Which Score to Use When Screening for Sepsis on the Wards

Aaron B. Holley, MD


April 21, 2017

Quick Sepsis-Related Organ Failure Assessment, Systemic Inflammatory Response Syndrome, and Early Warning Scores for Detecting Clinical Deterioration in Infected Patients Outside the Intensive Care Unit

Churpek MM, Snyder A, Han X, et al
Am J Respir Crit Care Med. 2017;195:906-911


The Sepsis-3 consensus statement recommends replacing the systemic inflammatory response syndrome (SIRS) criteria with the quick Sequential Organ Failure Assessment (qSOFA).[1,2] Several editorials published in response raised concern that sacrificing the sensitivity obtained with SIRS for increased specificity would lead to delayed identification.[3,4] A review published later in the year noted that studies cited in Sepsis-3 enrolled emergency department and intensive care unit (ICU) patients, but not ward patients.[5] The academic debate highlights how little we know about improving clinical outcomes for hospitalized patients who develop sepsis outside of the ICU.

The Study

A new study by Churpek and colleagues (the authors of the 2016 review[5]) supplies data to drive this debate. They sampled retrospective data from the University of Chicago over an 8-year period. SIRS, qSOFA, Modified Early Warning Score (MEWS),[6] and National Early Warning Score (NEWS)[7] were calculated for patients in the emergency department and on the wards. Intravenous antibiotic administration and culture orders were used to identify suspicion of infection. The primary outcome was hospital mortality, and the secondary outcome was a composite of death or ICU admission any time after suspicion of infection.

The researchers found the NEWS and MEWS had the best area under the receiver operating characteristic (AUROC) curve for predicting outcomes. qSOFA was next, and SIRS was last. A SIRS score ≥ 2 had a sensitivity of 91% and a specificity of 13%. Most of the patients met these criteria. Patients had ≥ 2 SIRS, ≥ 2 and > 1 qSOFA criteria 17, 5, and 17 hours before the composite, secondary outcome, respectively.


As noted in the accompanying editorial,[8] early identification of sepsis is governed by basic mathematical truths. If you want early and sensitive, you must sacrifice specificity and overall accuracy. If you want a higher AUROC, you'll need a more complicated model (the MEWS and NEWS scores contain five and six variables, respectively, and run from 1 to 9 and 1 to 15).

The University of Chicago data are hugely helpful because they allow hospitals and health systems to quantify trade-offs. Decisions regarding which score to use should be governed by resources and patient population. Hospitals with advanced electronic health records looking to optimize finite ICU capacity may opt for the MEWS or NEWS. Those with excess capacity or a lower prevalence of sepsis might want to use SIRS. There's no such thing as a free lunch, but these data help explain what we get for our money.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.