Saturday, November 9, 2013

If Early Goal Directed Therapy Had a Bar Mitzvah...

Manage Sepsis like a Baawwss! "Wwughh!"

It's been nearly 13 years since the publication of the landmark Rivers early goal directed therapy study in the NEJM. As with a Bar Mitzvah, the Jewish rite of passage from youth to manhood, it's now time for EGDT's transformation and to face post-pubescent changes that go hand-in-hand with maturity, development, and the right to start drinking Maneschwitz. The developments I allude to here are reflections of the various (improved, more accurate) modalities many of us now use to detect the different goal posts set out in Rivers' algorithmic approach to severe sepsis. Let's explore this growth together.

First, consider a case scenario:

A 66 year old female (history of diabetes, hypertension, CHF) from home presents with fever and productive cough with the vitals: BP 88/50, HR 120, RR 20, T 39 C. She is confused and has dark urine.

You place her on a monitor, insert a foley catheter, add supplemental oxygen via nasal cannula, and place an IV to send labs and administer an initial bolus of 20-30 cc/kg of crystalloid fluid. Antibiotics are running in the IV. 15 -20 minutes later, she is still hypotensive. Do you give more fluid? Or do you now place a central line and add pressors (like norepinephrine)? 

In the original Rivers model, once the presence of severe sepsis/septic shock is recognized and the decision is made to  proceed with EGDT to guide us in reaching the ultimate point in a severely septic patient--normalize tissue perfusion--we will first and foremost satisfy the most important goal of "filling up the tank" completely and adequately with IV fluid. Only then can we begin moving further to subsequent potential deficiencies/goals that demand addressing: "tightening the pipes" (MAP > 65 with pressors), "optimizing the pump" (optimize cardiac output with inotropy), and "adding passengers" (transfusion of blood product to increase oxygen carrying capacity). With each subsequent step, you keep going back to check and see if you've met your ultimate goal of improved tissue perfusion/oxygenation by looking to surrogate marker of perfusion: ScvO2 (Rivers) or, in the 2008 Jones non-invasive model, serum lactate clearance.  

Here's the algorithm:

For a more detailed explanation of the above analogy on tank/pipes/pump/and passengers, see the previous blog post on water slide physiology

We kosher so far? Good. Let's move along.

The meat of this discussion and important question to ask yourself next is "have I really given enough fluid? Was the initial crystalloid bolus adequate for volume repletion or does the patient still require more to meet the primary goal of filling up the tank?" Rivers' answered this dilemma by dropping a central line in all his patients and measuring the CVP: if it was less than 8-12 mmHg, he gave more fluid until the goal pressure (8-12) was achieved. If CVP was eventually optimized to 8-12 mmHg ('an adequate preload') and the patient was still hypotensive, he'd move down the chain and begin pressors to address the MAP ('optimize afterload'). (See above figure). The assumption here of course is that CVP is a good marker of a patient's fluid status--a fuel tank gauge on the physiologic dashboard, if you will. 

Today we know that isn't true. There are two huge shortcomings to this approach:
1. CVP does NOT in fact correlate well with volume status, if at all! Reference article by Paul Marik
2. Measuring CVP requires insertion of a central venous catheter, which is invasive, carrying risks and complications

With the risk of giving too little or too much fluid and increasing the morbidity/mortality, we welcome a better, more accurate evolutionary approach to detecting a patient's volume status at the bedside...

Ultrasound! What else??
There is now sufficient data to support the use of ultrasound-based measurement of IVC diameter fluctuation during the respiratory cycle  in order to gauge volume status. (EMCrit reviews them here). The background concept is quite simple:

-In a volume depleted spontaneously breathing patient: each breath lowers intrathoracic pressure, which in turn increases cardiac return (like a suction pump). As the already low reserve of intravascular volume is shifted up into the chest with each inspiration, the highly-compliant IVC will collapse. 
(ignore the bit about a sniff test)
Balloon is the IVC, collapsed with sucked-out volume
-In a volume depleted mechanically ventilated patient: each administered breath from the vent increases intrathoracic pressure, which in turn lowers cardiac return and pushes blood back down into the abdomen (like a piston). The depleted IVC, now receiving an inspiratory surge of volume, will noticeably distend. 
Man is the ventilator; balloon is IVC
Most of the papers validating the use of IVC ultrasound are based on mechanically ventilated patients, since tidal volume is set and each breath is controlled, The literature is not as clear for spontaneously breathing patients due to an inherent variability with each breath taken. These are the patients we care for more frequently though and for obvious reasons we prefer to use a less invasive approach to fluid monitoring when possible.

That's ok. 

There is good news however; research does agree with good consistency and relatively high specificity that on the lowest extreme of the fluid status spectrum--volume depletion--grossly (easily) visible IVC collapse (of greater than 40%) on B-mode sonography is suggestive of a fluid depleted state where patients will benefit from further volume resuscitation. As Weingart puts it: "if you see it collapse give more fluid, if you see it collapse give more fluid, if you see it collapse give more fluid!" 

So true is the above notion regarding IVC that its diameter measurement is now tantamount to assessing a patient's volume status in the 2012 Surviving Sepsis Campaign Guidelines

So move over CVP! Make way please. 

At this point we're beginning to run long, so I'm going to take a much-needed pause here. We will resume our discussion in part II, where I'll be addressing the pressing concern of how to determine if your patient requires more fluid therapy if or when the IVC does NOT show obvious fluctuation with respirations  (>40% collapse in spontaneously breathing and <18% dissension in mechanically ventilated patients). This is the scenario, unfortunately, that we are more likely to encounter--"the gray zone" as it were--while working up septic patients--especially if checking the IVC after an initial bolus. In other words, just because it's not collapsing (distending in MV patients) does it imply they no longer need IVF, that the tank is completely full? Be sure to check back in soon for the explanations and illustrations. 



  1. I love reading your story CPR CLASS . I'm sorry things we to hell when everything had been progressing great. Thank goodness I know things ultimately end well.

  2. magnificent post, very informative. I wonder why the opposite specialists of this
    sector don't realize this. You should continue your writing.
    I'm confident, you've a huge readers' base already!

    Also visit my homepage; Car Crashes

  3. I am al so agree with all above comments really appreciation work here...

    Dr Najeeb Lectures