Incremental healthcare costs of the simple SOF measure of phenotypic frailty in community-dwelling older adults.

Abstract

Frailty defined by the Cardiovascular Health Study (CHS) phenotype is associated with higher healthcare expenditures in community-dwelling Medicare beneficiaries after accounting for claims-based cost indicators. However, frailty assessment using the CHS phenotype is often not feasible in routine clinical practice. We evaluated whether frailty identified by the simple Study of Osteoporotic Fractures (SOF) phenotype is associated with subsequent incremental costs after accounting for claims-derived cost indicators. Prospective study utilizing data from four cohort studies of older adults linked with Medicare claims composed of 8264 community-dwelling fee-for-service beneficiaries (4389 women, 3875 men). SOF Frailty Phenotype (three components: weight loss, poor energy, and inability to rise from chair five times without using arms) and CHS Frailty Phenotype (operationalized using five components) derived from cohort data. Participants were classified as robust, prefrail, or frail using each phenotype. Multimorbidity index (CMS Hierarchical Conditions Categories score) and Kim frailty indicator (approximating the deficit accumulation index) derived from claims. Annualized total and sector-specific healthcare costs ascertained for 36 months after frailty assessment. Average annualized total healthcare costs (2023 US dollars) were $15,021 in women and $15,711 in men. After accounting for claims-based multimorbidity and frailty indicators, average incremental costs of SOF phenotypic frailty (two or three components) versus robust (none) were $7142 in women and $5961 in men, only modestly lower than incremental costs of CHS phenotypic frailty ($9422 in women, $6479 in men). SOF phenotypic frailty in both sexes was associated with higher subsequent expenditures in the inpatient, skilled nursing facility, and home healthcare sectors. As observed with CHS phenotypic frailty, SOF phenotypic frailty is associated with higher subsequent total and sector-specific expenditures after accounting for claims-derived indicators. The parsimonious SOF phenotype can be readily assessed in space-constrained and time-limited practice settings to improve identification of older adults at high risk of costly care.