1912. Implementation of Sepsis-3 Definition in the Emergency Department: Proposed of Case Detection in Real-Life Practice

BACKGROUND: Sepsis is a major public health concern. Revised definitions of sepsis in 2016 from the systemic inflammatory response syndrome (SIRS) criteria to the Sequential Organ Failure Assessment (SOFA) score made change in the sepsis case detection. One-hour bundle proposed by the Surviving Sepsis Campaign Bundle 2018 made the process more practically challenging because of its time-constraint. METHODS: We retrospectively reviewed medical records of patients aged over 15 who visited the emergency department (ED) and got admitted to the internal medicine department from January to February 2018 in Somdech Phra Pinklao Hospital. Our study excluded pregnant women and patients who died within 48 hours after the admission. Data needed to complete SIRS, SOFA and quick SOFA (qSOFA) score was collected. Patients’ diagnosis, treatments and in-hospital mortality were also reviewed. Prevalence of sepsis according to each definition was calculated. Test performances were summarized separately using sensitivity, specificity, ROC and AUC. RESULTS: We identified 217 cases, excluding one pregnancy and five patients who died within 48 hours. Prevalence of sepsis was 45.0% from SIRS ≥2, 30.3% from SOFA ≥2, and 11.8% from qSOFA ≥2. Because the high number of missing PaO(2)/FiO(2) (96/188, 51.1%), we also calculated adjusted SOFA by excluding the factor. Sensitivity of SOFA ≥2 was 0.60 (0.49–0.70), specificity was 0.94 (0.88–0.98) and AUC was 0.77 (0.72–0.82), comparing to SIRS criteria. In-hospital mortality prediction using SIRS ≥2 had sensitivity of 0.78 (0.58–0.91), specificity of 0.38 (0.30–0.48), and AUC of 0.58 (0.49–0.67), while applying SOFA score ≥2 had sensitivity of 0.67 (0.46–0.83), specificity of 0.62 (0.52–0.70), and AUC of 0.64 (0.79–0.92). Two-step approach by screening patients who had SIRS ≥2, followed by detecting who had SOFA ≥2 had sensitivity of 0.81 (0.58–0.95), specificity of 0.46 (0.34–0.58), and AUC of 0.63 (0.53–0.74). No significant difference was found between applying adjusted or completed SOFA score. By using two-step approach, about 55% decrease in number of patients needed to complete the SOFA score. CONCLUSION: Although SOFA score was a better diagnostic tool to detect sepsis than SIRS, applying the method for all patients in the ED is difficult to be practically implemented. We proposed two-step approach by using SIRS ≥2 followed by SOFA score ≥2 for sepsis case detection. DISCLOSURES: All authors: No reported disclosures.