Abstract
Introduction
Faced with growing populations of older, medically complex patients, health systems are now incentivized to deliver cost-effective, high-value care. We evaluated a new method that builds upon existing Medicare spending concentration studies to further segment these expenditures, revealing use patterns to inform care redesign.
Methods
We obtained monthly Medicare expenditure data and derived baseline comparison data using typical methods for identifying a yearly high-cost subpopulation. We then applied the new methodology, ordering monthly patient expenditures from highest to lowest to more extensively segment the baseline data. Our evaluation examined the following within the new more extensive segmentation: monthly expenditure distribution, corresponding patient counts, and occupancy of specific patient subgroups within the extended segmentation of baseline data.
Results
Compared to the baseline data, we found further spending concentration, with 16.7 percent of high-cost patients being responsible for about two-thirds of baseline expenditures. The remaining 83.3 percent of the high-cost subpopulation exhibited lower spending, collectively accounting for about one third of baseline expenditures. Additionally, we found that unique patient subgroups occupied different segments over time, with specific subgroups comprising 8.3 percent of the study subpopulation patients migrating into and out of each highest spending segment, accounting for almost half of monthly baseline expenditures.
Conclusions
With monthly health care expenditures concentrated among small numbers of migrating patients, our evaluation suggested potential cost-effectiveness in tiered care delivery models, where small subgroups receive direct, active care interactions, while the remainder experience surveillance-level care, designed to both address ongoing medical needs and to detect emergent migration.
Introduction
Throughout much of the world, health systems are facing challenges associated with growing populations of older, more medically complex patients [1]. In the United States (US), for example, the number of Americans age 65 and older is expected to rise to almost 100 million by 2060, ultimately representing as much as one quarter of the total population [2]. Because budgetary resources are often limited, it is necessary to deliver cost-effective, high-value care for this aging population. Within the Medicare health insurance program, which covers older Americans, there are imperatives to shift away from a legacy of expensive volume-based incentives to support more value-based care delivery [3]. Studies of Medicare expenditures can shed light on utilization patterns intrinsic to people in the U.S. 65 years and older, as well as providing context for demonstrations of comparative cost-effectiveness. Such characterizations could specifically inform approaches for designing health systems to more efficiently meet the needs of America’s growing population of older adults.
Historically, studies of Medicare expenditures have highlighted spending concentration, in which high-cost beneficiaries disproportionately drive overall yearly spending [4, 5, 6, 7, 8]. These well-established concentration patterns have been used as a basis for care redesign. Specifically, providers have considered potentially more efficient care delivery models that target high-cost beneficiaries [9, 10] by adding infrastructure [11] and/or human resources in the form of care coordinators or disease managers [12]. Results from numerous programs including an extensive series of Medicare demonstrations indicate that these additional resources were sometimes effective at reducing high-cost services (e.g. hospitalizations), provided they developed substantial, direct interactions between care coordinators, patients and their physicians [12]. However, most of these programs were not cost-effective overall, after accounting for the fees associated with providing the additional resources [12]. Thus, more extensive, in-depth characterizations of Medicare spending may inform care redesign by providing a clearer basis for targeting of care coordination and other resources.
In several recent studies, researchers identified more extensive segmentations of expenditures for high-cost Medicare beneficiaries [13, 14, 15]. Based mainly on clinical criteria, certain patient subgroups (e.g., characterized by frailty) identified from within the top 10 percent of Medicare spenders were found to occupy distinctly higher spending segments [13, 14, 15]. In the present study, we develop more extensive spending segmentation for these same high-cost beneficiaries, but we do not attempt to delve further into clinical considerations. Instead, we evaluate a new segmentation method that indicates the extent to which certain Medicare patient subgroups have high monthly health care expenditures and who often migrate in and out of the highest spending segments on a monthly basis. The hypothesis underlying our investigation was that indiscriminately or persistently providing care coordination or other resources—even within the top 10 percent of Medicare spenders—could inadequately target care delivery to the subgroups who may migrate at any time to occupy the highest spending segments. By quantifying attributes of monthly spending migration for these patient subgroups, the new approach could complement existing clinical segmentation methodologies and inform better targeted care, delivered when and where it is most needed. Although initially developed using Medicare data, this new methodology can be adapted to evaluate expenditure data from other programs facing similar demographic trends and challenges.