Friday, December 10, 2010

Does early goal directed therapy matter? A point-counterpoint

In response to a study published in Chest showing a disturbing low rate of adherence to early goal directed therapy (EGDT) at a large medical center, editorials expressing opposing views were presented in the same issue.


First Dr. Emanuel Rivers, principal investigator in the landmark EGDT study published almost a decade ago, defended the protocol. He cited the very strong basic science rationale that had been developed long before the EGDT trial. He also noted, as I have many times before in this blog, the many studies over the ensuing years after the trial that replicated its findings.


The counterpoint editorial was written by Dr. Gary Schmidt from the pulmonary-critical care section at the University of Iowa. His was largely and argumentum ad populum: adherence to EGDT is low; this must reflect informed skepticism; therefore EGDT is not validated. He did point out evidence that lactate guided therapy may be a surrogate for serial measurement of ScvO2. At least one study, however, found that lactate levels were complementary to the measurement of ScvO2. Finally, he noted three trials in progress which are examining various components of EGDT:


These include the Australasian Resuscitation in Sepsis Evaluation (ARISE) trial (1,600 subjects), the Protocolized Care for Early Septic Shock (ProCESS) trial centered in Pittsburgh (1,935 subjects), and the Protocolised Management of Sepsis (ProMISe) trial in the United Kingdom (1,260 subjects).


The results of multiple studies which replicate Rivers' original findings are compelling and I doubt that these three trials will negate EGDT in any way. More likely, by examining individual components of EGDT they will provide better understanding and insight, and perhaps some tweaks that will refine early hemodynamic resuscitation of septic patients.


Dr. Rivers offered a rebuttal in which he questioned the equipoise of continued trials. Dr. Schmidt countered once again.

No comments: