https://www.ajmc.com/journals/evidence-based-oncology/2019/june-2019/twosided-risk-in-the-oncology-care-model
Two-Sided Risk in the Oncology Care Model

Kashyap Patel, MD, ABOIM, BCMAS; Maharshi Patel, MBA; Taylor Lavender, BS, PA; Dhwani Mehta, MS, RD; Asutosh Gor, MD, Sashi Naidu, MD; and Chuck Newton, BS

The US healthcare system remains one of the most inefficient healthcare systems in the world. The Bloomberg Health-Care Efficiency Index ranked the United States 54th among 56 countries in 2018, tied with Azerbaijan and only ahead of Bulgaria.1 This occurs even though the United States spends $10,244 per capita annually on healthcare, a figure representing 17% of the gross domestic product.2

Our expensive yet inefficient healthcare system has been blamed on a fragmented, disorganized, and uncoordinated delivery system, with silos and redundancies that create inefficiency.3 Despite rapid advancements in treatment, the discovery of new drugs, and new technology aimed at improving patient outcomes, the overall performance of the US healthcare system in aligning incentives has not met expectations.

Although these inefficiencies have become evermore glaring, oncology care is advancing at a rapid pace; improvements in survival rates in many types of cancer have led to more than 15.5 million US cancer survivors as of January 2016.4 In 2018, the FDA approved 19 new cancer drugs and biologics, and 38 new indications,5 as well as new molecular tests and companion diagnostics that will ultimately scale personalized treatment in oncology. 

Reports from the National Academy of Medicine, formerly the Institute of Medicine, concluded that cancer treatment in the United States lacks in consistent quality and is neither patient-centric nor well-coordinated.3,6 The fee-for-service (FFS) payment model, used for decades in oncology and elsewhere in the United States, was not designed emphasize value or quality in cancer care. Under FFS, medical services are not bundled; they are paid for individually, thus incentivizing the provision of high-quantity, not necessarily high-quality, healthcare. The rapid, but hurried and disorganized, multidirectional advancements in payment models, such as pathways and capitations, were never formulated to affect the quality of care under the volume-driven model. Hence, despite many efforts to explore improving outcomes by using pathways and capitated models, outside of scant models demonstrating desired outcomes, most capitated models did not fulfill payers’ expectations.7-9 

In the United States, total spending on cancer care rose from $27 billion in 1990 to $124 billion in 2010, with projections of around $157 billion by 2020.10,11 Total costs of cancer care for the US population are predicted to increase across all phases of care.12 Cost drivers include technological innovation, rising hospital costs, and demographic shifts; as the population ages and people live longer, the risk of malignancy rises.13 In the United States, oncology drug expenditures, excluding supportive care agents, rose 18% from 2014 to 2015, one of the largest single-year increases on record, driven by both the first full year of implementation of the Affordable Care Act (ACA) and the arrival of programmed cell death protein-1 inhibitors.14 

Under the ACA, CMS established the Center for Medicare & Medicaid Innovation (CMMI) to test innovative payment models to incorporate the value element in the delivery of healthcare. CMS has developed value-based care (VBC) programs that reward healthcare providers with incentive payments for improving the quality of care, reducing costs, and improving the patient experience for Medicare beneficiaries. In October 2016, CMS finalized the rule for the Medicare Access and CHIP Reauthorization Act (MACRA) (Figure 1) of 201515 that implemented the Quality Payment Program (QPP) (Figure 2).16,17 An underlying tenant of VBC is to move away from the FFS model and toward performance-based payments. These programs tie payments to provider performance based on meeting specified quality metrics and practice reforms, with some practices already entering into payment arrangements that include financial and performance accountability for episodes of care involving chemotherapy administration to patients with cancer. 

The QPP began in January 2017, with payment adjustments based on performance (to be fully implemented by January 2019). The QPP offers payment according to 1 of 2 tracks: (1) the Merit-based Incentive Payment System (MIPS) linked to performance including following defined, evidence-based clinical quality measures, and (2) Advanced Alternative Payment Models (AAPMs) that provide financial incentives to clinicians to provide high-quality and cost-efficient care (Figure 2).16,17 

