Why Pioneer ACOs Are Disappearing and 3 Trends to Expect from the Exodus
In March of 2010, the Patient Protection and Affordable Care Act (ACA) was signed into law, marking the largest expansion of healthcare in the United States since the 1960s. The ACA contained provisions that authorized CMS to create accountable care organizations (ACOs), groups of primary care doctors, specialists, hospitals, and other caregivers that provide healthcare services to a defined population of people. CMS initiated the Pioneer ACO program along with other shared savings programs. The Pioneer program was developed for healthcare organizations that already had deep expertise in coordinating care for patients across care settings, utilization management, and clinical process improvement.
Pioneer ACOs were seen as leaders in the movement from fee-for-service payment models toward value-based reimbursement. Over the first three performance years (2012 to 2014), these Pioneer ACOs generated $304 million in savings, with significant increases each year. Half of the Pioneers received shared savings payments for their quality improvement work.
With this apparent level of success, why did the number of Pioneer ACOs drop from 32 in January 2012 to just 9 in January of 2016?
Accountable Care Organizations in Review
The intent of an ACO is to improve the quality and coordination of healthcare delivered to a defined population of patients. After the ACA was signed into law, CMS initiated several programs to give organizations a few options for creating their ACOs. Many organizations and groups voluntarily entered into similar arrangements with commercial payers because of market pressure. Groups that entered into these types of arrangements have also been commonly referred to as ACOs, even though they are not defined by the ACA or administered by Medicare. All ACOs typically operate under payment models that are moving away from a fee-for-service and toward value-based reimbursement (essentially paying the ACO to keep people healthy instead of reimbursing for services).
Even though Pioneer ACOs produced more than $120 million in savings during 2014, six received no shared savings payment, and three owed Medicare a total of $9 million in losses.
Let’s examine the exodus from the Pioneer program and what alternatives await these organizations that want to lead the way in value-based care.
Solving the Problem of Overutilization
In order for an ACO contract to be financially beneficial, the organization must be able to reduce the cost to provide care for its defined population of patients. Operating as a successful ACO during the first few years was primarily about limiting unnecessary and expensive utilization. However, some health systems that joined the Pioneer ACO Program were already doing a good job of reducing excess utilization, which obviated this potentially low-hanging fruit.
One good example of this, Dartmouth-Hitchcock Medical Center, is described in an October 22, 2015 article from Becker’s Hospital Review. Dartmouth has been working on coordinating care and reducing unnecessary utilization for years, so many of the large improvement opportunities had already been realized. Dartmouth was unable to achieve savings compared with its benchmark in 2013 and ended up paying back $1.4 million to CMS. Again in 2014, when Dartmouth-Hitchcock did save almost 4 percent over the previous year, and beat the comparison group (other non-ACO patients in its region) by 1.2 percent, it still paid CMS another $3.7 million because it came up 0.5 percent short of the savings target. Dartmouth-Hitchcock was penalized for not attaining enough success when compared with themselves. Many efficient systems faced with this problem have felt the need to look for more attractive options outside the Pioneer program.
Robert A. Greene, MD, Dartmouth-Hitchcock’s executive vice president and chief population health management officer, summed it up this way: “It’s like a runner who runs a 12-minute mile [with a goal of running a four-minute mile]. If they want to train and exercise, they can probably get that down to a 10-minute mile… it is much harder for organizations that are running a five-minute mile to work up to running a four-minute mile.”
Performance Benchmarking Issues
At the outset, Pioneers agreed to accept the risk of losses as a tradeoff for potential upside shared savings they would receive if they could control Medicare spending and deliver quality care. If ACOs kept their expenditures for patients below the benchmarks, they received shared savings dollars. If ACOs exceeded the targets, they were subject to financial penalties if they had taken on downside risk. In the first performance year (2012), Pioneer ACOs could avoid losses by agreeing to accept smaller shared savings payments, but in subsequent years they were required to take on risk of loss. For Pioneer ACOs without a ton of waste to eliminate, the risk for losses is substantial without a lot of potential for gain.
