The Clinical Frailty Scale offers little utility as part of a prediction model for community-dwelling older fallers at risk of re-presenting to the emergency department

Abstract

Background: There is no published literature on the predictive ability of the Clinical Frailty Scale (CFS) for falls risk specific to the Emergency Department (ED) population. This study aims to develop a prognostic model to determine the predictive ability of the CFS for ED falls' re-presentation in community-dwelling older people.

Methods: A retrospective observational cohort study was completed from July 2019 to July 2022 on community dwelling people aged 75 years and over who presented to the ED with an extrinsic fall and had a CFS score recorded. The primary outcome was fall-related re-presentation to ED; the secondary outcome was mortality. A flexible parametric survival model was applied with time to falls re-presentation, and post-estimation, used to predict the probability of another fall re-presentation within 6 months. Calibration was assessed and a decision support curve generated.

Results: The model demonstrated reasonable calibration-in-the-large (Slope = 0.999) and fit between CFS and probability of fall re-presentation. The CFS model displayed negligible discriminant ability (C-statistic = 0.534) for identifying older people at risk of falls-related ED re-presentations within 6 months of index presentation.

Conclusions: The CFS cannot be used to prognosticate an individual's risk of ED re-presentation within 6 months of an index extrinsic fall.

Keywords: Accidental Falls; Aged; Frailty; Geriatric Assessment; Patient Readmission.