical situation, blood, blood products, or colloid maybe appropriate. For those patients that are not volume responsive considering this algorithm, SVV < 13%, we interrogate the physiology at the level of the stroke index and depending on the clinical scenario apply physiology-based therapy. c. Without getting into a lot of detail, I will provide some clinical clarification on Pathways 1, 2 and 3. In Pathway 1 when the stroke index is normal, these are typically resuscitated patients who may be vasodilated, i.e. septic shock, and a pressor is indicated knowing we have adequately volume resuscitated the patient. Pathway 3: When the stroke index is elevated especially considering the data from the FACTT Trial for those patients with acute lung injury or ARDS we are know not only stopping further volume infusion but also using diuretics. (The NHLBI ARDS Clinical Trials Network: NEJM 2006;354:2564-2575) Pathway 2: Finally for those groups in the middle who are no longer volume responsive but still have inadequate cardiac performance an inotrope may be indicated. Almost always in this group of patients we try to obtain echocardiography before the application of inotropes as this adds to our physiologic understanding of the patient’s physiology and the intelligent application of inotropes. Considering that brief introduction, let’s analyze a clinical case treated in the ICU at Baystate Medical Center.

slide 2

● Slide #3-5 is a case I am sure we all recognize with multiple co-morbidities, recent hospitalization, who presents febrile, in shock, tachycardic, and tachypneic to the emergency room.


slide 3


● Slide 4. The chest x-ray is diffusely abnormal. Does this represent diffuse pneumonia, hydrostatic pulmonary edema or acute lung injury? The white count is elevated with a significant left shift strongly suggesting an infectious process. The patient has an elevated anion gap with acute renal failure. Is there an element of lactic acidosis complicating the metabolic acidosis from the renal failure. A lactate measurement would be useful in this regard. The blood gas shows both respiratory and metabolic acidosis with hypoxemia on supplemental oxygen. On the EKG we see a left bundle branch block again emphasizing at least a possible contribution from coronary ischemia to this clinical picture.

slide 4

● Slide 5. Our impression, clinical septic shock complicated by acute kidney injury possibly an acute coronary syndrome and pulmonary edema. But for those of you reading this, you understand that this is what we do on a daily basis. On the other hand, it highlights the complexity of our job and anything that can make our tasks easier is welcome in my practice.

slide 5

● Slide 6. Shock is being treated with volume resuscitation and Levophed. The patient is not making urine and because of this, we would like to know two things: 1. Volume responsiveness(SVV); i.e. is further volume resuscitation warranted, 2. Cardiac performance utilizing the indexed stroke volume (SI). What is the output of the heart on a beat to beat basis? These data will allow titrated precise management of the hemodynamic physiology. Initial Treatment:

slide 6

● Slides 7 & 8. What we find from evaluating the physiology is that stroke volume variation is elevated at 24% suggesting that the patient would still benefit from further fluid therapy and the stroke index is low at 22cc/beat/m2. The patient receives a 2L bolus and not only does the stroke index increase; the outcome that we are looking for with the fluid therapy, the stroke volume variation decreases to 17%. Further Treatment:

slide 7

slide 8

slide 8


● Slides 8 & 9;10 & 11. As we proceed through the algorithm SVV is still greater than our threshold for volume responsiveness of 13% and the patient receives an additional 2L bolus. With this treatment the SVV further decreases and most importantly the stroke index continues to rise. An additional benefit of the improved cardiac performance is the patient now begins to make urine. As the SVV remained above our threshold value of 13% another 2L bolus was given. This resulted in a further increase in the stroke index to 43 within the normal range. Additionally, the SVV of 7 suggested no significant further benefit from additional volume therapy. The important outcome of following these data in a systematic fashion is not only that it guides resuscitation; it also provides a physiologic end-point of volume therapy. With the stroke index in the normal range, the output of the heart on a beat per beat basis is adequate. There are multiple variations on this theme. One question that sometime comes up in a patient who is continuing to require a vasopressor, is can I make the cardiac performance even better and then eliminate the need for vasopressors. In this case, the information is clear and precise volume management is important for a patient with both acute kidney injury and acute lung injury. However, if we need to prove this to ourselves and sometimes we do, an additional fluid bolus focusing more on the change in stroke index would not necessarily be wrong and in fact, may further our understanding of how to use this technology. Clearly if there is a no additional increment in stroke index we can prove to ourselves independent of stroke volume variation that the patient is on the flat part of the Frank Starling Curve and volume is not the answer. As we become familiar with this technology and the application of dynamic variables of volume responsiveness, we will all have occasion to do this.