One of the primary complaints of all Medicare ACOs has been about benchmarking and this definitely affected Pioneer ACOs. The calculations used in the both the Medicare Shared Savings Program and the Pioneer ACO Program were not very transparent. Part of this was because of data sharing limitations put in place by the programs and part was because of the design of the benchmarks. Some ACOs have struggled with performance targets because variations in their attributed populations and benchmarks during the course of the year made it difficult to accurately predict financial outcomes. Benchmarks have not previously included regional adjustments but instead compare ACOs to themselves in prior years with adjustments for the risk profiles of the attributed populations in each given year. Successful ACOs face a more difficult task each year to get better than they were the previous years. On top of that, they must do this in an environment where CFOs feel like they’re flying blind, not knowing how they measure up against the targets until long after the year ends.
Prospective vs. Retrospective Attribution
Under the Pioneer program, patients have been assigned to ACOs using prospective attribution, which is based on the patient’s use of services in the prior year. This method assigns responsibility for a future period based on past usage. Quality and cost metrics are reported throughout the year. Though patients are allowed to seek care from any provider, they are assigned to providers they have visited most in the past.
While the merits of prospective attribution are widely understood, and criticism of retrospective attribution is abundant, there are disadvantages that Pioneer ACOs have felt as compared to their peers in other ACO programs. For example, the Medicare Shared Savings Program (MSSP) uses retrospective attribution, in which patients are assigned at the end of the year based on their use of services during the actual performance year. Retrospective attribution assigns responsibility for the past period based on usage and claims data. Quality and cost metrics are reported at the end of the year. While retrospective attribution means you don’t know what your population is until after the year is over, it also ensures that your population is most closely aligned with the actual patients being cared for by the ACO. Initiatives to improve patient care based on actual encounters will have a greater chance of impacting the financial results of the ACO than with prospective attribution.
The arguments from Pioneer ACOs against prospective attribution have been that 55 percent of the patients they treated weren’t attributed to them, and those patients weren’t counted in their performance review. At the same time, 17 percent of the patients under prospective attribution did not receive care from providers at their assigned ACO. While the responsibility was there, the impact from care initiatives was limited for the assigned population.
The arguments for and against prospective attribution have a lot of merit. Invariably, ACOs have agreed that CMS needs to improve the way attribution is handled in all of its ACO programs. This is one reason why some ACOs have reason to move from the Pioneer program into other programs such as the Next Generation model or Track 3 of the MSSP.
The ACO Outlook is Still Bright
Despite the movement away from the Pioneer program, it has generated more savings and improved quality each year. Total model savings per Pioneer ACO increased from $2.7 million to $4.2 million to $6.0 million over the first three performance years.
According to an August 2015 fact sheet from CMS, Pioneer ACOs increased their average quality scores from 71.8 percent to 85.2 percent to 87.2 percent over the first three performance years. The organizations improved in 28 of 33 quality measures between performance years two and three. The biggest improvements were in medication reconciliation (70 percent to 84 percent), screening for clinical depression and follow-up plan (50 percent to 60 percent), and qualification for an electronic health record incentive payment (77 percent to 86 percent). In addition to these measures, average patient and caregiver experience scores improved in five out of seven measures between performance years two and three, a strong indicator that utilization and financial streamlining can be done without negatively impacting the patient experience.
ACO Success Requires Learning New Skills
The skills health systems have learned and mastered over the years, while still important, are not sufficient to guarantee success as an ACO. Some of the skills have typically included payer contract negotiation, supply chain management, and network management (to bring in referrals). For healthcare entities, profitability was measured by revenue (payment received for rendered services) minus costs (of which labor and supplies are the most substantial). In the new paradigm of value-based care, things that used to count toward revenue now actually hit bottom-line costs, since providers are no longer paid incrementally for every service they deliver. Under the old system, when patients didn’t come into the hospital, revenue dried up. However, under value-based care, hospital utilization is a cost, not a revenue driver.
