Screening and managing frailty in older nursing home residents with Frail-VIG index: Feasibility, reliability and predictive validity for mortality

Frailty is prevalent among older nursing home residents, although there is limited evidence regarding frailty screening and management in this setting. OBJECTIVE: To evaluate the measurement properties of the Frail Index based on the Comprehensive Geriatric Assessment (Frail-VIG). DESIGN: Prospective observational longitudinal study of 571 residents from 3 nursing homes. Frail-VIG scores were calculated at baseline and at 6 and 12 months. Sociodemographic variables were studied. Feasibility was assessed based on simplicity of application and requirements for score calculation. Reliability was evaluated through inter-rater agreement and test-retest assessments. Construct and content validity were examined by comparing it with other frailty indexes. Predictive validity was evaluated using log-rank tests and AUC-ROC curves for mortality prediction. RESULTS: Mean (SD) resident age was 88.2 (6.5) years, and 80.6 % were women. The mortality rate was 11.4 % at 6 months and 20 % at 12 months. Calculating Frail-VIG scores required 5.15 min and no additional space or equipment, and there was low risk of missing data. The inter-rater consistency and score stability over time indicate strong reliability. The Frail-VIG maintains the characteristics of other established frailty indexes and shows strong convergent validity with the FRAIL-NH and CFS scales. Baseline scores have an AUC-ROC curve (confidence interval) of 0.69 (95 % CI, 0.63-0.76) at 6 months and 0.65 (95 % CI, 0.6-0.71) at 12 months. CONCLUSIONS: The measurement properties of the Frail-VIG in older nursing home residents validate its use in this population and setting. Its predictive ability for mortality suggests important implications for advanced care planning.