In theory, everyone wins when you create an accountable care organization. Providers earn financial incentives for providing higher-quality, more coordinated care while avoiding wasteful spending, and insurers spend less on a system that’s better tailored to patients’ needs. That’s the theory, anyway.
It’s unclear whether ACOs are a panacea for healthcare reform. Since their creation in 2010 under the Affordable Care Act, ACOs have delivered mixed results, suggesting they might not be the best healthcare model for saving money, particularly for smaller organizations.
“ACOs are a good concept,” says Michael Millenson, president of Health Quality Advisors. “A lot of criticism has come not because of the central concept but because people are engaged in disputes over what they are and how to measure how much they saved and over what period of time.”
Initially built to work under Medicare, ACOs are a group of healthcare providers who agree to be held accountable for the cost and quality of care delivered to a defined group of patients. In return, the providers agree to financial rewards or penalties based on their ability to meet benchmarks. The goal is to reduce spending on unnecessary tests and procedures because patients are tracked through one system in which all their providers share data. This tracking is designed to keep patients from falling through cracks in the system, thus improving outcomes. But after seven years on the market, it’s too early to tell whether ACOs will improve the quality of care while cutting costs.
You would have to follow an ACO for a number of years to determine its effectiveness—a challenge for employers who change their benefit programs on an annual cycle. One way to gauge whether to jump on board is to look at the data. “We do it in other areas of life,” Millenson says, “but the jargon in healthcare tends to be more impenetrable.”
The most readily available ACO performance data are found on the organizations participating in programs under the Centers for Medicare & Medicaid Services (CMS), including the Medicare Shared Savings Program (MSSP) and the Pioneer ACO Model. The most recent figures available from CMS show that Medicare ACOs participating in the shared savings program saved $652 million in 2016. Yet only about half of the participating ACOs saved money.
And CMS paid $691 million in bonuses, which go to providers who give high-quality care while bringing spending below CMS benchmarks. Overall, the program ended up losing $39 million.
“The results have been mixed if you look at the Medicare Shared Savings Program,” says Yulan Egan, a practice manager with Advisory Board, a healthcare consulting firm based in Washington, D.C. “I would say the results from last year were the most promising yet. There were a lot that didn’t achieve savings, but we did see more that did this year, and CMS came close to breaking even” when the MSSP, Pioneer and other programs were combined.
Rachel Sokol, a research practice manager at Advisory Board, says the most successful ACOs are those that have been at it the longest. Because the ACO model changes the way providers partner with plans, she says, it takes time for providers to learn the new system. This trend generally is borne out when you dive deeper into the statistics. ACOs that joined the shared savings program in 2012, its inaugural year, saved $299 million in 2016. Those that joined in 2016 saved only $5 million their first year. The UT Southwestern Accountable Care Network, for example, has saved more than $73 million over a three-year period, but its first year of operation netted only $2.9 million in savings.
A handful of other major ACOs also showed positive results. The Aetna Whole Health—Memorial Hermann Accountable Care Network estimates it saved one plan sponsor $5.9 million over two years, while the cost of care for its diabetes patients dropped an average of about $9,000 over two years (compared to the same treatment for patients not enrolled in the ACO).
Yet many ACOs simply can’t make things work. The CMS Pioneer program is down to nine participants from its original 32 (most dropouts cited “financial uncertainty”). Some ACOs have disbanded, such as Wisconsin’s Integrated Health Network, which spanned the state with more than 8,000 providers and 50 hospitals at its peak. But it scaled down in 2016 and closed last year because of marketplace changes and the costs of running such a grand system. “An organization should join an ACO because its employees will be better taken care of,” Millenson says. “It shouldn’t expect miraculous financial savings to sprout from a barren earth as soon as it signs on the dotted line.”
This bears out what most experts contend: changing the healthcare system is a very slow proposition.
“You have to ask how patient companies are willing to be in terms of waiting on savings,” Egan says. “One of the reasons Medicare has been so aggressive with these is because it has a pretty stable patient population. It’s not an issue like in the employer realm, where you have high rates of turnover. If their average turnover is five years, are they willing to make an investment in an approach that might take 10-plus years to pay off?”
Are ACOs an option for small and midsize employers?
“By definition, this is a strategy that works best where you have a density of members and providers,” says Paul McBride, CEO of accountable care solutions business development from Aetna. “It requires a large population on the side of the provider to invest in the kind of change that is necessary.”
