Comparison of different scoring tools in early detection of sepsis

Abstract

<p>Background: Recently, the diagnosis of sepsis was redefined and of today there is no gold standard for diagnosing the syndrome. The increasing use of different screening tools for identifying sepsis in the Emergency Department (ED) calls for validation. Objective: To evaluate the clinical usefulness of qSOFA, SIRS, TILT and NEWS as early warning scores for sepsis and in prediction of mortality in patients with suspected infection admitted to the ED. To assess if a modification by including risk factors to the different scoring tools could improve early recognition of sepsis. <p>Methods: The study was a retrospective study performed in the ED at a single center hospital in Norway in the period October 1. 2017 – January 14. 2018. The study sample consisted of patients (n=391) who were either received by The Emergency Medical Team (EMT) or were later admitted to the Department of Infection with either a yellow, orange or red triage according to the Rapid Emergency Triage and Treatment System (RETTS). Patients were selected using data from DIPS (the hospitals electronic health record). We measured sensitivity, specificity and area under the receiver characteristic curve (AUC) for detection of sepsis and mortality as end point. <p>Results: Of 391 patients screened, 270 patients were included and 139 had sepsis. NEWS 4 was of most clinical usefulness in detection of sepsis with a sensitivity of 0.78 (95% CI: 0.71-0.84) and a specificity of 0.59 (95% CI: 0.50-0.67). qSOFA 2 had lowest sensitivity with 0.48 (95% CI: 0.40-0.56), but highest specificity with 0.95 (95% CI: 0.90-0.98). Overall mortality was 27 %. NEWS identified most patients who experienced death within 7-days, 30-days and 1-year although the ROC curve of qSOFA was higher than of NEWS in predicting mortality. All modified screening tools demonstrated an increased ability to identify sepsis. <p>Conclusions: All scoring systems were able to recognize patients with sepsis. NEWS was found to be of more clinical usefulness compared to qSOFA, SIRS and TILT in early identification of sepsis. NEWS is at least equivalent or better than the other screening tools across most measures in predicting mortality. Our finding suggests that the implementation of risk factors in different screening tools could increase their clinical usefulness.