Increasing the Longevity of Accountability, Designing Cooperative Models, and Transferring Risk Across a Lifetime of Accountable Care

We have a lifelong relationship with our health. We depend on others to help finance our health risks through insurance, doctors to advise and treat us, drug companies to create treatments, and hospitals to house and care for us when we are in our worst health. Accountability is both an individual and a shared responsibility. Many physicians, nurses, other clinicians, and even employees of health care organizations feel a sense of responsibility when a patient’s health deteriorates. Was something missed? Could we have reached out more effectively? Could we have known something sooner? Would a stronger relationship with the person have made a difference? This professional accountability has always been at the heart of health care as a profession.¹ Starting in 2006, a refined concept of tying this professional accountability to financial accountability emerged: the accountable care organization (ACO).² Since the 2012 launch of the Medicare Shared Savings Program (MSSP), the flagship ACO program, concepts have turned into operations and spread widely. Yet Medicare, both traditional and Medicare Advantage, has continued to draw most of the attention and has nearly twice as much participation in ACOs.³

Stability Presents Incentives for Long-Term Alignment

The Medicare program holds a unique advantage in accountable care. Once a person is on Medicare, they are likely to remain on it for the rest of their life. This is in stark contrast to the commercial and Medicaid spaces, where eligibility fluctuates. People become eligible, ineligible, and sometimes eligible again for Medicaid. If commercially insured, they may change employers, move, and or have their children age off family plans. This dynamic nature of coverage in non-Medicare spaces creates unique challenges in accountable care.

In many cases, a person’s relationship with their physician will last longer than their length of coverage with an insurer. In other cases, changing coverage may require a change in physician. This churn creates unique challenges in accountable care outside of Medicare.

Population health efforts are inherently long-term, often taking years to bear fruit. For instance, managing blood pressure is a lifelong endeavor that yields cost savings only when acute events are prevented. Work is done every year, but the savings are only realized when an event is avoided. The longevity of a person’s coverage with an insurer can also affect the distribution of these savings, making longevity a powerful force in accountable care.

Accountable care seeks to create financial alignment between primary care and health as expressed by the need for health care services. I will focus on the relationship between a person and their primary care physician. (Although some organizations are making efforts to create primary care relationships between a person and an organization rather than a person and a clinician, time will tell how that works out. Today, nearly all Americans define their primary care relationship as one with an individual clinician, and most total cost of care payments go to ACOs.) The patient-physician relationship has turnover, too. Patients move, physicians retire, patients choose to switch. Yet the turnover between physicians and patients could be much less, particularly the turnover driven by insurance coverage. A physician can decide to accept a wide range of insurance types, allowing the accountable relationship to endure even when the patient changes coverage.

Understanding and Targeting Risk to Achieve Longer Accountable Relationships

The present environment around accountable and primary care does not take advantage of this opportunity. There is an urgent need for a tool or model that can create a bridge from one insurer to another. The key metric in any ACO is expected costs. Three commonly used elements to generate expected costs are a population’s historical costs, average costs, and cumulative risk scores. From the insurer’s perspective, all 3 change when a person moves from one insurer to another. Even though the population changes, the premiums change, the benefit design changes, the preferred risk model may change, and nearly everything around the person and the primary care physician changes, the person and the primary care physician do not.

Risk breaks down into 4 categories: insurance, program, individual, and performance. I describe insurance risk as the primary reason people have health coverage. These are the accidents, the unavoidable cancers, the genetic traits lying dormant, the infectious diseases contracted from asymptomatic people—essentially all the things that neither a person nor the primary care physician can influence. Program risk is variability in premium pricing, benefit design, the administrative side of quality reporting, and everything else the organization providing health insurance coverage decides on. Individual risk is the choices made by the person. Do they engage in high-risk behaviors, such as riding motorcycles or skipping their medications, thereby increasing their insurance risk? Society has largely decided that it prefers community rating systems that do not use many variables from the individual risk side. Finally, performance risk encompasses the use of and performance by health care providers. Primary care is at the center of performance risk. It is this centrality that puts primary care and accountable care hand in hand.

