MEDS Score for Sepsis Might Best Predict ED Mortality
Clinical question: Which illness severity score best predicts outcomes in emergency department (ED) patients presenting with infection?
Background: Several scoring models have been developed to predict illness severity and mortality in patients with infection. Some scores were developed specifically for patients with sepsis and others for patients in a general critical care setting. These different scoring models have not been specifically compared and validated in the ED setting in patients with infection of various severities.
Study design: Prospective, observational study.
Setting: Adult ED in a metropolitan tertiary, university-affiliated hospital.
Synopsis: Investigators prospectively identified 8,871 adult inpatients with infection from a single-center ED. Data to calculate five prediction models were collected. The models were:
- Mortality in Emergency Department Sepsis (MEDS) score;
- Acute Physiology and Chronic Health Evaluation II (APACHE II);
- Simplified Acute Physiology Score II (SAPS II);
- Sequential Organ Failure Assessment (SOFA); and
- Severe Sepsis Score (SSS).
Severity score performance was assessed for the overall cohort and for subgroups, including infection without systemic inflammatory response syndrome, sepsis, severe sepsis, and septic shock. The MEDS score best predicted mortality in the cohort, with an area under the receiver operating characteristics curve of 0.92. However, older scoring models such as the APACHE II and SAPS II still discriminated well, especially in patients who were admitted to the ICU. All scores tended to overestimate mortality.
Bottom line: The MEDS score may best predict illness severity in septic patients presenting to the ED, but other scoring models may be better-suited for specific patient populations.
Citation: Williams JM, Greenslade JH, Chu K, Brown AF, Lipman J. Severity scores in emergency department patients with presumed infection: a prospective validation study. Crit Care Med. 2016;44(3):539-547.