You would typically need 2,000 to 5,000 covered lives to measure quality, spread out risk and get providers to collaborate on services like expanded access, according to Emma Hoo, a director at the Pacific Business Group on Health. It’s also more difficult for providers to justify working with small groups (particularly in the commercial realm) because of the volatility of their population.
Take, for example, UT Southwestern Accountable Care Network, which aggregates data from its 350,000 patients to identify areas where it can make improvements, such as increasing vaccination rates or reducing ER visits. That would be tough to pin down in a group of 200 people. “We can take that to the bank and say we need to work on specific areas because we have large numbers of credible data,” says Danny Irland, UT Southwestern’s CEO.
David Smith, chief development officer for Leavitt Partners, says it’s not as easy for small businesses to be included in ACOs but they can make their way in through a couple of different vehicles. Employers can work directly with a primary care provider or specialty network and create a contract with them for a particular set of services. Or they can work with a group to organize enough mass in a market so providers will be willing to start an ACO.
A small number of organizations are creating products malleable enough to work with small businesses. One is Regence BlueShield of Ohio, which plans to roll out its Regence Medical Neighborhood in 2018. The ACO allows patients to choose a primary care medical group to coordinate their care through affiliated hospitals and specialists. Regence has created 17 quality-improvement targets for providers in the ACO to meet in an attempt to provide high-quality care at a low cost. The organization says participating physicians have already worked to reduce unnecessary care, prevent readmissions, monitor medications and better manage chronic care in Regence’s other ACO programs. The plan will be available to both individuals and businesses with 51 or more employees.
Reed Bjergo, vice president for next-generation products at UnitedHealthcare, says the company spent nearly $60 billion on value-based payments in 2017, nearly half of that year’s output. It is looking to expand that number, and one place it is focused on is its NexusACO product.
Already available in 22 markets, UnitedHealthcare plans to release this commercial ACO to nearly a dozen more areas next year. UnitedHealthcare makes its way into markets by working its top-rated groups in each place into a NexusACO. The providers sign on to meet UnitedHealthcare’s quality standards and cost-of-care and delivery criteria. The benefit plan allows patients to see other providers, but it incentivizes them to use ACO providers by offering the highest level of benefits with those visits. Because the providers are different and each market has different needs, the NexusACO contract looks slightly different everywhere.
“We meet providers where they are within their process and work on a plan together where we can focus on shared business objectives and get to the point where it makes sense for both entities,” Bjergo says.
He says UnitedHealth Group tends to enter a market working with national accounts because that’s where it sees demand for the product. But the ACO is also a solution for smaller businesses.
“We have some clients in the midsize space,” he says. “We certainly see it as viable for small businesses with two to 10 lives. To look at quality and cost results, you need to have a sizable population flowing through and driving enough volume. But we have developed a product that can be delivered from large groups down to small businesses, most of which are fully insured, and they can co-exist because it is a national product.”
Though some ACOs show savings in the millions, they’re also investing millions to alter the way they provide care. And they should be. In fact, Sokol says one way to determine whether providers are truly trying to change their practices is to see where they make investments and allocate resources.
According to a 2017 survey by Leavitt Partners and the National Association of ACOs, these were the four most common health-management tools chosen by ACOs:
- Medication reconciliation and management at outpatient visits
- Chronic disease management
- Reduction of preventable hospital readmissions
- Post-acute care integration.
Most ACOs are also utilizing some sort of care coordination or care management.
Providers are spending substantial sums to implement these strategies. CMS originally estimated it would take $1.8 million to start an ACO and operate it the first year. This number is close to what the Leavitt study found: providers spend an average of $1.1 million on care management and $600,000 annually on technology.
But a 2011 study conducted by the American Hospital Association found providers can spend as much as $26 million to successfully start and operate an ACO. Danny Irland, CEO of the UT Southwestern Accountable Care Network, says his organization budgeted more than $100 million for population health-management activities for its nearly 350,000 covered lives in fiscal year 2018. Even in smaller markets, the costs can be extensive. A 2016 study on costs for primary care providers in rural areas found that, from 2011 to 2012, the mean cost per visit for non-ACO providers rose by 4.4%, while the cost for providers who joined an ACO rose by 13.5%.
