Hong Kong Journal of Emergency Medicine (Oct 2024)

Prediction of clinical deterioration risk at 8 hours after arrival in emergency department in non‐traumatic patients using trends in modified early warning score level

  • Wen‐Chen Lin,
  • Chin‐Fu Chang,
  • Yan‐Ren Lin,
  • Chih‐Wen Twu,
  • Mei‐Chu Chen,
  • Yu‐Pin Ku,
  • Kang‐Ping Lin,
  • Ching‐Hsiung Lin

DOI
https://doi.org/10.1002/hkj2.12041
Journal volume & issue
Vol. 31, no. 5
pp. 242 – 252

Abstract

Read online

Abstract Background Improving the quality of medical care in hospitals is a major priority for all departments. The early warning score (EWS) trend is an effective early risk stratification tool that reflects the changes in patient condition and allows better assessment of deterioration risk. Objective The aim of this study was to investigate whether utilizing the trend of the modified early warning score (MEWS) level within 4 h of a patient's arrival in the emergency department (ED) could identify patients at risk of clinical deterioration at 8 h after arrival in the ED. Methods We conducted a retrospective observational study of non‐trauma patients who had at least two vital sign measurements (Glasgow Coma Scale score, heart rate, blood pressure, respiratory rate, and body temperature) within 8 h of arriving in the ED. The primary outcome was patients who had MEWS ≥ 4 at 8 h after arrival in the ED. We performed multivariate logistic regression analysis using age, sex, MEWS level at arrival in the ED, MEWS level within 4 h after arrival in the ED, and MEWS level trend over time. Results Among the 5825 patients, 680 (11.7%) were at risk of deterioration at 8 h after arrival in the ED. To predict the risk of deteriorating conditions (MEWS ≥ 4), utilizing the MEWS level trend within 4 h of arrival in the ED was more effective in identifying patients at risk of deterioration after 8 h of arrival in the ED compared to using a single MEWS value during the ED stay. The corresponding areas under the receiver operating characteristic curve were 0.756 (95% confidence interval (CI) 0.734–0.778) and 0.846 (95% CI 0.827–0.865), respectively (p < 0.01). Conclusions The proposed trend‐based predictive model for MEWS levels can alert healthcare personnel regarding patients at increased risk of deterioration (MEWS ≥ 4), potentially reducing mortality rates during ED stays.

Keywords