Insights on Site-of-Care Cancer Research: Both Quality and Cost Information Are Necessary to Guide Policy
Kavita Patel, MD, MPH, and A. Mark Fendrick, MD
Hospital mergers and consolidations among oncology practices are considered to be key contributors to recent increases in the utilization of and spending on cancer-related care. Different payment methodologies, often based on the site of service, are critical driving forces behind this phenomenon. In an attempt to slow the unsustainable rate of increases in healthcare expenditures, policy makers have implemented strategies to equalize such payments. One notable example, the 21st Century Cures Act signed by President Barack Obama in 2016, mandates site-neutral payments for new facilities in the Medicare program.1
Even with the popularity of the 21st Century Cures Act and similar policy efforts, there is an insufficient amount of empirical research exploring the impact of different sites of service on Medicare expenditures and the association of spending levels with patient-centered outcomes. This dearth of rigorous study is particularly evident for patients with a diagnosis of cancer who are receiving chemotherapy. In this issue of The American Journal of Managed Care® (AJMC®), Kalidindi and colleagues provide new insights on site-of-care spending that have important implications for the Medicare program and, specifically, value-based specialty payment programs, such as the Oncology Care Model (OCM).2
Their 4-year analysis of Medicare beneficiaries receiving chemotherapy found that risk-adjusted chemotherapy drug spending per beneficiary was $2451 lower in hospital outpatient departments (HOPDs) compared with those whose chemotherapy was provided in physician offices. This finding, driven primarily by the fact that patients in physician offices received chemotherapy drugs more frequently than those in HOPDs, differs from that of an often-cited analysis that found increased Medicare spending in hospital-based facilities.3 Kalidindi et al attributed the differences in the results to patient risk factors, particularly cancer type. The authors found that once such clinical differences were adequately accounted for, spending on chemotherapy drugs was, in fact, lower in hospital-based settings.
The implications of this AJMC®paper and other site-of-care research are increasingly timely as the Trump administration considers changing reimbursement for physician-administered drugs in Medicare Part B, which would include most chemotherapeutic agents.4 Such a change in payment for Part B drugs could have significant potential downstream effects on utilization, overall cost, and patient-centered outcomes. Moreover, the interpretation of the results of site-of-care research may have even more far-reaching consequences as pressure grows to extend site-neutral payment policies from new to existing facilities. (The 21st Century Cures Act, significantly, applies only to new facilities.)
It is important to note that the data used by Kalidindi et al were collected prior to the launch of the OCM, Medicare’s first large-scale oncology value-based payment initiative. Launched in July 2016, the OCM includes 187 practices, mostly community-based, that deliver care to 25% of the total Medicare beneficiaries receiving cancer care. The OCM provides monthly payments per Medicare beneficiary for comprehensive care while patients are receiving chemotherapy treatments, a patient population similar to that studied by Kalidindi and colleagues. The OCM also includes a novel therapy adjustment to attempt to account for the emergent use of innovative cancer drugs in Medicare parts B and D.
Early findings from the OCM have demonstrated signs of progress, notably that most practices are engaged in significant clinical transformation activities, such as after-hours access and extended patient support services. However, it has yet to be determined how this payment model might alter overall spending patterns, particularly for community-based practices that are independent of hospital affiliation.5
Through the OCM and other programs, the Center for Medicare and Medicaid Innovation is making a large investment in the premise that value-based payments will encourage better quality at a lower cost for all patients, independent of the site of care. Although this hypothesis has yet to be confirmed in oncology practice, predecessors using alternative payment models in the commercial setting have performed well, particularly on metrics related to improved care coordination. However, a single universal payment model that applies broadly to the entire spectrum of cancer care has yet to be identified. In fact, when one large commercial insurer engaged in a value-based oncology care pilot with community-based practices, spending on chemotherapy drugs increased, but aggregate spending decreased.6 This finding highlights that, in many situations, payment incentives might not directly correlate with a decrease in expenditures on certain aspects of care, such as spending on drugs. Likewise, financial savings do not (and should not) imply improvements in quality. It is important to note that the site-of-care analysis by Kalidindi et al did not assess the relationship between spending levels and quality of care.
An additional important takeaway from the growing (and sometimes conflicting) literature on site-of-care and other value-based initiatives is the need for robust risk adjustment. Because this requirement is heightened in a patient population with a cancer diagnosis, Kalidindi et al included distinct control variables, such as cancer type and presence or absence of metastatic disease, in their analysis. Similarly, in the hopes of better enhancing risk adjustments, present-day value-based models, such as the OCM, are also collecting detailed clinical data. Ultimately, the refinement of these factors leading to better risk-adjustment methods will have a dramatic impact on our understanding of the clinical and financial influence of past and future policy decisions for the Medicare program and commercial payers.
There is a growing sense of urgency to move away from fee-for-service compensation and instead align payment with quality-driven measures. Although consensus for this change exists at a high level, the details on how best to move from volume-based to value-based reimbursement are fraught with political, social, and administrative challenges. As the findings of Kalidindi et al and several other publications illustrate, most acute and chronic illnesses are extremely complex and often clinically heterogeneous, making direct comparisons difficult. The implementation of alternative payment models that successfully capture clinical heterogeneity, without adding unacceptable levels of administrative complexity, may be equally or more important than site-neutral payment policies. A robust evidence base concludes that patient-centered outcomes may be impacted by where care is provided, especially in those clinical scenarios where results are sensitive to physician or health system experience. Notwithstanding, until enhanced data collection and risk-adjustment techniques become available, the ability to associate site-based expenditures with quality of care and the patient experience remains an important work in progress.