Frailty classification challenges in the emergency department: agreement and variability in clinical frailty scale scoring

Le résumé fourni par la source

BACKGROUND: Frailty assessment in the emergency department (ED) is essential but challenging. The Clinical Frailty Scale (CFS) is widely used, although inter-rater variability has been reported across assessors and assessment methods. AIMS: To assess agreement between ED-assigned and retrospectively assigned CFS scores, and to explore characteristics associated with changes in frailty classification. METHODS: We included 500 patients aged ≥ 65 years admitted with suspected pneumonia to Haukeland University Hospital (2019–2023). CFS was initially scored by nurses in the ED and reassessed retrospectively by a geriatric-trained physician using chart review. The retrospective assessment had access to broader and more objective information, including formal documentation from care services on daily function. CFS scores were categorized as fit (1–3), prefrail (4) and frail (5–9). Agreement was measured by Intraclass Correlation Coefficient (ICC). RESULTS: CFS scores changed numerically in 252 (50.4%) patients and frailty category in 144 (28.8%). Agreement was moderate to good (ICC 0.73; 95% CI 0.68–0.77; p < 0.001), with retrospective assessment yielding higher scores. Agreement was highest in frail (89.7%) and lowest in prefrail patients (44.8%). Recategorized patients were older, more comorbid, more often community-dwelling with greater care dependency, and had higher 1-year mortality. DISCUSSION: Real-time frailty assessments in the ED may underestimate frailty, particularly among patients who are community-dwelling, older, or have complex health conditions. Retrospective assessments, informed by objective documentation of functional needs, likely reflect baseline frailty more accurately. CONCLUSIONS: Frailty is often underestimated in the ED. Improved access to collateral information may improve assessment accuracy.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.