Skip to Main Content

Better health usually isn’t the result of higher-quality health care. Factors outside the current health care system, social determinants like income, education, employment, food security, housing, and social inclusion, generally make a bigger difference — especially in disadvantaged communities. We need to rethink how health care organizations can help their patients stay healthy and out of the hospital by addressing these essential factors.

What’s needed is a full-risk model, one that holds provider organizations fully accountable for the health outcomes of their patients. In this model, practices are paid a fee for each patient and then cover all the costs of caring for that patient, whether it’s an emergency department visit, a hospitalization, a surgery, a medication, or a stay in a skilled nursing facility. Only with this degree of accountability can provider organizations be fully aligned with the interests of their patients and invest in what they truly need.


A full-risk model is daunting for most organizations. Alternatives have emerged to pursue value-based care without taking that plunge, though my colleagues and I believe that full risk is the most direct path to achieving high-quality care at a low cost while also creating incentives to invest in the services that patients need.

Shared savings: the middle ground isn’t enough

The shared-savings model for health care is, at first glance, an appealing financial choice that lets providers progress towards value-based care. In this model, a payer gives a provider organization a fixed budget to care for a group of patients. If the cost of the care provided comes in under budget, the provider shares the savings with the payer. There is, however, no penalty for failing to generate savings. Without this disincentive, a provider organization faces no real pressure to innovate.

Shared-savings models are often considered a good first step towards value-based care, as they appear to be the less-risky option for providers: There’s a chance to win without risk of loss.


But it isn’t enough.

While well-intentioned, shared-savings models have not consistently delivered lower costs. Results from the most common shared savings model for accountable care organizations, the Medicare Shared Savings Program (MSSP), are mixed at best. In its first year, accountable care organizations participating in MSSP experienced 1.4 percent savings; the next cohort experienced no significant savings.

With so much promise, why have the results of shared-savings models been lackluster? It’s simple: In a shared-savings model, providers who are able to lower the cost of care create savings they must then share with the payer. That means the provider makes the full investment and does all the work, but doesn’t get the full benefit. While a shared-savings model may feel less risky, it also reduces the incentive for innovation and investment in patient care.

Consider a hypothetical program that provides low-income patients with subsidized transportation to their primary care appointments. This service could prevent waste by improving access and reducing the number of missed appointments, which in turn would decrease the number of costly hospitalizations. If this program reduced costs, under a shared-savings model the provider would have to turn over some of the savings to its payer partner.

Such sharing, however, could mean that the return doesn’t cover the payer’s investment in the program. So while the program helps patients and saves money, the financial arrangement may prevent the provider from building this obviously advantageous program. A full-risk organization, in contrast, has a strong incentive to invest in patient transportation services, since every dollar the provider saves is a dollar it earns — and keeps.

While a shared-savings model may seem like a conservative approach to value-based care, it doesn’t deliver the same results as a full-risk model, either in theory or in practice.

Health care providers should thrive only if their patients thrive. Otherwise the health care system is rewarded for expense, including waste, that doesn’t help patients. And without full accountability, organizations invest in what is billable, not what is necessary for patients. Innovators must take the plunge and adopt full-risk models rather than half measures. Doing so is less risky in the long run.

Let health care providers capture the value they create

We need health care models that encourage provider organizations to invest in what they know is needed by the communities they serve. It’s time to transition from “primary care practices” to “social determinants practices” — delivery organizations that acknowledge the importance of these factors in creating better health outcomes.

Here’s an example from Oak Street Health, a full-risk network of 40 primary care centers focused on older adults in medically underserved communities that I co-founded in 2012.

At age 67, Ike (not his real name) was turned away from a homeless shelter one cold Chicago afternoon because of a cough. (Ike also had a history of substance abuse, depression, and housing insecurity.) The shelter’s staff members feared he would spread his cold to other residents and told him he couldn’t stay without a physician’s note explaining that he was receiving treatment and was not contagious. Ike fully expected to spend the frigid night outside, or in an emergency department. Fortunately, he called his care team at Oak Street Health, which he had met during the team’s visit to his shelter months earlier.

Given the freedom and accountability of a fully value-based health care organization, we provided Ike with transportation to one of our health centers, made sure he was seen by a clinician, got him the prescription he needed for his cough, documented his plan, and transported him back to the shelter in time for the evening meal.

By investing in care that addressed several social determinants — Ike’s need for transportation, appropriate and timely medical attention, assistance getting medication from our pharmacy, and the like — we were able to avoid a costly emergency department stay or an even costlier hospital visit down the line. In other words, value-based care allowed us to do what was best for Ike and ultimately what was best for the organization.

It’s true that the transition to a full-risk, value-based model of health care is difficult. But the status quo, in which patients suffer under a health care system that can’t properly serve them, is intolerable. Full-risk, value-based systems may not be the best fit for every doctor or health system, but it’s good for patients, and thus must be our goal — as taxpayers, as clinicians, and as humans. For all of us who experience first-hand the daily impact of this kind of accountability on the patients we serve, we share an urgency that isn’t going away soon.

Griffin Myers, M.D., is the chief medical officer of Oak Street Health, which is based in Chicago.

  • So we agree that the bulk of the evidence shows shared savings doesn’t work.
    So why would we double down on this poor bet for full risk models? At the very least, if we are being honest we should only do so after we have built a substantial body of evidence that this works…Please show us that evidence.
    Isn’t it odd that health systems that deliver far better results for far less money don’t rely on these risk models?

  • When managed care is done right it is a win-win-win for patients, providers, and health plans. We have been caring for patients in a risk environment for over 17 years. The key to success is educating providers on population management and always putting patient’s needs first. As we grew our managed care population, our financial risk decreased. Physicians need to understand that in the full-risk model world seeing less patients often translates to increase in revenue and improved patient care. It is hard work and can be very lucrative when done in a compliant, patient centric manner.

  • My first impression was that Dr. Myers is a youngster. HMOs were the rage from the early eighties to the mid-nineties. There was even experimentation with so called social HMOs, which were close to what Dr. Myers is describing.

    The concept and this type of organization largely died for a number of reasons, including one that Dr. Myers alluded to, sharing of savings. Only back then the insurance companies often dumped the risk on the providers without adequate safeguards against underutilization and inadequate compensation. A lot of medical groups lost their shirts.

    Some HMOs such as Kaiser Permanente have survived and flourished but partly because of a long history of providing pre paid or capitated care. Even it’s model has evolved over the last 20 or so years to adapt to the changing marketplace.

    With better information and regulatory oversight, some of the risks of over and underutilization can be mitigated. However, as I see it, the biggest barriers to shifting all the risks, and the rewards to the provider side, are the aging of the population with its attendant explosion in diseases and the need to treat them; and the explosion in new technologies which are usually quite expensive. These two treads alone make it very difficult to predict utilization of services and financial risk to whomever wants to assume it.

    So, I don’t see Dr Myers’ concepts taking hold anytime soon.

  • I think you must still remember what was learned from the capitated HMO models of the past. Some provider organizations denied or delayed care making patients jump through hoops. I believe safeguards must be in place and reimbursement must be tied to patient outcomes as well. Much as I hate to say it, not all medical providers are altruistic.

Comments are closed.