Appendix Exhibit A3.
Description of prediction models for formal care - Samples, outcomes, main predictors, and covariates
| Main predictor: Family Availability | Model | Outcome: Formal care from | Control variables | Sample restrictions | |
|---|---|---|---|---|---|
| Spouse availability | presence of spouse | 1 | Any type | Base controlb | Base sample restrictiona (N=4,259) |
| 2 | Nursing home employee | ||||
| disability status of spouse | 3 | Any type | Base sample restrictiona + spouse present (N=1,783) | ||
| 4 | Nursing home employee | ||||
| working status of spouse | 5 | Any type | Base controlb + spouse disability (IADL/ADL) | ||
| 6 | Nursing home employee | ||||
| Child availability | presence of adult children | 7 | Any type | Base controlb + marital status of respondent | Base sample restrictiona (N=4,259) |
| 8 | Nursing home employee | ||||
| number of adult children | 9 | Any type | Base sample restrictiona + adult child present (N=3,795) | ||
| 10 | Nursing home employee | ||||
| presence of daughter | 11 | Any type | Base controlb + marital status of respondent + number of adult biological children | ||
| 12 | Nursing home employee | ||||
| working status of adult children | 13 | Any type | |||
| 14 | Nursing home employee | ||||
| distance to adult child | 15 | Any type | |||
| 16 | Nursing home employee | ||||
a
Base sample restriction is to include adults 55+ with dementia and at least one ADL limitation in interview year T (T=2002–2014) and community-dwelling in year T-2.
b
Base control includes year, interview mode, proxy status, age, gender, race/ethnicity, education, wealth in quartile, and the number of ADL limitations in interview year T-2.