slide 9

slide 10

slide 10

slide 11

slide 11

● Slide 12, Physiologic Optimization Program, reviews the case and the application of the algorithm. To summarize, our patient was admitted with clinical septic shock complicated by acute kidney injury, acute lung injury and possibly an acute coronary syndrome. Following the Physiologic Optimization Program we were able to move through his resuscitation with a high degree of certainty of our fluid management using the data pair; stroke volume variation and stroke volume index. Despite a fairly good initial resuscitation these data suggested that more volume was required. An additional six liters of crystalloid were given as the stroke volume variability remained above 13%. Following this, stroke index normalized, stroke volume variation was 7. Throughout the patient’s clinical diagnosis was septic shock. We had both a clear end-point now to determine the adequacy of our resuscitation and we expected the patient to be vasodilated because of the septic shock diagnosis and competently could use a vasopressor without volume therapy at this point. Ultimately as shown at the bottom of slide 10, stroke index continued to increase to reach a supra physiologic value of 58cc/beat/m2 later that afternoon. The patient had a clinical diagnosis of acute lung injury/ARDS all along, but with these data it was simple and physiologically correct to begin early diuretic therapy, despite the fact that the patient was on a vasopressor. Had the patient had clear x-ray and lungs another appropriate course of action would have been simply watchful waiting to see if diuretics were required later in his course. As our patient always had significant lung injury, minimizing the additional physiologic derangement consequential to volume overload, diuretics were indicated. The impact of diuretic therapy can also be followed precisely. We are often able to diurese similar patients, perhaps not this early in the course of their illness, without significant change in cardiac performance. Moving them to the left in the Frank Starling Curve while simultaneously assuring good cardiac performance. Alternatively, if stroke index drops significantly with this therapy as long as we are paying attention it is simple to replace the diuresed fluid. That is why we left the “?” in box under diuretic therapy feeling that it is indicated for ARDS based on the outcome of the FACTT trial, but not necessarily so if the patient does not have acute lung injury/ARDS (The NHLBI ARDS Clinical Trials Network: NEJM 2006;354:2564-2575) .

slide 12

● The final outcome is presented in the last slide #13 and although this is a single case, I think it emphasizes the impact of using physiology at the bedside to manage hemodynamics and specifically fluid therapy in the critically ill. 1

slide 13

June , 2011 - Blog # 2 - Dr. McGee on Hemodynamics

May , 2011 - Blog # 1 - Dr. McGee on Hemodynamics

William T. McGee, M.D., MHA, FCCP is Associate Professor of Medicine and Surgery at Tufts University School of Medicine, Boston, Massachusetts.  His interests are in ARDS, vascular access (pulmonary artery catheterization), sepsis, nutrition, and nosocomial pneumonia.  He has published > 67 papers, chapters and abstracts.  He is the principal investigator for clinical trials studying the efficacy and safety of rfPAF-AH for the prevention of ARDS in patients with severe sepsis.  He is a three-time recipient of Excellence in Teaching Award from Tufts University School of Medicine, the Society of Critical Care Medicine Internal Medicine Specialty Award for “Influence of Insurance Status on Pulmonary Artery Catheter Use” and The Presidential Citation Award from the Society of Critical Care Medicine for outstanding contribution to the Society during 1999.  He is a Director for the Fundamentals of Critical Care Support Course for the Society of Critical Care Medicine.  He is on the review board for the Journal of Intensive Care Medicine.