Tag Archives: EGDT


EGDT sepsis bundle challenged

An article in American Journal of Emergency Medicine by two intensivists challenges the science behind Rivers’ early goal-directed therapy (EGDT) protocol for severe sepsis / septic shock. In a nutshell:

  • Rivers’ study was small (n = 263), nonblinded, industry-supported and single-center
  • early fluids and antibiotics are a sound idea, but other components of EGDT are flawed
  • targeting a CVP is meaningless and could result in hypovolaemia or pulmonary oedema; dynamic markers of preload responsiveness such as pulse pressure variation or IVC diameter variation are better guides to fluid resuscitation
  • ScvO2 may be normal or elevated in septic shock patients; the low average ScvO2 in Rivers’ study has not been reproduced in subsequent studies.
  • packed cells have significant side effects and their non-deformability, pro-inflammatory and pro-thrombotic effects may impair microvascular perfusion and paradoxically worsen tissue oxygen delivery
  • dobutamine can potentially further worsen the haemodynamic status of patients with hypovolaemia, vasodilation, or a hyperdynamic circulation, which cannot be differentiated using CVP and ScvO2

Early goal-directed therapy: on terminal life support?
Am J Emerg Med. 2010 Feb;28(2):243-5

I like this paper, mainly because I have been uncomfortable with the chasing of arbitrary targets for some time. My own practice is to try to improve markers of organ hypoperfusion (such as lactate, urine output, mental status, and skin perfusion as well as blood pressure) by early antibiotics, fluid resuscitation guided by clinical and sonographic (IVC) signs, and vasoactive drugs guided by clinical and sonographic (basic echo) findings. I place a central venous catheter for access for the vasoactive drugs, rather than to get a CVP reading. I do measure ScvO2 with a central venous blood gas, but have rarely seen one below 70% even in severely shocked patients – I’m far more interested in clearing the lactate, as are these guys.

ScvO2 in sepsis: high is bad too

ScvO2 values are obtained by measuring the oxygen saturation in venous blood returning to the heart, and reflect the balance between oxygen delivery and oxygen consumption.

Low (<70%) ScvO2 values were targeted by Rivers in his Early Goal Directed Therapy study: by improving the macrocirculation with fluids, vasoactive drugs, and packed red cells the aim is to improve oxygen delivery to tissues, and therefore a higher oxygen saturation is found in the venous blood returning to the heart in adequately resuscitated patients. The story is more complex, however, as mechanisms of oxygen supply (macrocirculatory flow), distribution (microcirculatory flow), and processing (mitochondrial function) must all function at an adequate level to maintain normal physiology.

Although low ScvO2 values may be a marker for macrocirculatory failure, high ScvO2 values may reflect microcirculatory or mitochondrial failure.

A multicentre study demonstrated a higher mortality on patients whose ScvO2 in the ED was high (90-100%) compared with those with a normal ScvO2.
Mortality associated with three groups according to their highest recorded ScvO2 in the ED was:

Hypoxia group (ScvO2 <70%) – 40% mortality (95% CI 29-53)
Normoxia group (ScvO2 71-89%) – 21% mortality (95% CI 17-25)
Hyperoxia group (ScvO2 90-100%) – 34% mortality (95% CI 25-44)

The study design could not control for many potential confounders, but this opens the door for further study, and reminds us that the unthinking pursuit of a single physiological target may miss the bigger clinical picture.

Multicenter Study of Central Venous Oxygen Saturation (ScvO2) as a Predictor of Mortality in Patients With Sepsis
Annals of Emergency Medicine 2010;55(1):40-46