Impact of Medicare eligibility on informal caregiving for surgery and stroke

Abstract

Objective: To assess whether the intensity of family and friend care changes after older individuals enroll in Medicare at age 65.

Data sources: Health and Retirement Study survey data (1998-2018).

Study design: We compared informal care received by patients hospitalized for stroke, heart surgery, or joint surgery and who were stratified into propensity-weighted pre- and post-Medicare eligibility cohorts. A regression discontinuity design compared the self-reported likelihood of any care receipt, weekly hours of overall informal care, and intensity of informal care (hours among those receiving any care) at Medicare eligibility.

Data collection: Not applicable.

Principal findings: A total of 2270 individuals were included; 1674 (73.7%) stroke, 240 (10.6%) heart surgery, and 356 (15.7%) joint surgery patients. Mean (SD) care received was 20.0 (42.1) weekly hours. Of the 1214 (53.5%) patients who received informal care, the mean (SD) care receipt was 37.4 (51.7) weekly hours. Mean (SD) overall weekly care received was 23.4 (45.5), 13.9 (35.8), and 7.8 (21.6) for stroke, heart surgery, and joint surgery patients, respectively. The onset of Medicare eligibility was associated with a 13.6 percentage-point decrease in the probability of informal care received for stroke patients (p = 0.003) but not in the other acute care cohorts. Men had a 16.8 percentage-point decrease (p = 0.002) in the probability of any care receipt.

Conclusions: Medicare coverage was associated with a substantial decrease in family and friend caregiving use for stroke patients. Informal care may substitute for rather than complement restorative care, given that Medicare is known to expand the use of postacute care. The observed spillover effect of Medicare coverage on informal caregiving has implications for patient function and caregiver burden and should be considered in episode-based reimbursement models that alter professional rehabilitative care intensity.

Keywords: Medicare; caregiving; discontinuity; stroke; surgery.

FIGURE 1

Change in overall weekly caregiving hours received at Medicare eligibility. HRS respondents ages 59 to 74, comparing pre and post‐eligibility cohorts, matched on sociodemographics and health characteristics. The year 67 was used as the cut‐point given the HRS' use of a 2‐year lookback window on interview questions about stroke and heart and joint surgeries. The pre and post‐Medicare eligibility trends are estimated using a two‐part regression model with a post indicator, a continuous age indicator, and an interaction of post and age; plotted values are unadjusted values of the dependent variable at given values of the age.

FIGURE A1

McCrary test results. The test was performed using ‘rddensity’ command in Stata. T = 1.24, p = 0.22. [Color figure can be viewed at wileyonlinelibrary.com]