Summarizing amount of difficulty in ADLs: a refined characterization of disability. Results from the women's health and aging study.

Activities of Daily Living (ADLs) rating scales often classify older persons as independent or dependent in self-care activities. However, with this type of classification system little information is available on people who, while not dependent, do report some difficulty in performing self-care activities. This 3-year prospective cohort study presents an ADL summary scale that assesses a gradient of difficulty in ADLs for moderately to severely disabled women aged 65 and older living in the Baltimore, Maryland area. At baseline and for each of six interviews done at 6-month intervals, an ADL summary score measuring ADL difficulty was created. ADL summary score slopes were created for each woman who completed at least the baseline and first three follow-up interviews. The baseline ADL summary score and slope of the ADL summary score were used to predict risk of future hospitalization, nursing home admission and death. Additionally, 6-month ADL summary change scores were calculated in order to assess change in the ADL summary score for women who did and did not report an acute event (MI, stroke or hip fracture) at follow-up. The ADL summary scale showed excellent reliability. Intraclass correlation coefficients ranged from 0.83 to 0.93 for measures made one week apart. The 6-month average correlation coefficient was 0.77. The ADL summary scale also performed well in tests of validity. Poorer scores on the ADL summary scale were significantly and inversely related to an objective physical performance scale. The baseline ADL summary score and slope of the ADL summary score predicted future risk of hospitalization, nursing home admission and death. Each unit increase in the slope of the ADL summary score was associated with a 19% (95% CI 1.10, 1.29) increased risk of hospitalization, a 57% (95% CI 1.37, 1.81) increased risk of nursing home admission and a 41% (95% CI 1.22, 1.64) increased risk of death, adjusting for potential confounders. The mean 6-month ADL change score for women reporting an acute event (2.59) was significantly different from those not reporting an acute event (0.52). The strongest and most significant mean 6-month ADL change score was observed for women who reported a hip fracture (4.52) followed by stroke (2.40) and MI (1.00). These results demonstrate that the ADL summary scale presented has validity, reliability and is sensitive to change. This scale, therefore, offers the opportunity to classify functional problems in a larger proportion of the population than do ADL dependence scales. Additionally, it gives us a way to identify earlier functional changes, and a way to track the natural history of functional problems as well as a response to interventions.