Date of Award
Medical Doctor (MD)
Charles R. Wira III
The recently introduced Sepsis-3 guidelines have yet to be applied to the emergency department (ED) setting. To compare these new definitions clinically with those of Sepsis-2, a retrospective analysis was performed using a dual-center ED registry of patients prospectively identified as having sepsis with organ dysfunction as defined by Sepsis-2. Of 446 registry patients meeting Sepsis-2 criteria, 61.0% (n = 272) had an elevated qSOFA score. 28-day in hospital mortality for patients with an elevated qSOFA was 15.8% compared to 10.9% in patients with non-elevation (p = .162). Rates of mechanical ventilation and vasopressor infusion were 38.6% and 36.4% respectively for patients with elevated qSOFA compared to 19.5% and 13.8% for patients without qSOFA elevation (p < .001). Patients with an elevated qSOFA and a serum lactate >2 mmol/L had a mortality rate of 18.1% vs 8.8% (p = .006) in patients with non-elevated qSOFA or normal lactate. Patients with septic shock as defined by Sepsis-3 (n = 50) had a mortality rate of 34.0% compared to 29.1% in patients with septic shock as defined by Sepsis-2 (n = 79; p = .565). 85.0% (n = 379) of patients had a SOFA score of 2 or higher thus meeting the definition of sepsis, with 30.5% (n = 136) of patients having all data available for calculating a complete SOFA score. Mortality rates were 15.6% in patients with a SOFA score of 2 or greater compared to 4.5% in those without (p = .016). While the Sepsis-3 definitions may help identify patients with a greater risk for mechanical ventilation, vasopressor infusion, and death, the percentage of patients meeting Sepsis-2 criteria who do not meet Sepsis-3 may limit the prognostic utility of these new definitions. Addition of serum lactate to the qSOFA may improve prognostic value and should be investigated further in the ED phase of care.
Zirker, Christopher Anzalone, "Evaluation Of Diagnostic Criteria In The Early Identification And Treatment Of Sepsis" (2018). Yale Medicine Thesis Digital Library. 3468.