Appendix E

Methodology: NHATS and NSOC Surveys

This report presents data on older adults and their family caregivers drawn from the public use files of the 2011 National Health and Aging Trends Study (NHATS) and the National Study of Caregiving (NSOC). They are two linked national studies led by the Johns Hopkins University Bloomberg School of Public Health, with data collection by Westat, and support from the National Institute on Aging for NHATS and the Assistant Secretary for Planning and Evaluation of the U.S. Department of Health and Human Services for NSOC (NHATS, 2015). Extensive technical documentation of the surveys’ designs is available at: https://www.nhats.org/scripts/TechnicalPapers.htm and http://www.nhats.org.

NHATS is nationally representative of Medicare beneficiaries aged 65 years and older. Respondents (or their proxies) living in the community and in residential care settings, other than nursing homes, participated in a 2-hour in-person interview that included self-reports and validated performance-based measures of disability (Kasper et al., 2013a). For those living in nursing homes, an interview was conducted with a member of the facility staff to learn about the respondent’s service environment. Study participants were asked whether and how they performed daily activities in the month before the interview. Among older adults who received assistance, a detailed helper roster was created listing the relationship and specific activities for each person providing assistance. Nursing home residents were not included in generating the helper roster.

NSOC respondents (i.e., family caregivers of the NHATS respondents) were family members or other unpaid helpers who provided assistance with mobility, self-care, household activities, transportation, or medically

oriented tasks. A telephone interview was conducted with up to five family caregivers (i.e., “helpers”) for each older adult. For older adults with more than five eligible helpers, the five helpers were selected at random.

Of 7,609 NHATS participants living in the community or in a residential care facility, 2,423 were included in the NSOC sampling frame, and 4,935 helpers met NSOC eligibility criteria. An NSOC non-response can arise from the NHATS participant (who may refuse to provide contact information for helpers) or his or her caregivers (who may refuse to participate) (Kasper et al., 2013a). The NHATS participants did not provide contact information for 1,573 eligible family caregivers, and 1,355 of the remaining 3,362 eligible family caregivers could not be located or refused to respond, yielding 68.1 percent and 59.7 percent of first-stage and second-stage response rates, respectively. In total, 2,007 family caregivers of 1,369 older adults responded in 2011 to the NSOC.

Observations from NHATS and NSOC are weighted to produce nationally representative estimates and to account for the surveys’ complex sampling designs. Weights adjust for differential probabilities of selection at both the NHATS sample person and caregiver levels. The analyses presented in this report were conducted with statistical software (Stata v.12) using the survey sampling weights provided to NSOC users.

CLASSIFYING NHATS PARTICIPANTS BY DEMENTIA STATUS

Several analyses presented in this report distinguish among three groups of NHATS participants—those with probable dementia, possible dementia, or no dementia. NHATS assigns these categories based on the following:

• A report by the sample person or proxy respondent that a doctor told the sample person that he/she had dementia or Alzheimer’s disease.
• A score indicating “probable dementia” on the AD8 Dementia Screening Interview (which was administered to proxy respondents to the NHATS interview). The AD8 is a brief informant interview used to detect dementia (Galvin et al., 2005, 2006).
• Cognitive tests that evaluate the sample person’s memory (immediate and delayed 10-word recall), orientation (date, month, year, and day of the week; naming the President and Vice President), and executive function (clock drawing test).

A report by either the NHATS participant or a proxy respondent that a doctor told the sample person that he/she had dementia or Alzheimer’s disease was used to classify persons as having probable dementia (Kasper et al., 2013b). Proxy respondents not reporting a diagnosis who gave

answers to the AD8 who met criteria for likely dementia (a score of 2 or higher) also were classified as having probable dementia. For all others—self-respondents not reporting a diagnosis and a small number (n = 79) with proxy respondents who had no diagnosis reported and did not meet AD8 criteria, but had test information—score cut-points applied to cognitive tests assessing three domains (memory, orientation, executive functioning) were used. Impairment was defined as scores at or below 1.5 standard deviations (SDs) from the mean for self-respondents. Impairment in at least two cognitive domains was required for probable dementia; a cut-point of <1.5 SDs below the mean in one domain was used for cognitive impairment, indicating possible dementia.

REFERENCES

This page intentionally left blank.