The NHIS-D employed a two-stage design to investigate the prevalence and impact of disability in the US. Phase 1 screened for conditions expected to have high dependency needs. Phase 2 comprised in-depth interviews of individuals with disability. The objective of the survey was to provide better data on underlying factors and services needed. The goal here is to investigate the impact of chronic balance/dizziness problems on functional abilities, however selection bias complicates the analysis. Since Phase 2 was conducted only for persons with disability, behaviors or other outcomes cannot easily be compared to the general population. Simply adjusting for overall health status did not much change the relative risks (RR), for example, of having difficulty in working or performing daily activities. Instead, the original weights (adjusted only for selection probability and age*sex*race distributions) were modified to account for the different patterns of self-reported health status between adults (18+ years of age) in Phase 1 and 2. RRs become much larger after adjusting separately for people with and without chronic balance/dizziness problems to match the health status of their counterparts in Phase 1. The RR of "having difficulty using the bathroom at home" among people with chronic balance/dizziness problems increases from 2.93 (95% confidence interval [CI]: 2.57-3.34) to 4.98 (CI: 4.25-5.84) and the RR of "retiring earlier" increases from 1.07 (CI: 0.98-1.29) to 1.68 (CI: 1.41-2.01). Detailed results of RR for demographic subgroups will be presented. Other adjustment techniques to obtain appropriate comparisons will also be considered.
Learning Objectives: N/A
Keywords: Disability, Survey
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.