How early do antibiotics have to be to impact mortality in severe sepsis? A prospective, observational study from an emergency department

BACKGROUND: The objective of this study was to assess the promptness of antibiotic administration to patients presenting with sepsis and the effects on survival and length of hospitalization. METHODS: Consecutive, adult patients presenting with Systemic Inflammatory Response Syndrome (SIRS) to the emergency department of the Aga Khan University hospital were enrolled in a prospective, observational study over a period of 4 months. Univariate, multivariate regression modeling and one-way ANOVA were used to examine the effects of various variables on survival and for significant differences between timing of antibiotic administration and survival, two-sided p values \u3c 0.05 were considered significant. RESULTS: One hundred and eleven patients were enrolled. Severe sepsis was present in 52% patients; the most frequent organism isolated was Salmonella typhi (18%). Overall mortality was 35.1%. One hundred (90.1%) patients received intravenous antibiotics in the Emergency room; average time from triage to actual administration was 2.48 +/- 1.86 hours. The timing of antibiotic administration was significantly associated with survival (F statistic 2.17, p = 0.003). Using a Cox Regression model, we were able to demonstrate that survival dropped acutely with every hourly delay in antibiotic administration. On multivariate analysis, use of vasopressors (adjusted OR 23.89, 95% CI 2.16,263, p = 0.01) and Escherichia coli sepsis (adjusted OR 6.22, 95% CI 1.21,32, p = 0.03) were adversely related with mortality. CONCLUSIONS: We demonstrated that in the population presenting to our emergency room, each hourly delay in antibiotic administration was associated with an increase in mortality