Disability and Place of Death in Older Americans: The Moderating Role of Household Wealth

Abstract

ContextHome-based deaths are increasing, yet, how wealth influences where people die in the presence of disability remains unknown. ObjectiveTo examine place of death by help with (instrumental) activities of daily living (I/ADLs) at the end of life (EOL) and the modifying role of wealth. MethodsRetrospective study of decedents from the Health and Retirement Study (n = 13,210). The exposure was intensity of help with I/ADLs at the EOL (no help/ lower intensity/higher intensity). The outcome was place of death (hospital/nursing home/home). Household wealth was an effect modifier with six categories: ≤$0, first-fifth quintile. Covariates included age, gender, race, marital status at the EOL, last place of residence, and receipt of hospice care. We used multinomial logit regression models with estimates reported as average marginal effects (AMEs). ResultsMean age was 79.8 years; 53.2% were female. In the adjusted models, compared to not receiving help at EOL, receiving higher-intensity help was associated with a lower probability of dying in a hospital (AME = −3.8 percentage points (pp), 95% CI = −6.3 to −1.3) and a higher probability of dying at home (AME = 3.6 pp, 95% CI = 1.4–5.7). Associations were most pronounced among decedents in the top two wealth quintiles; older adults who received higher-intensity help had a lower probability of dying in a hospital (AME = −9.0 pp, 95% CI = −14.8 to −3.1), and a higher probability of dying at home (AME = 8.4 pp, 95% CI = 3.8–13.0). ConclusionReceiving higher intensity of help with I/ADLs was associated with lower likelihood of dying in a hospital, and higher likelihood of dying at home, particularly among older adults with greater wealth.