Rivers Early Goal-Directed Therapy for Sepsis

Sepsis management in the 1990s was supportive and largely reactive. Clinicians gave fluids, started vasopressors when blood pressure fell, and treated the source. There was no widely accepted protocol that defined what resuscitation should look like in the first hours, and an analysis of 1995 hospital records from seven US states found that 28.6% of patients with severe sepsis died in hospital. Emanuel Rivers, an emergency physician at Henry Ford Hospital in Detroit, hypothesized that correcting deficits in tissue oxygen delivery in the emergency department, before organ dysfunction set in and before the patient reached the ICU, could change outcomes.

Rivers and colleagues randomized 263 patients presenting to the Henry Ford emergency department with severe sepsis or septic shock to a six-hour protocolized resuscitation bundle or to standard care. The bundle targeted specific physiologic endpoints in sequence: central venous pressure of 8 to 12 mmHg through fluid loading, mean arterial pressure of 65 mmHg or above with vasopressors, and central venous oxygen saturation of at least 70% achieved through packed red cell transfusion or dobutamine as needed. In-hospital mortality was 30.5% in the EGDT group versus 46.5% in controls. The paper appeared in the New England Journal of Medicine in November 2001.

The Surviving Sepsis Campaign's first guidelines, written in 2003 by experts from 11 international organizations and published in 2004, recommended early goal-directed resuscitation during the first six hours, and hospitals were encouraged to adopt it as part of sepsis bundles. Central venous catheters for ScvO2 monitoring became a standard part of early sepsis care in many institutions. Before the protocol, resuscitation varied from one physician to the next, and the 16-point drop in mortality in the Detroit trial was the main argument for starting aggressive treatment in the emergency department.

Three multicenter trials published between 2014 and 2015 tested EGDT against usual care in larger populations. ProCESS, coordinated from the University of Pittsburgh with Derek Angus as senior author, enrolled 1,341 patients at 31 US hospitals. ARISE enrolled 1,600 patients at 51 centers, most of them in Australia and New Zealand. ProMISe enrolled 1,260 patients across 56 NHS hospitals in England. Taken together, the three trials randomized more than 4,200 patients and none found a mortality benefit for protocolized EGDT compared with usual care, which by then had incorporated early antibiotics, fluids, and careful hemodynamic attention without the rigid ScvO2 targets.

The ProCESS investigators pointed to changes in routine critical care over the intervening decade and noted that the Detroit patients had been older, with more heart and liver disease and higher lactate levels; a later review added that the original emergency department served a low-income community and had a high control-group mortality. In 2016 the Surviving Sepsis guidelines kept early antibiotics and fluid resuscitation but dropped ScvO2 as a recommended target. A patient-level meta-analysis of the three trials, published in 2017, found 90-day mortality of 24.9% with EGDT and 25.4% with usual care among 3,723 patients at 138 hospitals.