If Not Cost, Then Quality
If ACOs are seeing mixed results in cost savings, what incentive do employers have to get into this market? Well, many ACOs do appear to be moving the needle on quality. In the CMS Pioneer model, the 23 participants taking part in year two of the program saw overall quality-of-care improvements. In the MSSP, nearly 30% improved their quality of care enough in the first year to qualify for shared savings. And the National Association of ACOs reports that MSSP participants improved on 91% of their quality measures the first year of reporting, 82% the second year and 84% in 2015.
But finding similar numbers in the commercial market is a challenge. Commercial ACOs don’t publicly report their cost and quality information, and providers are often hesitant to offer solid data. For example, of the large commercial groups we looked at, UT Southwestern reported saving CMS more than $73 million over three years taking part in Medicare programs while meeting a quality rating of 95%. But it declined to provide data on its commercial plans.
UnitedHealthcare Services, which launched its nationwide commercial NexusACO in 2017, saw 10% fewer hospital admissions and emergency room visits among its participating groups. Reed Bjergo, UnitedHealthcare’s vice president for next-generation products, says it’s too early to have more solid data on the program’s performance. But he says the organization sees ACOs as critical to UnitedHealthcare’s commitment to moving toward value-based care.
Aetna Whole Health, a partnership between Aetna and Memorial Hermann Accountable Care Network physicians, saw impressive numbers in 2015, including 40% fewer outpatient surgeries, 8% fewer hospital bed days and a 19% decrease in the number of patients with unhealthy glucose measures.
Paul McBride, CEO of accountable care solutions from Aetna, says the organization focused its energies on “a more transformational” type of ACO known as product-based. In this model, all of the employee members specifically select that ACO instead of other insurance plans (as opposed to an attribution-based model, in which employees are placed by their insurer and may not even know they are part of an ACO). “It’s a much more direct way to associate physicians and members in the partnership of that benefit period to drive improvements,” McBride says.
A 2016 Health Affairs study comparing the results of commercial and public ACOs found commercial ACOs tended to be larger and more integrated with hospitals and spend less than those in CMS programs. They are more likely to be led by physicians than hospitals, have a higher primary care doctor-to-patient ratio, offer more specialized services and have higher quality-of-care scores. But the authors offered a caveat: the nearly 400 ACOs they examined still had far to go to achieve systemic change in both groups.
To wit, only one third of the providers had the ability to monitor their financial performance. One third of the commercial ACOs and 15% of the public ACOs had shared electronic medical records (EMRs) among their providers. Fewer than half of the respondents had implemented chronic care programs, involved patients in healthcare decision making or integrated behavioral health with primary care. And neither commercial nor public ACOs were particularly active in reducing outpatient-visit costs or the use of imaging and post-acute care—all activities considered important benchmarks of population health management.
The authors concluded that ACOs generally lack adequate infrastructure to make major changes to provider compensation and improve quality. “Our findings demonstrate that ACO delivery systems remain at a nascent stage,” they noted. And the inability to make critical changes to the system, like implementing chronic care programs and shared EMRs, “could substantially limit delivery system transformation.”
Millenson says businesses taking part in commercial ACOs should be able to expect better-coordinated care and digital patient outreach. And most have some sort of care managers to coordinate among providers. Anything other than that is a bonus. “If you can buy a healthcare product for your workers where they are sick less, that is a good value for your money, and it will translate to healthier and happier employees. You have to make sure [the ACOs] are committed to this way of practicing medicine.”
What Makes a Good ACO?
In the Health Affairs study, the commercial ACOs did appear to use more disease-monitoring tools, patient satisfaction information and methods for improving quality than public ones did. These are all important aspects of a good ACO. To compare ACOs, the experts say, follow a checklist of preferable traits.
First, ACOs should be organized (and preferably owned) by a primary care provider. Physicians just tend to be more willing to cut costs like unnecessary ER and hospital visits because they don’t have the “mixed incentives” hospitals do (i.e., the need to keep beds full in order to make money), Millenson says. He says physicians also tend to be more innovative than hospitals.
Primary care providers have been critical to the success of the UT Southwestern Accountable Care Network. Irland says the network developed an infrastructure that allows them to work with local providers as well as their employed physicians. They installed EMRs in the practices to create a uniform platform on which to share providers’ information. The moves allowed the primary care providers to “homogenize” their quality and financial reporting to better understand how efficiently they managed costs and clinical aspects of care.
Millenson says it’s also important to have a large portion of providers’ payment base tied to value-based payments. Such a business model offers more incentive for providers to make systemic changes. “If they are dabbling in it—like 10% to 15% of their reimbursements are tied there—they will give good rhetoric,” Millenson says, “but they are just dabbling and you won’t get the best results.”
