What could be wrong about a campaign to promote sepsis survival? Looking back on the history of the Surviving Sepsis Campaign (SSC), just about everything. Understand, though, this knowledge is borne with the benefit of hindsight; there are lessons to be learned for all of us.

The initial SSC recommendations in 2004 were an attempt by many individuals in critical care and other professional societies (including ACEP) to improve sepsis care on the heels of what finally appeared to be new lifesaving interventions.¹

Although introduced and received with much enthusiasm, the SSC recommendations were soon called into question in a New England Journal of Medicine article critical of pharma sponsors that made products associated with these recommendations, particularly Eli Lilly who made Xigris.² Looking back, even the least cynical would acknowledge that many of the strongly recommended approaches from the initial 2004 and 2008 guidelines were later shown to be not only ineffective or unnecessary but also dangerous.^1,3 Specifically, the following were recommended in 2004 and then later removed:

• Activated protein C (Xigris): Demonstrated ineffective, associated with an increased risk of bleeding, and withdrawn from the market (Xigris recommended against in 2012).
• Dopamine among first choices for fluid-unresponsive septic shock: Associated with more arrhythmias and a higher mortality rate than norepinephrine (norepinephrine recommended in 2012).
• Tight glucose control (<150 mg/dL): Associated with severe hypoglycemia and increased mortality compared with target glucose levels up to 180 mg/dL (goal changed to <180 mg/dL in 2012).
• Red blood cell transfusion to maintain hemoglobin >10 g/dL: Not found to be associated with improved outcomes compared with >7 g/dL (goal changed to 7–9 g/dL in 2008).
• Early goal-directed therapy (EGDT) resuscitation based on serial central venous O2 saturation assessments: Found to have no survival benefit compared with usual care (EGDT parameters not mentioned in 2016).

In retrospect, a clear theme of the original recommendations was that many were based on early, positive, but not well-validated results. For example, the EGDT recommendation was based on a single-center trial that involved 263 participants.⁴ While the Xigris trial was multicenter and involved 1,690 participants, the results had not been validated in another investigation at the time the SSC recommended it.⁵

In response to the remarkably rigorous multisite, multinational debunking of EGDT—first by the ProCESS trial and later by the ARISE and ProMISe trials, with almost 5,000 total participants—the SSC responded by pointing out the 18 percent mortality in the usual care arm “illustrates a dramatic change in the management and outcomes of patients with septic shock” compared with the higher mortality rate in the original EGDT trial.⁴ Temporal associations of SSC bundle compliance and lower mortality rates were also offered as support of SSC efforts.