ACO model design should isolate performance risk as much as possible to achieve longer accountable relationships. Success in an ACO should be a substantial part of a physician’s revenue. In that case, it can’t go from significant to zero just because a person changed insurers. So how do we create accountability that does not reset to zero every time any factor changes? Two future-looking possibilities are improving predictive risk models and instituting a similarly designed ACO model for each payer.

Having a predictive risk model where the current state can narrowly and accurately predict future performance risk would be ideal. Many organizations are working to improve existing risk models. A perfect model would use existing information generated through the course of medical record keeping and administrative claims, be insensitive to efforts to influence the model results, break out the differences in performance risk and insurance risk, and be universal across payers. Efforts around improving risk models and developing new ones could lead to breakthroughs, allowing risk models to play a more prominent role in developing expected costs for ACOs across payers. The potential for these breakthroughs should inspire optimism about the future of ACOs.

From the payer’s perspective, program risk isn’t risk. It is the primary tool for the payer’s financial success. From the primary care physician’s perspective, program risk is the riskiest because decisions over which they have no influence affect the bottom line. Insurance risk can be modeled and forecasted, but program risk can come out of the blue. The ability to separate program risk from performance risk is valuable to payers and ACOs. ACOs do not want their revenue to drop because of decisions made in a payer C-suite, and payers do not want to pay ACOs for accrued savings that are due to the payer’s benefit design. As with risk modeling, there is still much work to do to reach an ideal ACO design. Closer to our grasp is more commonality among ACO models from payer to payer,⁴ creating stability when the payer-to-person relationship changes but the physician-to-person relationship is maintained. The main element of a common model is the recognition that an ACO model cannot start over at zero every time the person changes insurance coverage. Whether this takes the form of expected costs against a market average adjusted for program risk, a common risk adjustment model, or a combination, any model that can survive a transition from one payer to another must recognize past achievements in performance risk in a significant way.

Conclusions

Increasing the longevity of accountable care relationships is key to shifting the finances of health care from set rates for services to financial alignment around health and performance. Cooperation in the design of ACO models, commonality in how the different types of risk are transferred or more often not transferred to ACOs, and the constant improvement of our predictive risk models will lead to accountable care that flourishes in the multipayer system we have in America. It will be difficult because everyone’s best design will be slightly different from everyone else’s best design. Financial alignment will never be perfect in any singular year. Yet the benefits of everyone moving closer together to a model that creates sustainable revenue based on primary care outcomes can last a lifetime.

Author Information: Mr Broome is senior vice president for policy and economics at Aledade in Asheville, North Carolina, and a member of the editorial board of Population Health, Equity & Outcomes. The opinions expressed in this article are those of Mr Broome and not necessarily Aledade.

REFERENCES

1. McWilliams JM. Professionalism revealed: rethinking quality improvement in the wake of a pandemic. NEJM Catal Innov Care Deliv. 2020;1(5). doi:10.1056/CAT.20.0226

2. Tu T, Muhlestein D, Kocot SL, White R. The Impact of Accountable Care: Origins and Future of Accountable Care Organizations. Leavitt Partners. May 2015. Accessed November 7, 2024. https://www.brookings.edu/wp-content/uploads/2016/06/impact-of-accountable-careorigins-052015.pdf

3. 2023 APM measurement. Health Care Payment Learning & Action Network. Accessed November 7, 2024. https://hcp-lan.org/apm-measurement-effort/2023-apm/2023-infographic/

4. McNulty R. AHIP, AMA, NAACOS playbook recommends best practices for sustainable value-based care. AJMC^®. April 11, 2024. Accessed November 7, 2024. https://www.ajmc.com/view/ahip-ama-naacos-playbook-recommends-best-practices-for-sustainable-value-based-care