infection-prevention-2017-posters-abstracts

Page 64

conferenceseries

.com

Volume 5, Issue 7 (Suppl)

J Infect Dis Ther, an open access journal

ISSN: 2332-0877

Infection Prevention 2017

December 14-15, 2017

December 14-15, 2017 | Rome, Italy

World Congress on

INFECTION PREVENTION AND CONTROL

Prospective cohort study of the qSOFA score versus the SIRS criteria in the determination and

prognostication of sepsis in a Philippine tertiary hospital

Stephanie Rachel C, Onion Gerald V, Maria Fe Raymundo-Tayzon, Cybele Lara R, Karl Evans R Henson, Irmingarda P Gueco

and

Jude Erric L Cinco

The Medical City, Philippines

epsis is a leading cause of mortality both locally and worldwide. Despite this, early diagnosis of sepsis remains difficult, with

a significant number not fulfilling SIRS criteria. In 2016, the Sepsis-3 guidelines modified its definition to include qSOFA

score. To compare the two, 295 adult patients in the emergency room with suspected infection were included in the study and

simultaneously determined their qSOFA score and SIRS criteria. The presence of sepsis was adjudicated by three infectious

disease specialists, and outcomes within the first 48 hours were acquired. Sensitivity, specificity, positive predictive and negative

predictive values for qSOFA and SIRS were computed using constructed confusion matrices, and overall predictive accuracy

was measured by the AUROC. The qSOFA score was specific (95.5%) but poorly sensitive (46.3%) test compared to the SIRS

criteria (sensitivity 73.7% and specificity 60%). Both qSOFA and the SIRS criteria significantly co-related with sepsis positivity

but the qSOFA score had superior overall predictive accuracy at 70.9% compared to the SIRS criteria. The adjudicators had

moderate strength in agreement (Fleiss’ kappa=0.39) and a percentage agreement of 60%. Based on our findings, we conclude

that the qSOFA score is a more accurate predictor of sepsis, but should not be used as a preliminary sepsis screening tool.

J Infect Dis Ther 2017, 5:7(Suppl)

DOI: 10.4172/2332-0877-C1-036