Abstract

Objectives: To use a theoretical framework based in transaction cost economics/ relational contracting to draw insights from mixed-methods data concerning the relationship of post-acute care (PAC) providers with Medicare accountable care organizations (ACOs); and to consider implications of the Medicare ACO model for long-term services and supports (LTSS) providers and populations.

Background: In the US, ACOs represent a new organizational form charged with containing total Medicare expenditures for a population of assigned beneficiaries. Medicare covers skilled health services provided after acute hospitalization by skilled nursing facilities (SNFs) and home health (HH) agencies; PAC expenditures have grown rapidly and vary substantially. Although nursing homes and HH agencies serve LTSS populations along with PAC patients, LTSS is not covered by Medicare. Medicare ACOs have the potential to contain PAC costs through steering discharged beneficiaries to more efficient PAC providers, improving care transitions to reduce rehospitalization, and increasing the efficiency of PAC partners. Further gains might be achieved through collaboration with LTSS providers to coordinate services for assigned beneficiaries with functional needs.

Methods: This study used qualitative and quantitative data on PAC providers for three market areas. Structured interviews revealed how ACOs contract with providers. Claims data for fee-for-service Medicare beneficiaries were used to investigate changes in length of stay, rehospitalization, and concentration of utilization for HH and SNF sectors by market for 2012 through 2014, for all and for ACO-assigned beneficiaries.

Results ACOs manage PAC utilization by sending discharged hospital patients to providers they own (vertical integration) or by developing continuing negotiated relationships (relational contracting) with providers; or they continue standing patterns of hospital-PAC referrals. ACO hospital discharges to HH can be served by one system-owned provider, because that HH agency’s personnel travel to cover the market area. In contrast, discharges to SNFs must be distributed over independent facilities near beneficiaries’ homes, encouraging ACOs to develop relationships with multiple SNFs. Some HH agencies reported coordinated admissions and system investment in information technology (IT). SNF respondents described levels of relational contracting, including offers of “preferred” status, promising increased – or no reduction in – admissions in exchange for alignment with ACO goals; ACO or SNF investment in assets specific to the relationship, including IT and staff training; and direct clinical management by ACO personnel. ACO reporting requirements and direct case management could be burdensome, and some PAC providers doubted that gains from the relationship exceeded costs. Both SNFs and HH agencies described investments for smoother care transitions, but felt that shorter SNF stays burdened patients and family caregivers. Because SNFs rely on Medicare revenue, reduced expenditures may constrain access for LTSS residents. No PAC provider was sharing gains of PAC containment with an ACO.

Conclusions: Provision of PAC services has become more concentrated in the HH and SNF sectors in three market areas, and utilization is being constrained. Although aligning PAC providers with ACO goals can move Medicare toward more coordinated and integrated care for beneficiaries, implementation is still a work in progress and is not yet supported by mutual gains.