Do Medicare Accountable Care Organizations Impact Health Care Utilization among Long-Stay Nursing Home Residents?

Abstract

Objectives

Nursing home (NH) residents are high-cost, high-need Medicare beneficiaries. Accountable Care Organizations (ACOs) have the potential to improve quality of care and reduce potentially unnecessary health care utilization. This study aimed to assess the impact of Medicare Shared Savings Program (MSSP) ACOs on health care utilization among long-stay NH residents.

Design

“Intention-to-treat” and quasi-experimental design.

Setting and Participants

A national cohort of 158,259 fee-for-service Medicare beneficiaries who were long-stay NH residents in 2011 or 2018. In each year, residents were included in the sample the first time their Minimum Data Set (MDS) assessments (ie, index MDS) met the following inclusion criteria: (1) aged 66+; (2) dependence in 2 or more activities of daily living; (3) neither enrolled in hospice nor in coma; and (4) NH length of stay ≥90 days.

Methods

We followed residents’ health care utilization and Medicare expenditures for 1 year after their index MDS date. Outcomes included any health care utilization in different care settings (ie, inpatient, outpatient emergency room visit/observational stay, skilled nursing facility, hospice) and corresponding Medicare expenditures. We used difference-in-differences models to estimate the association between ACO attribution and health care utilization in 2018, using 2011 as the pre-ACO baseline. To determine ACO attribution among the 2011 cohort, we developed an algorithm to replicate the ACO attribution in 2018 and used it to identify residents who would have been attributed to 2018 ACOs back in 2011. To address the endogeneity issue between ACO attribution and utilization outcomes, we used an “intention-to-treat” design to determine ACO attribution.

Results

Adjusted difference-in-differences results showed a lack of significant associations between ACO attribution and health care utilization or Medicare expenditures among long-stay NH residents.

Conclusions and Implications

ACOs did not affect health care utilization of long-stay NH residents. Future payment reforms need to ensure that their benefits could reach these vulnerable older adults.