1068

Abstract

Introduction: To examine the independent predictive value of reaching a central venous oxygen saturation (ScvO2) of 70% or greater among emergency department (ED) patients with severe sepsis or septic shock treated with early goal-directed therapy (EGDT). Methods: Retrospective analysis of a prospective database of 2894 patients with severe sepsis or septic shock treated with EGDT in 21 EDs between March 2010 and September 2012. Patients were included if they qualified for ScvO2-guided therapies as defined by two consecutive ScvO2 measurements less than 70% at least 15 minutes apart with a concomitant CVP of at least 8 mmHg and a MAP of at least 65 mmHg. Variables abstracted during the first six hours of EGDT included age, gender, initial ScvO2, final ScvO2, highest CVP, initial serum lactate, percent serum lactate clearance, mechanical ventilation, shock index, vasopressor or dobutamine administration, red blood cell (RBC) transfusion, lowest hemoglobin, 6-hour Sequential Organ Failure Assessment (SOFA) score and a modified electronic Simplified Acute Physiology Score III (eSAPS 3). Multivariable logistic regression was used to assess the predictive value of each variable on the outcome of in-hospital mortality. Results: 580 (20%) patients were included. In-hospital mortality was 21%. Patients with in-hospital mortality were older(p-0.001), more likely to require mechanical ventilation (p<0.001) or vasopressors (p=0.02), had higher initial serum lactates (p<0.001), 6-hour SOFA (P<0.001) and eSAPS 3 scores (p<0.001), and had lower percentage serum lactate clearances (p=0.002). On logistic regression analysis (c statistic = 0.75), variables associated with the outcome were 6-hour SOFA (OR = 1.14 [1.03-1.26], p=0.014), eSAPS 3 (OR = 1.06 [1.02-1.10], p=0.007), initial lactate (OR = 1.15 [1.03-1.27], p=0.005) and lactate clearance less than 50% (OR = 1.90 [1.03-3.51], p=0.04). Dobutamine administration, RBC transfusion, initial ScvO2 or final ScvO2 were not statistically associated with the outcome before or after adjustment for covariates. Conclusions: When CVP and MAP goals were reached, successful resuscitation to an ScvO2 of 70% or greater was not an independent predictor of in-hospital mortality in this adjusted retrospective analysis. These findings suggest that lactate clearance may be an alternate resuscitation endpoint in this cohort.