Spending Changes After Moving to Areas With Greater ACO Participation Among Nonattributed Medicare Beneficiaries.

Importance: Accountable care organizations (ACOs) under the Medicare Shared Savings Program have long been envisioned as a pathway to improved care efficiency and quality of care for all Medicare beneficiaries. However, little is known about whether the changes in health care spending associated with ACOs have extended beyond ACO-attributed Medicare beneficiaries to all Medicare beneficiaries. Objective: To estimate spending changes by non-ACO-attributed Medicare beneficiaries after moving to geographic areas with greater ACO participation. Design, Setting, and Participants: This repeated cross-sectional study analyzed claims from a 20% representative sample of all Medicare beneficiaries, aged 65 to 99 years, from 2009 to 2017. The sample consisted of Medicare beneficiaries who were never attributed to an ACO and moved once across hospital service areas (HSAs) during the study period (movers) and was supplemented by a 20% random sample of beneficiaries who never moved (nonmovers). Data analysis took place from November 2022 to October 2024. Exposure: Changes in the ACO market penetration rate triggered by nonattributed Medicare beneficiaries moving across HSAs. Main Outcomes and Measures: Annual standardized Medicare spending per beneficiary on acute inpatient, outpatient facility, physician services, and total acute care as well as on hospital outpatient department, evaluation and management, and nonadmitted emergency department visits. Results: The estimation sample included 62 618 movers (388 263 beneficiary-years; mean [SD] age, 75 [7] years; 134 503 [65%] female-years) and 433 298 nonmovers (2 066 404 beneficiary-years; mean [SD] age, 76 [8] years; 1 273 154 [62%] female-years). In the base model, moving into a market with a 1-SD higher ACO penetration was associated with a 5.8% (95% CI, 4.1%-7.4%) decrease in spending on outpatient facilities and with a 1.6% (95% CI, 0.9% to 2.2%) increase in spending on physician services. Limited changes in total spending on acute inpatient and total acute care were found. These estimated changes were robust to controlling for a rich set of patient- and market-level characteristics and origin-destination HSA combinations. Conclusions and Relevance: This study provided novel evidence of market-level ACO spillovers to non-ACO-attributed Medicare beneficiaries. Although no substantial ACO spillovers in total acute care spending occurred, the substitution in spending on outpatient facility and physician services suggested that outpatient care may shift away from higher-cost facility settings for all Medicare beneficiaries in markets with greater ACO penetration.