Activity Limitation Stages empirically derived for Activities of Daily Living (ADL) and Instrumental ADL in the U.S. Adult community-dwelling Medicare population

Abstract

Background: Stages quantify severity like conventional measures but further specify the activities that people are still able to perform without difficulty.

Objective: To develop Activity Limitation Stages for defining and monitoring groups of adult community-dwelling Medicare beneficiaries.

Design: Cross-sectional.

Setting: Community.

Participants: There were 14,670 respondents to the 2006 Medicare Current Beneficiary Survey.

Methods: Stages were empirically derived for the Activities of Daily Living (ADLs) and the Instrumental Activities of Daily Living (IADLs) by profiling the distribution of performance difficulties as reported by beneficiaries or their proxies. Stage prevalence estimates were determined, and associations with demographic and health variables were examined for all community-dwelling Medicare beneficiaries.

Main outcome measurements: ADL and IADL stage prevalence.

Results: Stages (0-IV) define 5 groups across the separate ADL and IADL domains according to hierarchically organized profiles of retained abilities and difficulties. For example, at ADL-I, people are guaranteed to be able to eat, toilet, dress, and bathe/shower without difficulty, whereas they experience limitations getting in and out of bed or chairs and/or difficulties walking. In 2006, an estimated 6.0, 2.9, 2.2, and 0.5 million beneficiaries had mild (ADL-I), moderate (ADL-II), severe (ADL-III), and complete (ADL-IV) difficulties, respectively, with estimates for IADL stages even higher. ADL and IADL stages showed expected associations with age and health-related concepts, supporting construct validity. Stages showed the strongest associations with conditions that impair cognition.

Conclusions: Stages as aggregate measures reveal the ADLs and IADLs that people are still able to do without difficulty, along with those activities in which they report having difficulty, consequently emphasizing how groups of people with difficulties can still participate in their own lives. Over the coming decades, stages applied to populations served by vertically integrated clinical practices could facilitate large-scale planning, with the goal of maximizing personal autonomy among groups of community-dwelling people with disabilities.

Figure 1

(a) Activities of Daily Living (ADL) stage patterns of association between different types of mental and physical conditions. (b) Instrumental Activities of Daily Living (IADL) stage patterns of association between different types of mental and physical conditions. Patterns of association of ADL and IADL stages with different type(s) of mental and physical chronic health conditions. A multinominal logistic regression was applied to address type of condition. Absence of the condition served as reference for each type of condition. Each plotted association was adjusted for age, sex, race/ethnicity, and all other conditions. Relative risk ratios are indicated at the central points with the vertical lines representing 95% confidence intervals. All relative risk ratio estimates were weighted applying 2006 sample weights. Each set of 4 vertical lines shows stage I, II, III, or IV compared to stage 0 (no limitation). The Y axis shows the logarithm of relative risk with estimates weighted applying 2006 sample weights.

Figure 2

(a) Activities of Daily Living (ADL) stage patterns of association with perceived health status. (b) Instrumental Activities of Daily Living (IADL) stage patterns of association with perceived health status. Patterns of association of ADL and IADL stages with perceived health status rated by individual or proxy referenced to individuals of the same age as excellent (reference), very good, good, fair, or poor. Excellent perceived health served as reference. Comparisons were according to ADL-0 and IADL-0 as reference. Each set of 4 vertical lines shows stage I, II, III, or IV compared to stage 0 (no limitation). The Y-axis shows the relative risk with estimates weighted applying 2006 sample weights.