Model Can Predict Sepsis Risk for Emergency Medical Admissions
Computer-aided National Early Warning Score model makes use of data already collected by clinicians
MONDAY, April 8, 2019 (HealthDay News) -- A computer-aided National Early Warning Score (cNEWS) model accurately predicts sepsis for emergency medical admissions, according to a study published online April 8 in CMAJ, the journal of the Canadian Medical Association.
Muhammad Faisal, Ph.D., from the University of Bradford in the United Kingdom, and colleagues compared three cNEWS models for predicting sepsis risk (M0: NEWS alone; M1: NEWS + age + sex; M2: M1 + subcomponents of NEWS + diastolic blood pressure). All emergency medical admissions of patients aged 16 years and older were included from two acute care hospital centers (York Hospital [YH] for model development; combined data set from two National Health [NH] Service hospitals for external model validation).
The researchers found that the C statistic for predicting sepsis increased across models (YH: M0, 0.705; M1, 0.763; M2, 0.777; NH: M0, 0.708; M1, 0.777; M2, 0.791). For NEWS of 5 or greater, there were increases from M0 to M2, respectively, in sensitivity (YH: 47.24 versus 50.56 versus 52.69 percent; NH: 37.91 versus 43.35 versus 48.07 percent), positive likelihood ratio (YH: 2.77 versus 2.99 versus 3.06; NH: 3.18 versus 3.32 versus 3.45), and positive predictive value (YH: 11.44 versus 12.24 versus 12.49 percent; NH: 22.75 versus 23.55 versus 24.21 percent).
"These risk scores should support, rather than replace, clinical judgment," a coauthor said in a statement. "We hope they will heighten awareness of sepsis with additional information on this serious condition."