Andrew P Costa

Background Little is known about the prevalence of frailty among patients with memory concerns attending a primary care-based memory clinic. Objective This study aims to describe the prevalence of frailty among patients attending a primary care-based memory clinic and to determine if prevalence rates differ based on the screening tool that is used. Methods We conducted a retrospective medical record review for all consecutive patients assessed in a primary care-based memory clinic over 8 months. Frailty was measured in 258 patients using the Fried frailty criteria, which relies on physical measures, and the Clinical Frailty Scale (CFS), which relies on functional status. Weighted kappa statistics were calculated to compare the Fried frailty and the CFS. Results The prevalence of frailty was 16% by Fried criteria and 48% by the CFS. Agreement between Fried frailty and CFS was fair for CFS 5+ (kappa = 0.22; 95% confidence interval: 0.13, 0.32) and moderate for CFS 6+ (kappa = 0.47; 0.34, 0.61). Dual-trait measures of hand grip strength with gait speed were found to be a valid proxy for Fried frailty phenotype. Conclusions Among primary care patients with memory concerns, frailty prevalence rates differed based on the measure used. Screening for frailty in this population using measures relying on physical performance may be a more efficient approach for persons already at risk of further health instability from cognitive impairment. Our findings demonstrate how measure selection should be based on the objectives and context in which frailty screening occurs.