Mortality from Severe Sepsis Decreasing in the U.S.

Thomas Glück, MD

Clinical trial findings and administrative data suggest that severe sepsis mortality decreased from 47% in 1991–1995 to 29% in 2006–2009 — a 3% annual reduction.

Epidemiologic investigations in the U.S. have suggested increasing incidence of severe sepsis and septic shock — conditions with a considerable risk for mortality. Is the death toll from these conditions also rising? To shed light on this issue, researchers examined discharge data for 1993 through 2009 from the Nationwide Inpatient Sample, the Healthcare Cost and Utilization Project, and the Agency for Healthcare Research and Quality, and findings from multicenter clinical sepsis trials (many of them multinational) with enrollment between 1991 and 2009.

Data from 36 sepsis trials involving 14,418 patients with severe sepsis or septic shock who did not receive investigational treatments showed a significant, 3% annual decline in 28-day mortality — from 47% in 1991–1995 to 29% in 2006–2009, although the mortality rate predicted using baseline severity-of-disease scores did not change. The decline in the observed mortality rate was similar among trials, whether the trials included patients in the U.S. or only enrolled patients in other countries.

The administrative data revealed a similar trend in severe sepsis mortality: 3.7% and 4.5% declines, using the definitions of Martin and Angus, respectively (NEJM JW Infect Dis Mar 27 2013). Although mortality rates identified using the Martin definition closely matched the clinical trial data, those based on the Angus definition were approximately 10% lower.