Comparison of the Sepsis-2 and Sepsis-3 Definitions of Sepsis and Their Ability to Predict Mortality in a Prospective Intensive Care Unit Cohort. (4th October 2017)

• Record Type: Journal Article
• Authors: Poutsiaka, Debra D; Porto, Maura; Perry, Whitney; Hudcova, Jana; Tybor, David; Hadley, Susan; Doron, Shira; Reich, John Adam; Snydman, David; Nasraway, Stanley
• Is Part Of: Open forum infectious diseases. Volume 4(2017)Supplement 1
• Page Start: S602
• Page End: S602
• Publication Date: 2017-10-04
• Subjects: Communicable diseases -- Periodicals; Medical microbiology -- Periodicals; Infection -- Periodicals; 616.9
• DOI: 10.1093/ofid/ofx163.1579
• Languages: English
• ISSNs: 2328-8957

Abstract

Abstract: Background: We compared Sepsis-2 (S2) and Sepsis-3 (S3) in classifying patients with sepsis and mortality risk. Methods: The cohort was assembled for a prospective study of a sepsis biomarker. Entry criteria included ≥ 2 of 4 Systemic Inflammatory Response Syndrome (SIRS) clinical criteria, age ≥ 21 y, and informed consent for sample collection. Classification of sepsis or prediction of death by 30 d by S2 (≥ 2 SIRS criteria + infection [I]), S3 (prescreening by quick Sequential Organ Failure Assessment [qSOFA] of ≥ 2 of 3 clinical criteria followed by the complete SOFA [requiring clinical and laboratory assessments] score change ≥2 + I) and an amended S3 definition, iqSOFA (qSOFA ≥2 + I) were compared. All statistics were performed with SAS 9.4 (Cary NC). Results: Of 176 patients with SIRS, 95% were cared for in the ICU. 80 of 105 patients classified by S2 as having sepsis also had sepsis per S3 or iqSOFA but 25 did not (Kappa=0.72, 95% CI 0.62, 0.82). Sepsis classifications by S3 and iqSOFA were identical. Mortality data were available for 169 patients. 25 (14.8%) of the total cohort died, (20 of 100 with sepsis per S2 [20%], and 20 of 77 [26.0%] with sepsis per S3 or iqSOFA). AUC (95% CI) of ROC curves for identifying those who died were 0.54 (0.41, 0.68) for S2, 0.84 (0.74, 0.93) for S3 and 0.69 (0.60, 0.79) for iqSOFA. AUC for S3 was significantly greater than for the other two definitions ( P < 0.01, Mann–Whitney Test). S3 was more specific than S2 in distinguishing those who died (Table). iqSOFA performed identically to S3 in test performance. Unadjusted hazard ratios for death associated with sepsis are shown in the Figure and were not changed after adjusting for age or gender. Conclusion: S3 and iqSOFA were more stringent definitions of sepsis and better at predicting death than S2. The S3 ROC AUC was higher than for iqSOFA which otherwise performed identically to S3. The SOFA score used in the S3 definition required more data collection compared with the iqSOFA score. Thus, the complex S3 definition adds little advantage to the simpler iqSOFA in classifying patients with sepsis or forecasting mortality. Disclosures: D. D. Poutsiaka, Mochida Pharmaceuticals: Investigator, Research support; J. Hudcova, Mochida Pharmaceuticals: Investigator, Research support; S. Nasraway, Mochida Pharmaceuticals: Investigator, Research support