Objective. Public e-penditures for home health care grew rapidly in the 1990s, but it remains unclear to whom the additional services were targeted. This study tests whether the rapidly increasing e-penditures were targeted to the elderly with high levels of disability and low levels of social support, 2 groups that have historically been higher users of paid home health and nursing home services.
Methods. The Asset and Health Dynamics Study, a nationally representative, longitudinal survey of people >=70 years of age (n = 7,443), was used to determine the association of level of disability and level of social support with the use of paid home care services in both 1993 and 1995. Multivariable regression models were used to adjust for sociodemographics, recent hospital or nursing home admissions, chronic medical conditions, and receipt of informal care from family members.
Results. Those with higher levels of disability received more adjusted weekly hours of paid home care in both 1993 and 1995. In 1993, users of paid home care with the least social support (unmarried living alone) received more adjusted weekly hours of care than the unmarried elderly living with others (24 versus 13 hours, P <0.01) and the married (24 versus 18 hours, P = 0.06). However, by 1995, those who were unmarried and living with others were receiving the most paid home care: 40 versus 26 hours for the unmarried living alone (P <0.05) and 24 hours for the married (P <0.05).
Conclusions. The recent large increase in formal home care services went disproportionately to those with greater social support. Home care policy changes in the early 1990s resulted in a shift in the distribution of home care services toward the elderly living with their children.