Nuances in Resuscitation Part II: EGDT In Sepsis

By Justin Bright, MD
Senior Staff Physician
Henry Ford Hospital, Dept. of Emergency Medicine
Detroit, MI

Edited by Alex Koyfman, MD

 
In November 2001, Dr. Manny Rivers and his colleagues published an article in the New England Journal of Medicine on Early Goal Directed Therapy in Sepsis.  At the time, sepsis was not a new concept, nor was the treatment of it.  Where I believe the real genius in EGDT lies is in a fanatical focus on early recognition of sepsis by utilizing SIRS criteria, as well as developing an algorithm with definable objective treatment goals to assist providers in understanding if their treatment selections are in fact working.  The basic questions in EGDT therapy are:

  1. How much fluid is enough?
  2. Are the vital organs being perfused appropriately?
  3. Is there adequate oxygen delivery and utilization by those vital organs?

Since Rivers published his article in 2001, it has been met with both acclaim and controversy.  EGDT utilizes central venous pressure monitoring, lactate trending, SvO2 monitoring, vasopressor therapy, and sometimes, blood transfusions to optimize resuscitation of the septic patient.  I believe the controversy is not in whether or not it works, as multiple studies have demonstrated a reduction in morbidity and mortality.  Instead, the controversy lies in what is the best modality to answer the basic questions of sepsis resuscitation, and whether some of the aggressive steps recommended in the initial study are necessary or even practical in many emergency departments across the country.

The first step in the resuscitation of septic shock is recognizing it early, with the goal being identifying patients with 2 or more SIRS criteria before you have identified a clear source of infection.  The morbidity and mortality of a patient identified with 2 SIRS criteria is 7%, compared to 20% for those with severe sepsis, and 43% for septic shock.

The SIRS criteria are:

  • HR > 90
  • RR > 20 or PCO2 < 32
  • WBC > 12,000 or < 4,000 or more than 10% bands
  • Temperature > 38 C or below 36 C

Septic shock occurs when an identifiable source of infection causes an inflammatory cascade and reduction in adequate tissue perfusion, thereby resulting in organ dysfunction.  At its very core is a lack of oxygen delivery to vital tissue, causing a flip from aerobic to anaerobic metabolism to generate ATP.  Anaerobic metabolism is a very inefficient way to generate fuel for the body, and comes at a great metabolic cost.  A byproduct of anaerobic metabolism is the generation of lactate.  A serum lactate > 4 mmol/L has been demonstrated to have a significantly higher morbidity (28% vs 9% for levels less than 4).  Lactate is also a very helpful marker because you can repetitively draw levels throughout the resuscitation process to establish whether or not what you are doing is helping the patient.

In contrast to the hypovolemic shock of trauma, septic shock causes vasodilation of the vasculature.  As a result, you can continue to aggressively infuse IV crystalloid with no change in blood pressure because there is a dramatic reduction in systemic vascular resistance, or “squeeze.”  So the question becomes, how much fluid hydration is necessary, and how do you know when you have reached your hydration end-point.  Or as many physicians like to say, “How do you know when the tank is full?”  In EGDT, the goal is fluid resuscitation with .9NS until you reach a central venous pressure (CVP) of 8-12 mmHg (utility has been widely challenged).  The CVP is measured using a central venous catheter placed in either the internal jugular vein or subclavian vein.  Femoral central lines are not conducive to accurate CVP measurement.  If the CVP is below 8 and the patient is hypotensive, you continue to administer more fluids.  If the CVP is adequate, you need to move on to the next step of EGDT or run the risk of fluid overload.  Since the initial publication, much discussion has focused on less invasive and alternative ways to measure the CVP.  Use of bedside ultrasound measurement of the collapsibility of the inferior vena cava has become the new trendy way of monitoring central venous pressure.  It is certainly less invasive, and more and more emergency departments have access to bedside ultrasound as well as providers who are comfortable using it in day-to-day medical practice.  I think the important thing to remember is, CVP monitoring is very important to adequately resuscitate a septic patient.   The manner in which you get there is less important.

Once the tank is full, the next priority is blood pressure optimization with a goal mean arterial pressure (MAP) of 65 mm Hg.  Patients with a CVP 8-12 mmHg who are still hypotensive are in need of vasopressor therapy, which is a stark contrast to persistently hypovolemic trauma patients.  Dr. Rivers did not recommend a particular vasopressor in his initial publication, nor did the subsequent Surviving Sepsis Campaign.  However, the recommended vasopressor at this point is norepinephrine (“levophed”) rather than dopamine.  I believe my own “vasopressors for dummies” simplifies things.  Norepinephrine has a strong alpha receptor effect relative to beta receptor, so it increases blood pressure by increasing peripheral vascular resistance.  It has very little effect on the heart rate, which is preferable because most septic patients are already tachycardic, and increases in heart rate are associated with a significantly higher metabolic consumption.  Dopamine on the other hand has a stronger beta receptor effect, so it increases heart rate and blood pressure.  For the previously mentioned reasons, dopamine is less desirable unless the patient is uncharacteristically bradycardic.

