This was a retrospective cohort study of nontrauma adult patients who had experienced in-hospital cardiac arrest during emergency department stays at an urban, 2600-bed tertiary medical center in Taiwan from February 2011 to July 2013. Data regarding patients' characteristics, Charlson Comorbidity Score, MEWS score before events, mode of arrest, and outcome details were extracted following the Utstein guidelines for uniform reporting of cardiac arrest.
During the 30-month period, 234 patients suffered in-hospital cardiac arrest during emergency department stays, and 99 patients with periarrest MEWS were included in the final analysis. The MEWS at triage did not differ significantly between survival-to-discharge and mortality groups (3.42 ± 2.2 vs. 4.02 ± 2.65, p = 0.811). Periarrest MEWS was lower in the survival-to-discharge group (4.41 ± 2.28 vs. 5.82 ± 2.84, p = 0.053). In multivariate logistic regression analysis, periarrest MEWS was an independent predictors for survival to discharge. A rise in periarrest MEWS reduced the chance of survival to discharge by 0.77-fold (95% confidence interval: 0.60–0.97, p = 0.028).
The simplest MEWS system not only can be used as a prevention measure, but the periarrest MEWS could also be considered as an independent predictor of mortality after in-hospital cardiac arrest.