When looking at contracts, the San Francisco nonprofit Pacific Business Group on Health (PBGH) recommends looking for ACOs in which providers have some downside risk to ensure they have skin in the game and aren’t just rewarded for good outcomes. Instead of just rewarding for meeting targets, an ACO can penalize an organization for errors, inappropriate use of services and “never” events.
In order to be rewarded—or penalized—an ACO has to be able to provide clear, understandable quality targets. Those that take part in CMS programs have set benchmarks to meet, but the commercial side is more nebulous. If benchmarks exist, they are set by the owner of the ACO and are often unknown to payers and businesses taking part.
“The devil is in the details, and that’s what you really need to look at,” Millenson says. “Who and what are they measuring? How much of the doctors’ income is at risk if they don’t meet their targets?”
Millenson says it’s difficult for employers to grasp the complexities of these contracts, which makes it imperative for brokers to dig into the details and ask probing questions. “In some ways, it is an opportunity for brokers to increase their own level of sophistication,” he says. “As healthcare changes, broker sophistication has to grow as well.”
Emma Hoo, director of pay for value at Pacific Business Group on Health, says one of the group’s major initiatives in the ACO realm has been to build a common set of measures for commercial ACOs, allowing their results to be better quantified and compared. PBGH offers a comprehensive questionnaire that can be used to gain insight into an ACO’s strategy and offerings in a particular market. The tool includes detailed questions about what quality metrics are used, their standards, the frequency of reporting and the follow-up processes taken when benchmarks are not met. It also offers a list of measures against which the ACO can compare its information.
“When employers are purchasing coverage through a health plan, all of the data are blended, and you don’t necessarily see the difference in services, quality and price point,” she says. “And having an ACO offering creates the opportunity to look under the hood and compare provider organizations and get the information to provide better access to services.”
Aetna’s McBride agrees. Demonstrating you can meet quality markers is crucial for an ACO, he says, particularly when you are moving into midsize and large markets.
“We are trying to create greater transparency and buy-in,” McBride says. “Most members thought they were paying for quality, and now they hear they aren’t and are confused. The basic question is, ‘Why would I pay more for a program like Memorial Hermann when I was always paying for good care?’”
McBride says it took a concerted effort, serious investments, data tracking and cultural change to build the Memorial Hermann ACO. All of it, he contends, is required for the success of this kind of endeavor.
“Lots of organizations are talking about coming together as a clinically integrated network, but a minority of organizations are making sizable investments in infrastructure and capabilities,” he says. “Becoming an ACO is a significant undertaking, and there is a big difference in terms of what they do and the results they get. You have to invest, see your results, have a feedback loop and modify it. It is a cycle, and many providers are on this journey but are nowhere near the end.”
Broker ACO Checklist
- How many ACO contracts does the provider hold?
- How does the ACO differentiate services for different health plans (including those outside the ACO)?
- What does the provider do differently for the ACO (compared to its traditional health plan contracts)?
- What is the provider’s road map for performance improvement and managing the cost of care?
Patients and providers
- Do the members know they are part of an ACO if it’s an attributed relationship?
- Does the provider know when it is seeing patients in the ACO?
- Is the provider aware it’s part of an ACO?
- What is the provider required to commit to?
- How does the provider gain access to needed metrics?
- Is improvement monitored?
- Are ACO-attributed patients getting different care management from those who are not part of an ACO seen by the same providers?
- How are complex patients identified? Are their needs addressed differently from others?
- If behavioral health services are offered, how is that information coordinated?
- What performance metrics are measured?
- Are the metrics reported to physicians and other providers? If so, how often?
- Are systematic improvement approaches in place to support needed redesign and monitor progress?
- Are provider payments based on quality or fee for service?
- What kind of alternative payment models are part of the ACO (e.g., bundled payments or capitation)?
- Is information on claims and authorization coordinated if a provider uses a pharmacy benefit manager?
- What portion of doctors have EMR access that supports medication management services?
- Do protocols exist to optimize efforts like generic prescribing, step therapy or utilization management?
- What information does the plan provide the ACO, and how often is it given?
- Does it have a common EMR platform for providers? How many use it? For those that don’t, how is information exchanged between the provider and the ACO?
- Do providers and care coordinators exchange information in real time?
- Do care coordinators and physicians access the same information?