Avaliação longitudinal da fragilidade em idosos comunitários e comparação entre a Escala de Fragilidade de Edmonton (EFS) e o Índice de Vulnerabilidade Clínico-Funcional-20 (IVCF-20)

Abstract

The aging-associated particularities have become more evident with the increased proportion of elderly individuals in the general population, with more occurrences in all countries, especially in developing countries, such as Brazil. The present study aimed to longitudinally assess frailty in community elderly and to compare the Edmonton Frail Scale (ESF) with the Clinical-Functional Vulnerability Index (IVCF-20) instrument, applied in the same population. A longitudinal prospective and analytical study was performed in order to assess the factors associated with worsened frailty. To compare frailty in the elderly by different instruments, a cross-sectional study nested to a population-based cohort was carried out. Data collection at baseline was performed at the elderly’ homes from May to July 2013 via random sampling, by conglomerates. The data collection of the first wave in this study was performed in average 42 months, between November 2016 and March, 2017. Demographic and socioeconomic variables, morbidities and health services use were evaluated. The frailty was assessed by the Edmonton Frail Scale (EFS), during the two longitudinal evaluated periods, and by the Clinical-Functional Vulnerability Index (IVCF-20) in the cross-sectional study. Demographic, sociodemographic, morbidities and health services use variables were evaluated. Frailty was assessed via EFS, during the two data collection periods in the longitudinal study, and via IVCF-20 in the cross-sectional study. The adjusted prevalence ratios were obtained by multiple Poisson regression analysis with robust variance. The Kappa statistics evaluated the agreement degree, considering the frailty dichotomization (fragile vs. not fragile), while the Pearson coefficient assessed the instruments correlation, considering the total scores of each instrument. 394 elderly subjects agreed to participate in the study, being 21.8% with frailty worsening. The variables associated with the transition to a worse frailty state were: polypharmacy, negative self-perception of health, weight loss and hospitalization in the last 12 months. In the cross-sectional study, the frailty prevalence was 36.8% via EFS and the Clinical-Functional Vulnerability high-risk prevalence was 19.5% via IVCF-20. The Kappa statistic was 0.599, and the Pearson correlation coefficient was 0.755 (p<0.001). The variables that remained statistically associated to frailty after multiple analysis, in both instruments, were: age equal or superior to 80 years old, stroke history, polypharmacy, negative self-perception of health, fall and hospitalization in the previous 12 months. Education less than four years, osteoarticular disease and weight loss were associated with frailty only by EFS, while having caregiver associated with a high risk of clinicalfunctional vulnerability only by IVCF-20. Frailty was confirmed as a dynamic process, characterized by transition between frailty levels over time. Health-associated factors showed longitudinal association with frailty worsening. Although the EFS and IVCF-20 instruments have moderate agreement and strong positive correlation and present similar characteristics for the identification of frailty associated factors, the frailty prevalence pointed was discrepant. These findings highlight the need for standardization of an instrument to assess frailty in community elderly.