Modified Early Warning Score and Risk of Mortality after Acute Stroke

Abstract
Objective
An accurate prediction tool may facilitate optimal management of patients with acute stroke from an early stage. We evaluated the association between admission modified early warning score (MEWS) and mortality in patients with acute stroke.

Method
Data from the Anglia Stroke Clinical Network Evaluation Study (ASCNES) were analysed. We evaluated the association between admission MEWS and four outcomes; in-patient, 7-day, 30-day and 1-year mortality. Logistic regression models were used to calculate the odds of all mortality timeframes, whereas Cox proportional hazards models were used to calculate mortality at 1 year. Five univariate and multivariate models were constructed, adjusting for confounders. Patients with a moderate (2−3) or high (≥4) scores were compared to patients with a low score (0–1).

Results
The study population consisted of 2006 patients. A total of 1196 patients had low MEWS, 666 had moderate MEWS and 144 had a high MEWS. A high MEWS was associated with increased mortality as an in-patient (OR 4.93, 95 % CI: 2.88–8.42), at 7 days (OR 7.53, 95 % CI: 4.24–13.38), at 30 days (OR 5.74, 95 % CI: 3.38–9.76) and 1-year (HR 2.52, 95 % CI 1.88–3.39). At 1 year, model 5 had a 1.02 OR (95 % CI 0.83–1.24) with moderate MEWS and 2.52 (95 % CI 1.88–3.39) with high MEWS.

Conclusion
Elevated MEWS on admission is a potential marker for acute-stroke mortality and may therefore be a useful risk prediction tool, able to guide clinicians attempting to prognosticate outcomes for patients with acute-stroke.