Identifying hospitalized older patients at varying risk for physical performance decline: A new approach

OBJECTIVE: A classification tree analysis identifies patient groups at varying risk for decline in physical performance 1 year after hospitalization. DESIGN: Prospective cohort study. SETTING: Tertiary care VAMC. PARTICIPANTS: A total of 507 acutely ill hospitalized male veterans aged 65 years and older. MEASUREMENTS: Eighteen admission characteristics were considered as potential predictors: demographic data, medical diagnoses, functional status (e.g., ADL and IADL), geriatric conditions (e.g., incontinence, vision impairment, weight change), mental status, depression, and physical functioning (measured by self-report (MOS-PFR) and the Physical Performance and Mobility Examination (PPME)). Outcome measure was change in PPME status at 12-months postadmission. RESULTS: Patients with the greatest risk for decline had both high baseline physical performance (PPME greater than or equal to 9) and at least moderate self-report limitations on physical functioning (MOS-PFR less than or equal to 36, mean = 30.8). Patients with the lowest risk of decline had impaired baseline physical performance (PPME less than or equal to 8) but fewer self-report limitations on physical functioning (MOS-PFR greater than or equal to 31, mean = 37.4) and two or less geriatric conditions. CONCLUSIONS: The predictive role of self-report functioning suggests that perception of the impact of health on one's own physical functioning is associated with future performance. The number of geriatric conditions is also an important predictor of physical performance change. By identifying patient risk groups based on geriatric conditions, physical performance, and self-report physical functioning, future targeting strategies may improve physical performance outcomes for hospitalized older adults.