Policy Changes in Medicare Home Health Care: Challenges to Providing Family-Centered, Community-Based Care for Older Adults

Abstract

The Balanced Budget Act of 1997 (BBA) established new reimbursement systems in the Medicare home health fee-for-service benefit. Reimbursements were reduced to 1993 levels and per-beneficiary capitated limits were introduced for the first time. This article analyzes the impact of these changes on chronically ill older adults and their families. The study combined a secondary analysis of the Provider of Service file (1996, 1999, 2002, and the Medicare Current Beneficiary Survey (1996, 1998) with qualitative interviews of home health agency directors. The greatest decreases in staff and visits were for medical social work and home health aide services. Patients with caregivers saw greater decreases in visits and reimbursements for all visits, skilled nursing, medical social work, and home health aide visits. Agency directors reported that they increased caregiver education, training, and involvement in care in order to discharge patients sooner. Additional research is needed to understand the long-term, adverse impact of these policy changes on chronically ill patients and their families.

KEYWORDS:

• access
• home health
• medicare
• policy

The author would like to thank the following organizations for their generous support of this research: the Andrus Foundation, the John A. Hartford Foundation, the Centers for Medicare and Medicaid Services, and the Rivitz Fund, Bryn Mawr College.

Notes

Note. Data source: CMS's Provider of Service file, 1996, 1999, 2002.

Note. Data source: CMS's Provider of Service file, 1996, 1999, 2002.

Note. Data Source: MCBS. ∗p < / = .05; ∗∗p < / = .01; ∗∗∗p < / = .001; N = 2437.

Note. Data Source: MCBS. ∗p < / = .05; ∗∗p < / = .01; ∗∗∗p < / = .001; †Results from multiple logistic regression models with interactions with year 1998. N = 2,437; b_i = direct variable effects; b_j = year 1998 effects; b_ij = interaction between variable and year 1998; a = the difference in likelihood between 1996 and 1998 derived from the following formula [(e^b−1)100]. All models include the full set of control variables: age, sex, marital status, years of education, race, cognitive impairment, supplemental insurance, Medicaid eligibility, census region, metro versus nonmetro residence, number of informal caregivers, functional status, health status, and number of diagnoses.

Note. Data Source: MCBS. ∗p < / = .05; ∗∗p < / = .01; ∗∗∗p < / = .001; † = The main interaction variable in all models is year 1998 or b_j; b_i = direct variable effects; b_ij = interaction between variable and year 1998. a = calculated from log transformed multivariate regression models the percent change in visits between 1996 and 1998 derived from the following formula [(e^b−1)100]. c = nonskilled services include medical social work and home health aide. All models include the full set of control variables: age, sex, marital status, years of education, race, cognitive impairment, supplemental insurance, Medicaid eligibility, census region, metro versus nonmetro residence, number of informal caregivers, functional status, health status, and number of diagnoses.