One Score Fits All? A Narrative Review on Early Warning Scores for Older Adults in the Emergency Department in the Era of Personalized Medicine

TL;DR

This review examines how well early warning scores work for older adults in emergency departments and suggests personalized approaches may be better.

Contribution

Highlights limitations of traditional early warning scores in elderly patients and proposes adaptive, personalized models as a novel direction.

Findings

Traditional EWSs show reduced accuracy in older and frail patients.

Uniform cut-off values fail to capture individual vulnerability in elderly patients.

Integrating age, frailty, and comorbidities improves risk stratification.

Abstract

Background: The growing use of Emergency Departments (EDs) by older adults highlights the need for early and accurate identification of clinical deterioration. Early Warning Scores (EWSs) are widely implemented tools based on standardized vital sign thresholds; however, their performance in elderly patients is inconsistent, likely reflecting the biological heterogeneity, multimorbidity, and reduced physiological reserve typical of this population. Objectives: This narrative review aims to summarize current evidence on the use of EWSs in adults aged ≥ 65 years presenting to the ED, with a specific focus on mortality and intensive care unit (ICU) admission, and to discuss their role within the evolving framework of personalized medicine. Sources: A narrative review of 36 clinical studies published between 2014 and 2025 was conducted. Content: Traditional scores such as National Early Warning Score (NEWS), National Early Warning Score 2 (NEWS2), Modified Early Warning Score (MEWS), VitalPAC Early Warning Score (ViEWS), Rapid Acute Physiology Score (RAPS) and Rapid Emergency Medicine Score (REMS) show variable and often reduced prognostic accuracy in older and frail patients. Evidence consistently suggests that applying uniform cut-off values fails to capture individual vulnerability in elderly patients. The integration of age, frailty, comorbidities, and baseline physiological status improves risk stratification. Second-generation tools—including Copeptin-NEWS, NEWS-L, suPAR-NEWS, OPERA, and RISE UP—as well as artificial intelligence-based models, represent emerging personalized approaches to clinical deterioration prediction. Implications: No single score currently provides reliable early risk prediction for all elderly ED patients. Moving beyond “one-size-fits-all” EWSs toward adaptive, person-centered models may better reflect the complexity of geriatric emergency care and improve prognostic accuracy.