Health care burden as manifested by greater health care utilization rises markedly with advancing age due in part to increasing prevalence of multiple chronic medical conditions (multimorbidity). While multimorbidity is a robust risk factor for higher health care burden, current multimorbidity measures explain only a modest proportion of the variation in subsequent health care costs. In order to constrain health care burden without negatively impacting health outcomes, improved understanding of key determinants of health care utilization among older adults is essential. Risk factors from other domains (including potentially modifiable determinants) would enhance identification of older community-dwelling adults likely to require costly care. Thus, models of health care utilization including the potentially modifiable (but not routinely measured in clinical settings) domains of the frailty phenotype and self-reported functional limitations may improve characterization of patient risk profiles and therefore improve targeting and design of interventions to reduce subsequent health care burden. The overarching objective of this proposal is to construct interpretable, generalizable and validated models of health care costs and utilization among older community dwelling adults that identify clinical risk factors amenable to targeted interventions to reduce risk of subsequent extensive and costly health care. We will take advantage of the linkage between community-dwelling participants in four cohort studies of older adults (Study of Osteoporotic Fractures [SOF]; Osteoporotic Fractures in Men Study [MrOS]; Health Aging and Body Composition Study [Health ABC]; and National Health and Aging Trends Study [NHATS]) and their Medicare claims data. We will perform a comprehensive evaluation of key potential clinical risk factors (frailty phenotype, self-reported functional limitations and components of frailty phenotype and functional limitations) for higher health care costs and utilization, after accounting for demographics and claims-based indicators of multimorbidity and frailty. Our application is consistent with the NIA mission to conduct biological, clinical, behavioral, social, and economic research related to the diseases and conditions associated with the aging process. Our study combines a wide range of data from four large epidemiologic studies leveraging the value of these cohorts in an efficient manner. If our hypotheses are confirmed, findings from our analyses will more accurately characterize subsets of older community-dwelling adults at risk for intense, costly health care; benefit health care systems/payers estimating costs vs. benefits of interventions aimed at delaying progression to frailty and disability; support assessment of the frailty phenotype, functional status, or specific individual components of these domains in the outpatient primary care practice setting; and direct the design of future targeted intervention trials aimed at reducing health care costs and utilization among the aged population.

Older adults have a disproportionate share of health care utilization, but current models inadequately characterize individuals likely to require extensive and costly care. We hypothesize that direct assessments of the frailty phenotype, functional limitations or their individual components will improve prediction of health care utilization in community-dwelling older adults beyond that provided by demographics and claims-based measures of multimorbidity and frailty. Findings will enhance clinical decision-making improving targeting of interventions to delay or prevent progression to frailty and disability thereby facilitating efforts to better tailor integrative patient-centered care to high risk individuals.