The effect of self-reported and performance-based functional impairment on future hospital costs of community-dwelling older persons

Purpose: We determined the prognostic value of selfreported and performance-based measurement of function, including functional transitions and combining different measurement approaches, on utilization. Design and Methods: Our cohort study used the 6th, 7th, and 10th waves of three sites of the Established Populations for Epidemiologic Studies of the Elderly, linked to 1- and 4-year Medicare Part A hospital costs. We examined mean hospital expenditures based on (a) 1- and 4-year transitions in selfreported functional status; (b) 4-year transitions in performance-based functional status; (c) combined baseline self-reported and perform a nce-based functional status; and (d) poorest self-reported and performance-based functional status during a 4-year period. Results: Even modest declines in selfreported or performance-based functional status were associated with increased expenditures. When baseline self-reported and perform a nce-based assessments were combined, mean 1-and 4-year adjusted costs were higher with progressively worse performance-based scores, even among those who were independent in self-reported function. When the poorest 4-year self-reported and performance-based functions were examined, self-reported functioning was the most important determinant of hospital costs, but within each self-reported functional level, poorer performance-based function was associated with progressively higher costs. Implications: The costs associated with even modest functional decline are high. Combining self-reported and performance-based measurements can provide more precise estimates of future hospital costs.