Email updates

Keep up to date with the latest news and content from Critical Care and BioMed Central.

This article is part of the supplement: 26th International Symposium on Intensive Care and Emergency Medicine

Open Badges Poster presentation

A novel biomarker panel with a Multimarker Index value for the diagnosis of sepsis in the Emergency Department

E Rivers1, N Shapiro2, JE Hollander3, R Birkhahn4, R Otero1, T Osborn5, D Milzman6, E Moretti7, B Nguyen8, S Trzeciak9, K Gunnerson10 and the Prospective Observational Evaluation of Sepsis (POEMS) Group

Author Affiliations

1 Henry Ford Hospital, Detroit, MI, USA

2 Beth Israel Deaconess Medical Center, Boston, MA, USA

3 University of Pennsylvania, Philadelphia, PA, USA

4 New York Methodist Hospital, Brooklyn, NY, USA

5 University of Virginia, Charlottesville, VA, USA

6 Medstar Research Institute, Washington, USA

7 Duke University Medical Center, Durham, NC, USA

8 Loma Linda University Medical Center, Loma Linda, CA, USA

9 Cooper Health System, Camden, NJ, USA

10 Virginia Commonwealth University, Richmond, VA, USA

For all author emails, please log on.

Critical Care 2006, 10(Suppl 1):P82  doi:10.1186/cc4429

The electronic version of this article is the complete one and can be found online at:

Published:21 March 2006

© 2006 BioMed Central Ltd


To define a panel of biomarkers to aid in the assessment and diagnosis of patients presenting to the Emergency Department (ED) with suspected sepsis.


Patients presenting to the ED of 10 tertiary hospitals. The inclusion criteria were: 18 years or older, and two or more SIRS criteria with confirmed or suspected infection and/or lactate > 2.5 mmol/l. The exclusion criteria were: cardiac arrest; Do Not Resuscitate order; pregnancy.


Blood samples were collected upon enrollment and the plasma was frozen at -70°C within 1 hour. Measurement of biomarkers was performed in a blinded fashion by immunoassay (Biosite). The final diagnosis was determined based on standard definitions. A search engine based on the optimization of the area under the receiver-operator curve (ROC AUC) for diagnosis of sepsis was used to select a panel of six biomarkers and to optimize a Multimarker Index (MMX). The MMX combines the individual marker values into a single index value for the sample, which is reported as the test result.


The ROC AUC (95% CI) for the sepsis panel MMX for differentiating patients with SIRS (n = 73) from those with various sepsis conditions (n = 224), SIRS from all sepsis with positive blood culture (n = 41), SIRS from severe sepsis (n = 43) and SIRS from septic shock (n = 14) are, respectively, 0.76 (0.70–0.82), 0.85 (0.79–0.92), 0.89 (0.83–0.95) and 0.95 (0.91–0.99). For comparison, the ROC AUCs obtained with CRP are 0.62 (0.54–0.70), 0.69 (0.59–0.79), 0.63 (0.53–0.74) and 0.71 (0.57–0.85); for IL-6 the values are 0.64 (0.57–0.71), 0.69 (0.59–0.80), 0.66 (0.55–0.77) and 0.63 (0.45–0.82).


The biomarker panel MMX shows clinical utility in identifying sepsis as an underlying cause in patients presenting with SIRS. Because this level of accuracy is attained based on the first blood draw, these results suggest that it may be useful in the rapid assessment and diagnosis of patients presenting to the ED. These results need to be confirmed in additional populations.