So to recap, thus far we have optimized intravascular volume utilizing CVP monitoring, and we have helped the patient maintain a MAP of 65 mm Hg using norepinephrine.  The next step is monitoring of SvO2, which can be done second-to-second using a specialized (and expensive) central venous catheter.  I can say anecdotally, that I have worked in 9 different emergency departments as a resident and attending physician.  I have only utilized a SvO2 detection catheter here at Henry Ford.  In the other 8 EDs, SvO2 detection wasn’t done because the technology wasn’t available or deemed too expensive, nurses weren’t trained to assist with the equipment, or the patient simply didn’t stay in the emergency department long enough to make it necessary or realistic to trend it.

The concept of SvO2 is important because it is another way of understanding the metabolic demand of the septic patient.  The main vehicle of oxygen delivery to end organs is via hemoglobin.  A typical hemoglobin molecule carries 4 oxygen molecules, and delivers 1 to vital end-organ tissue.  That hemoglobin then returns to the lungs via venous return and gets that missing oxygen replaced.  A patient in septic shock has a dramatically increased metabolic consumption due to anaerobic production of ATP, and as a result each hemoglobin molecule is forced to give up additional oxygen molecules to the end-organ tissue.  This is the very essence of monitoring SvO2.  An ideal SvO2 is 70%, meaning essentially the hemoglobin is giving away 1 oxygen molecule and retaining 3.  So if you have already optimized a patient’s fluid volume and MAP, and the SvO2 is still less than 70%, the next step in EGDT is blood transfusion.  These patients need not be anemic to warrant transfusion.  Instead, the concept behind transfusion of these patients is supplementing the number of vehicles available in the blood to deliver oxygen to end organs.

Finally, what do you do if you have optimized CVP, have a perfect MAP, have transfused blood, and yet the SvO2 is still less than 70%? The final step in optimization is administration of an inotropic agent.  Dobutamine is the medication of choice in this instance.  Dobutamine has a strong beta-1 receptor effect, which increases cardiac output, thereby delivering more hemoglobin per contraction to the end organs.  This does come at significant metabolic cost, as each beat consumes the very oxygen you are trying to replenish.  But drastic times call for drastic measures, and patients who have reached this point in the pathway are without a doubt the sickest of the sick.  I personally have only needed to use dobutamine once in all the septic patients I have resuscitated in my career.

Is this all making sense so far?  Good!  Come back and check out Part III: Diabetic Ketoacidosis!

8 thoughts on “Nuances in Resuscitation Part II: EGDT In Sepsis”

  1. There is now available a peripherally inserted central catheter, that has CVP capability, manufactured by a Chicago-based company. This line is placed in the antecubital fossa with a 5 French introducer kit or through an already placed large bore IV in the same area that can dilated up over a wire. It is brand new to the market and the first one was placed at a hospital in Monroe, MI. It was reported that they were able to place the line, obtain the VBG w/ lactate, provide the 30cc/kg fluid bolus and measure the CVP in 10-15 minutes. Very interesting possibilities.

  2. Nice review of Manny Rivers protocol although, as you state, hotly debated. I think most shops have moved towards (if they weren’t already there) the ProCESS trial style resusc. Early identification is clearly key and aggressive resuscitation. The truth is we don’t know what markers to use to gauge the adequacy of our resusc. There is likely a role for all of the non-invasive measures being used in concert (lactate clearance, SCVO2, passive leg raise, US of IVC etc).

    I disagree on the choice of fluids for sure. I would use LR or a balanced electrolyte solution (BES) if available instead of NS. Although no good randomized prospective data, the available literature seems to show an association with chloride liberal resusc and AKI (http://jama.jamanetwork.com/article.aspx?articleid=1383234) and mortality in sepsis (http://www.ncbi.nlm.nih.gov/pubmed/24674927).

    Thanks for a great review!

  3. Nice review of Manny Rivers protocol although, as you state, hotly debated. I think most shops have moved towards (if they weren’t already there) the ProCESS trial style resusc. Early identification is clearly key and aggressive resuscitation. The truth is we don’t know what markers to use to gauge the adequacy of our resusc. There is likely a role for all of the non-invasive measures being used in concert (lactate clearance, SCVO2, passive leg raise, US of IVC etc).

    I disagree on the choice of fluids for sure. I would use LR or a balanced electrolyte solution (BES) if available instead of NS. Although no good randomized prospective data, the available literature seems to show an association with chloride liberal resusc and AKI (http://jama.jamanetwork.com/article.aspx?articleid=1383234) and mortality in sepsis (http://www.ncbi.nlm.nih.gov/pubmed/24674927).

    Thanks for a great review!

  4. Also interesting is the prospect of measuring regional staurations via a product like INVOS, we use it in pediatric cardio thoracic surgery patients to measure end organ perfusion, and it can show real time results with how we are doing with our interventions.

Leave a Reply

Your email address will not be published. Required fields are marked *