The frailty of constructs and the construct of frailty in geriatric medicine and research

Frailty has been defined as a “state of increased vulnerability” which increases the risk of adverse outcomes. Although there is currently no consensus definition, frailty has been consistently associated with death, institutionalization, and disability. These associations have driven strong, but mostly unheeded, calls from the research and clinical community to include frailty in clinical care processes. Geriatric research and practice rely on age-related constructs which diverge from the traditional biomedical model of disease: frailty, multimorbidity, and other geriatric syndromes are used to describe and characterize the variability in the health status of older adults. Despite interest for such constructs, issues related to their definition, reliability, and applicability in the context of clinical practice have not been well explored.The overarching objective for this dissertation was to explore the conceptual underpinning and clinical applicability of age-related constructs, with an emphasis on frailty. The first objective was to examine whether older adults show increasing heterogeneity with age, as a motivation for developing age-specific constructs. A novel application of methods was used to disentangle the within-age and between-age variability of 34 health characteristics within eight domains in 30,097 participants from the Canadian Longitudinal Study on Aging (CLSA). I demonstrated that heterogeneity increases with aging, but not for all health characteristics and domains, and not uniformly. Clinical implications and research opportunities of heterogeneity include the importance of the comprehensive geriatric assessment, of measurement and scaling of variables, and the need to develop new multidimensional constructs that distinguish older adults among themselves. The second objective was to examine the measurement of age-related constructs. I investigated the reliability and clinical correlates of reliability of the deficit-accumulation frailty index. Monte Carlo methods were used to simulate 12,000 studies comparing various implementations of the frailty index in 12,080 participants 65 years and older from the CLSA. I showed that the number and composition of items of individual FIs strongly influence their reliability. Descriptive estimates using frailty indices and predictive estimates between frailty indices and mortality varied between implementations. This lack of reliability and stability of estimates lowers the generalizability and clinical application of study findings.The third objective focused on determining the applicability of study results to clinical practice. A literature search was conducted to identify 26 existing frameworks appraising study “applicability.” I analyzed and synthesized frameworks and criteria according to the scope and level of aggregation of the evidence appraised, the target user, and the specific area of applicability. A novel framework to appraise clinical applicability was proposed which categorizes studies into three evidence domains (research domain, practice informing, and practice changing) using six criteria (Validity, Indication-informativeness, Clinical relevance, Originality, Risk-benefit comprehensiveness, and Transposability, VICORT).This framework was used for the fourth objective to appraise the clinical applicability of recently published articles on frailty. A mapping review was conducted by sampling 476 articles published in 2017–2018 and investigating whether these articles informed practice, changed practice, or belonged in the research domain. Among all articles, 63 (13%) articles were categorized as practice informing, 11 (2%) as potentially practice changing, and 1 (0.2%) as clearly practice changing. The lack of indication-informativeness (96%) and originality (83%) were the most important reasons hampering clinical applicability. Recommendations are proposed for future research on frailty, which may also extend to other age-related constructs

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.