Comparison of the long-term prognostic value of different frailty instruments in older inpatients: a 5-year prospective cohort study.

Background: Frailty is associated with increased mortality in older adults, but limited studies compare frailty instruments among inpatients with long-term follow-up.

Objective: To evaluate five frailty scales for predicting 5-year all-cause mortality in older inpatients.

Methods: This prospective cohort study enrolled 917 inpatients aged ≥ 65 years. We used five commonly used scales [Clinical Frailty Scale (CFS), FRAIL, Fried, Edmonton, and the comprehensive geriatric assessment-frailty index (CGA-FI)] to screen or assess frailty and then conducted a 5-year telephone follow-up. The primary endpoint was 5-year all-cause mortality. The predictive value of different frailty scales was compared using Kaplan-Meier (K-M) survival analysis, COX regression models, and the receiver operating characteristic (ROC) curves.

Results: The prevalence of frailty ranged from 19.5 to 36.5%. Both K-M survival curves and Cox regression confirmed that frailty patients had higher mortality risk across all scales. After multivariate adjustment, the hazard ratios from highest to lowest, were: CGA-FI, FRAIL, Fried, CFS, and Edmonton (all p < 0.05). Frailty demonstrated moderate performance, with area under the curves (AUCs) ranging from 0.70 to 0.75 (all p < 0.001). CGA-FI had the largest AUC of 0.724, revealing the best predictive value, while FRAIL had the smallest AUC of 0.666. The AUCs of Fried, Edmonton, and CFS gradually decreased, with no statistical differences. Furthermore, CFS has the highest sensitivity (77.5%).

Conclusions: Frailty identified by all scales is associated with an increased risk of long-term mortality. CFS is the preferred frailty screening scale, while CGA-FI is the most accurate assessment scale. Trial registration ChiCTR1800017204 (07/18/2018).