The Oncology Care Model 
The Oncology Care Model (OCM), which launched on July 1, 2016, is one pilot program that is intended to shift reimbursement away from volume and tie payments to value under the broader umbrella of transitioning to VBC. The OCM is a cancer care delivery model that encourages participating providers and practices to align financial incentives to improve care, add the quality component, and enhance the patient experience while reducing the costs of care. The OCM with 2-sided risk is considered an AAPM.18,19 It is a voluntary pilot and a part of CMS’s broader initiative to improve the effectiveness and efficiency of cancer care.20 This program aims to provide higher-quality, more coordinated oncology care at the same or lower cost to Medicare than traditional FFS payments. When the program was announced and CMMI invited requests for applications in the spring of 2015, more than 400 oncology practices from across the country submitted letters of intent to participate in the 5-year pilot (2017-2022). Overall, 196 practices (covering 3200 oncologists; approximately 20% of all practicing oncologists from different settings) were selected to participate in the OCM. In addition, CMS encouraged private commercial payers to participate; 17 payers initially signed up, and 16 national and regional payers began implementing the OCM in 2017. At the time of launch in July 2017, 192 practices and 14 commercial payers were participating in the OCM.20 

The OCM and the QPP have differences that are significant to oncologists (Figure 3). The OCM incorporates a 2-part payment system for physician practices: a per-beneficiary Monthly Enhanced Oncology Services (MEOS) payment and a performance-based incentive payment (PBP) (Figure 4). The MEOS payment allows practices to improve care coordination by effectively managing and coordinating episodes of care for patients with cancer. The PBP is calculated retrospectively on a semiannual basis, based on the practice’s achievements in quality measures and reductions in Medicare expenditures. 

Fundamental Tenets of OCM 
OCM providers are required to:  The multipayer OCM pilot’s early feedback has provided a blueprint for not only quality improvement and payments linked to quality, but also for the way physicians think about and deliver care. Under the OCM, patients, caregivers, and their microecosystems become the central force that directs care in accordance with patient input, preference, and choice. The OCM has the potential to reduce healthcare costs while improving patient quality of life and outcomes. 

CMMI has been reimbursing successful practices that have earned a PBP in the form of gain sharing to reward high performance and reduce the cost of care to below the target price. CMS will provide a discount of 4.0% for practices participating in the 1-sided risk model and 2.75% for practices participating in the 2-sided risk model. The savings offered will be adjusted by performance based on quality reporting. Participating practices are subject to monitoring, including on-site visits, by the CMMI team. CMMI has appropriately adjusted the risk factors to accommodate multiple comorbidities, dual-eligible status, surgical intervention, radiation therapy, and clinical trial participation. 

The OCM is the first major initiative by CMS to pilot the transition from FFS toward VBC. CMMI has been relatively flexible and accommodating of changes made to the program in response to stakeholder input. However, areas of concern remain, including the main issue of the negative penalty for exceeding the target price of care. 

Oncologists participating in the OCM are held responsible for the total cost of care, regardless of the origin of that cost. For example, if a patient under the OCM pilot is involved in a motor vehicle accident and requires hundreds of thousands of dollars’ worth of interventions, the cost of this care will be part of the bundle that the oncologist will be responsible for. However, according to the Winsorization formula—which identifies Medicare beneficiaries outside of statistical variables, which limits its practice attributes—risk exposure will be capped at 95%. 

Many lessons remain to be learned by oncologists participating in the OCM. This is compounded by the recent announcement of the MACRA final rule, where a negative risk exposure of an oncologist is limited to 3%, based on resource utilization, at the most versus the 20% risk exposure in the OCM.9  

Conversely, the upside bonus for top performers in the QPP track under the MACRA final rule could exceed 20%. This may become further complicated, given the uncertainty surrounding the 2017 change in the federal administration and its healthcare agendas. 

In terms of pure comparison, upon the launch of the OCM pilot, it appeared that OCM with 1-sided risk was the best option for oncologists, followed by MIPS and then OCM with 2-sided risk. However, acknowledging an extremely high downside risk— exposing a practice to almost 20% risk, amounting to in excess of $250,000 per provider—the original 2-sided risk parameters provided a model that could cause significant financial hardship to practices that fell on the wrong side of the model. That is why none of the participating practices opted for original 2-sided risk and why they instead remained under the no-risk model. As mentioned before, the upside potential for an exceptional performer practice or provider is significantly higher. This also contributed to a business case for providers to not opt for the original 2-sided risk model. Recognizing these limitations of 2-sided risks, CMMI altered parameters to limit risk and make 2-sided risk a more attractive proposition. However, CMS subsequently modified the 2-sided risk model, and the new alternative risk model has significantly less financial risk for the practice than the previous +/–20% approach. In this modified risk model, risk is based upon expenses attributed to a provider and not the total cost of care per episode (Figure 5, Tables 1-2). 

