Systematic Biases in Functional Status Assessment of Elderly Adults: Effects of Different Data Sources

Abstract

Measuring functional status using specific instruments is an important part of geriatric assessment. These instruments, however, often rely on data sources different from those with which they were originally validated. To study possible biasing effects of different data sources on functional status scores, we examined scores for two widely used instruments (the Lawton Personal Self-Maintenance Scale, PSMS, and Instrumental Activities of Daily Living, IADL, scale) on a group of hospitalized elderly (n = 61) using three different data sources (the patients themselves, the patients' nurses, and significant others). Analysis showed that PSMS scores derived from patients were significantly higher than scores derived from significant others (p < .025) and that patient-derived IADL scores were significantly higher than both nurse-derived scores (p < .001) and significant-other-derived scores (p < .001). We also compared scores for a group of nursing home patients (n = 68) on the Katz Activities of Daily Living (ADL) Scale, using data obtained from patients and their nurses. Again, the patient-derived scores were significantly higher than those from nurses (p < .001). We conclude that data sources for determining patient functional ability are not interchangeable and that patients may overstate their functional abilities, whereas significant others may understate them, relative to judgments of skilled nursing personnel.