View Abstract

A3298 - Relationships Among SIRS, qSOFA, and Sepsis-2 or Sepsis-3 Organ Dysfunction in Emergency Department Patients with Suspected Infection
Author Block: S. Parashar1, K. Lembke1, S. Q. Simpson2; 1Division of Pulmonary and Critical Care Medicine, University of Kansas, Kansas City, KS, United States, 2Pulmonary and Critical Care Medicine, Kansas Univ Med Ctr, Kansas City, KS, United States.
Rationale: We hypothesized that the recently proposed Sepsis-3 diagnostic criteria are not mutually exclusive with Sepsis-2 criteria. This study was performed to determine overlaps between SIRS and qSOFA in patients presenting to the emergency department and to evaluate for progression to organ dysfunction. Methods: Single center retrospective cohort analysis of patients aged ≥18 admitted through ED with suspected infection 3/2007-5/2016. Presenting SIRS and/or qSOFA was determined using vital signs within 2 hours and laboratory tests within 3 hours of triage. Suspected infection was determined by a combination of blood cultures and antibiotics within a 4-hour window. Progression to organ dysfunction 3 hours to 48 hours post-triage was determined based on Sepsis-2 and Sepsis-3 criteria. Results: 15,037 suspected infection patients were identified; 9,992 (66.5%) presented with SIRS and 3,427 (22.8%) presented with qSOFA. 3,993 patients (26.6%) developed Sepsis-2 organ dysfunction after 3 hours and 3,065 patients (20.4%) developed Sepsis-3 dysfunction. 187 patients (5.5%) presenting with qSOFA developed Sepsis-3 organ dysfunction in 3-48 hours vs. 1,830 patients (18.3%) who presented with SIRS. 694 patients (20.3%) who presented with qSOFA developed Sepsis-2 organ dysfunction, compared with 2,760 patients (27.6%) who presented with SIRS. Of 10,402 patients presenting with SIRS or qSOFA, 3,017 (29.0%) presented with both (30.2% of SIRS patients and 88.0% of qSOFA patients). ROC analysis of predictive ability for progression to either Sepsis 2 or 3 organ dysfunction yielded AROC of 0.4894 (95% CI 0.4803 - 0.4986) for SIRS and 0.4068 (95% CI 0.3984 - 0.4151) for qSOFA. Sensitivity of SIRS≥2 65.1%, specificity 32.6%; sensitivity of qSOFA ≥2 13.3%, specificity 70.8%. ROC analysis of SIRS predictive ability for mortality yielded AROC of 0.6226 (95% CI 0.6050 - 0.6403). SIRS ≥2 points had a sensitivity of 84.4% and a specificity of 34.6%. ROC of qSOFA predictive ability for mortality yielded an AROC of 0.6879 (95% CI 0.6706 - 0.7053). qSOFA ≥2 points had sensitivity of 48.9% and specificity of 78.7%. Conclusions: A majority of patients who present to ED with suspected infection have ≥2 SIRS. There is substantial overlap between qSOFA and SIRS. A greater proportion patients with SIRS develops organ dysfunction than patients with qSOFA. However, both SIRS and qSOFA show poor prognostic value for subsequent development of organ dysfunction. Predictive ability for mortality of SIRS and qSOFA is similar to previous studies.