Frailty does not always matter: insights from an elderly, multimorbid acute heart failure cohort (RICA-2 registry)
D Chivite Guillen, A Contra Carne, J L Corcoles Satorre, A Sanchez De Alcazar Del Rio, E Montero Hernandez, V Garrigasait Vilaseca, J P Ferreira, M Sanchez Marteles, J Perez Silvestre, J C TrullasAbstract
Background
Frailty is highly prevalent in elderly patients experiencing an episode of acute heart failure (AHF) and is associated with adverse outcomes. However, frailty instruments capture heterogeneous constructs, and their incremental prognostic value beyond functional dependence remains uncertain.
Purpose
To assess the concordance and added prognostic impact of two commonly used frailty instruments—the FRAIL instrument (FI) and the Clinical Frailty Scale (CFS)—on short- and mid-term outcomes in elderly patients with AHF.
Methods
Elderly, multimorbid patients from an Internal Medicine spanish multicentric registry of AHF patients (RICA-2) were prospectively evaluated. Pre-AHF frailty was assessed using FI and CFS, and functional dependence using the Barthel Index (BI). For comparative analyses, CFS was grouped into three categories (non-frail, pre-frail, frail) paralleling those of FI. Agreement between FI and CFS was evaluated using weighted Cohen’s kappa. Outcomes included all-cause mortality and heart failure readmission at 30 days and 12 months. Survival was assessed using Kaplan–Meier curves and log-rank tests. Cox proportional hazards models, sequentially adding FI or CFS and subsequently BI to a baseline clinical model were performed. Model discrimination was evaluated using the area under the receiver operating characteristic curve (AUC)
Results
A total of 786 elderly, multimorbid AHF patients were included (mean age 84 years, 51% female, mean age-adjusted Charlson Comorbidity Index 5). Agreement between FI and CFS was low to moderate in the overall cohort (weighted κ = 0.47) and improved among patients with complete 12-month follow-up (weighted κ = 0.53). Neither FI nor CFS independently predicted 30-day mortality or readmission. At 12 months, both frailty measures were associated with mortality in unadjusted analyses, with rates increasing from approximately 12–14% in non-frail patients to over 35% in frail patients. Kaplan–Meier analyses showed similar survival gradients across frailty categories for FI and CFS (log-rank p < 0.001). Baseline clinical models showed modest discrimination for 12-month mortality (AUC 0.699). Discrimination improved after addition of FI or CFS (AUC 0.726 and 0.721, respectively) and further increased after inclusion of BI (AUC 0.742 and 0.739, respectively). Cox regression analyses yielded consistent results, with frailty measures losing independent prognostic significance after adjustment for functional status, while BI remained a robust predictor in the final models (p < 0.001)
Conclusions
In elderly, multimorbid patients with AHF, FI and CFS show moderate concordance but similar unadjusted prognostic gradients. Although frailty measures modestly improve risk discrimination, funcional status emerges as the main driver of mid-term mortality, questioning the incremental prognostic value of routine frailty assessment when disability is comprehensively evaluated in this population.Kaplan–Meier 12-month survival curvesFor image description, please refer to the figure legend and surrounding text.ROC curves for 12-month survivalFor image description, please refer to the figure legend and surrounding text.