Accountable Care Organizations (ACOs) are composed of healthcare providers who assume responsibility for the cost and quality of all medical care and outcomes for a defined population. As of 2024, ACOs in the US covered 13.7 million Medicare beneficiaries [1], serving nearly half of all people with Traditional Medicare [2]. The overall Medicare population includes >50% of incident cancer cases in the US [3]. Moreover, cancer care accounts for a large proportion of total Medicare spending with 1 in 12 Medicare fee-for-service (FFS) dollars spent on cancer care [4]. Yet ACOs have not meaningfully impacted cancer care [[5], [6], [7]]. Understanding the factors contributing to ACOs lack of impact on cancer care coordination and cost reduction is critical for improving cancer care.

ACOs aim to improve healthcare outcomes while reducing costs through coordinated healthcare delivery and value-based payment systems [8]. Traditionally, ACOs have operated in the realm of primary care, despite specialty care driving over 50% of medical costs [9]. However, ACO contracts incentivize care-coordination and quality, which could lead to savings on costs of cancer care. Emerging evidence suggests that ACOs have not had a meaningful effect on spending, utilization, quality of care and outcomes for patients diagnosed with cancer [[5], [6], [7],10].

Seeking clinical care outside patients' assigned ACO may limit the ACOs impact on quality and cost, but this has not been evaluated among a cancer population. Counterintuitively, patients are generally unaware of their assignment within an ACO. Any provider outside of the patient's primary care physician's network is not part of the ACO, but still may be in network with the patient's insurer. ACO leakage refers to the receipt of healthcare by an ACO-assigned patient from an institution outside the assigned ACO network. ACO leakage for specialist visits is known to be prevalent, ranging from 61 to 72% across Medicare Shared Savings Program (MSSP) cohorts from 2010 to 2014 [11,12]. Such leakage directly results from patients' unrestricted choice of providers [11], and may also be a by-product of physicians referring patients with complex illnesses and elevated financial risk to tertiary/quaternary institutions [13]. Leakage appears to be highest when patients have high risk profiles and networks have low primary care physician-to-specialty ratios [11,12,14]. Leakage negatively impacts effective care coordination because the ACO remains financially responsible for the patient's cost of care but is not able to intervene to control costs when patients seek care externally. ACOs attempts to curtail leakage through controlling or internalizing specialty referrals appear to have had negligible impact [11].

We hypothesized that cancer-directed care services represent an important source of ACO leakage. Better understanding of the patterns of ACO leakage may inform strategies to encourage cancer patients to seek specialty care within the assigned ACO. In this study, we describe the overall characteristics of gynecologic cancer patients with stable ACO assignment from 2013 to 2017 using Medicare data. Further, we quantify incidence of ACO leakage, overall, by cancer type and cancer therapy type, among gynecologic cancer patients. We also identify factors, both patient- and ACO-level characteristics that are associated with ACO leakage. Lastly, we describe differences in healthcare costs by ACO leakage status among gynecologic cancer patients.