Although the metrics for no risk still are the most favorable for most practices, if a practice fails in the performance category for 3 consecutive periods, that practice is forced to accept the alternate model. Given the more favorable parameters of the alternative risk model, it is worth considering remaining in the OCM under 2-sided risk rather than leaving the model. 

Summary of Risk OCM Models and Financial Impact 
The 1-sided risk model allows for a practice to eliminate the financial challenges associated with downside risk. These practices may not qualify for the designation of AAPM and hence will not be eligible for the 5% AAPM bonus. They will also be subject to following the QPP requirement. However, by nature of the practice transformation process, these practices are already compliant with the majority of quality reporting, as well as other QPP requirements, and they will likely earn incremental bonuses up to 9% by 2022. Thus, a practice that is very efficient, has achieved a PBP, and has fulfilled all requirements is best suited to remain in the 1-sided risk model and optimize its performance in the QPP track. For those practices and providers considering 2-sided risk, the original 2-sided model provided the highest potential revenue for a top performer. However, the difference in top gains versus no risk would be less than $100,000, versus the difference in top losses versus no risk of almost $1 million. This high-risk profile is what led to no adoption of the original 2-sided risk model. 

The practices that fail to achieve PBPs for 3 successive episodes will either need to shift to the 2-sided risk model or drop out of the OCM. In our analysis (hypothetical, based upon our own case study), the numbers show that the new 2-sided risk model has a more favorable maximum loss than the no-risk model, although gains are approximately half of what they would be under the no-risk model. As such, this new 2-sided risk model is a financially sound model for a typical physicians’ office. Although the no-risk model remains the preferred OCM reimbursement model for most practices, the new 2-sided risk model is one that may not be financially toxic like its predecessor. As such, for practices without another option, it is still a viable model under which a practice can be a participant in the OCM and experience significant financial gains through its participation. The new 2-sided risk model is significantly more favorable than the previous model. In particular, for practices that have failed the 3-performance-periods test for the no-risk model, it may make sense to remain in the OCM under the new 2-sided model. This approach makes sense, since the financial downside can be mitigated by obtaining reinsurance, the cost of which can be offset by the gain in revenue through MEOS payments (Table 3, Figure 6). 

Protecting Downside Risk With Reinsurance 
There are 2 types of reinsurance concepts that a provider needs to understand when accepting downside risk. These concepts are: (1) severity risk and (2) frequency risk. 

Severity risk is the impact of a single large claim on the risk pool. For example, Medicaid plans worry about an infant in the newborn intensive care unit or a patient with hemophilia with medical costs exceeding $1 million. 

Frequency risk is the impact of many low-level claims on your risk pool. For example, a year with widespread flu cases can cause an increase in hospitalizations and associated treatments. No one treatment is costly, but the increase in utilization can drastically affect the risk pool. 

Severity risk is mitigated by what is called specific reinsurance and frequency risk is mitigated by aggregate reinsurance. 

Conclusions 
CMMI has paved the path for the transition from volume to value in oncology. Although concerns remain to be addressed, at least one-third of OCM practices have reached some type of PBP and accessed shared savings. It is possible to improve the patient experience and reduce cost of care while adhering to guidelines based on standard-of-care treatment. Some of the areas of cost improvement include expanded access, use of biosimilars and generic drugs, and following recommendations by the national professional societies (eg, the American Society of Hematology, the American Society of Clinical Oncology). Visits to the emergency department, reduced hospitalization, use of biosimilars and generics, and expanded access are low-hanging fruit for success, all of which can help practices optimize their performance in the OCM and/or newer payment models, even if they must adopt 2-sided risk. 

We believe that realigning the workforce to provide patient-centered care will enhance the team-based approach, improve employee morale, improve the patient experience and care coordination, and ultimately lead to true value in healthcare. Human potential is frequently ignored and undervalued, but capturing it is a highly rewarding step for success in any task, no matter what challenges exist. By properly training and incentivizing the workforce around patient-centered care, practices can provide a recipe for their success in the OCM. 

From CMMI’s perspective, it may be worth considering adjustments based on regional and/or socioeconomic and demographic risk factors. Factors such as having a rural versus a suburban location or access to urgent care can affect outcomes. Such an alternative would reduce disparities in performance that are beyond an individual provider’s control. 

Based on our experience, transitioning to VBC is possible. Given the newer improved risk profile in the 2-sided risk model for the practices that must take 2-sided risk, it is worth considering by identifying the possibility of reinsurance, either by sharing part of upside potentials or sharing part of the MEOS payments. 
Print | AJMC Printing...