Health systems must learn new skills to be successful in this new financial paradigm. They must begin to learn and execute skills and competencies that have been typical of insurance companies, such as understanding which patients are most likely going to incur high costs and providing care managers to help these patients navigate the healthcare system and manage their disease(s). One of the confounding problems for ACOs trying to learn these new skills is that some of the return will come after the timeframe of the risk contracts. ACOs must start practicing these skills, but not invest too heavily upfront on long-term improvements. For example, the return on investment from a successful care management program can be measured in one to three years—or more—potentially adding cost during an ACO contract term, but not increasing the potential for revenue from shared savings on that contract. Getting this balancing act wrong can literally put organizations out of business. This is why successful ACOs must start with heavier investments in short-term efficiencies, such as unnecessary and inefficient utilization, since the return on that investment is closer to real-time.
Overutilization and the Role of the Enterprise Data Warehouse
In order to understand what opportunities exist for reducing unnecessary utilization, health systems need to access the clinical data stored in their electronic health record system. This is typically accomplished by pulling this data into an enterprise data warehouse (EDW). From there, it can be visualized in business intelligence applications allowing clinicians and administrators to make decisions on where the best opportunities lie.
One of the first things that Health Catalyst does to help our ACO health systems is provide a tool that lets them see how patient expenditures and utilization (in terms of ER visits, inpatient admissions, high-cost services, readmissions, etc.) are trending across the different clinical groups and service lines. If ER visits or inpatient admissions are increasing, the health system can see what diseases and what service lines are contributing to the increased utilization. This gives healthcare entities a starting point for opportunity prioritization.
A health system with an accountable care contract will get claims data for the services provided to their contracted population. Health Catalyst loads that data into the data warehouse, along with other data sources, such as the electronic health record system, financial system, and cost accounting system. Once the data is loaded into the EDW, we use the Health Catalyst suite of ACO applications to understand where overutilization and unnecessary utilization is occurring, including the specific diseases and service lines responsible. This drilldown capability is essential for organizations that, in the first couple of years, need to get a handle on utilization while making population health investments over the longer term.
Adjusting the ACO Model and the Next Generation
As mentioned earlier, some of the Pioneer ACO organizations are moving over to the MSSP or other ACO-like contracts. Some are moving into the Next Generation model for ACOs. The next gen model retains many of the MSSP and Pioneer program elements, but adds four significant new features:
- Greater financial risk and reward with options for upside and downside risk
- Flexible payment options:
- normal fee-for-service
- fee-for-service with ACO support payment
- population-based payments
- Beneficiary engagement
- Better benchmarks
In addition, the next gen model allows for patients to voluntarily align themselves with an ACO by opting in to the program. CMS essentially is attempting to keep some of the benefits of retrospective attribution by continuing to use retrospective claims-based attribution as the default alignment mechanism, but is also adding some of the benefits of prospective attribution by allowing patients to opt-in to the ACO before the contract year begins.
Benchmarks are going to be set by looking at one year’s worth of historical data (not three years), and then adjusting it for risk and region. This will help get us to where we have health systems that are competing, not just to deliver care, but to keep their populations healthy based on what the expected cost should be for that population under contract.
Three Trends to Expect
With accountable care down the road, we expect to see three trends emerging.
- Consolidation in the healthcare ecosystem. Our Big 5 insurers are now the Big 3. Along with this, we’re already seeing a consolidation of healthcare systems trying to create entities to take on this risk and have a bigger seat at the table when they’re negotiating with insurers.
- More employer-led innovation in the accountable care space. Employers have exhausted a lot of value out of value-based insurance design in terms of high deductible plans, so they are going to try and change the way care is provided for their employees by partnering directly with health system entities, paying them a fixed amount, and receiving coverage for every employee. We’re already starting to see this with some companies.
- Continued innovation in the digital health space. We’re going to see more companies trying to provide innovative solutions for employers and health systems, including tools for health systems to manage the care of their populations.
The Long-term Future of Accountable Care
Everyone is rising with the tide generated by the ACA and the implementation of accountable care contracts by Medicare. Even as organizations push back on specific problems with individual programs, the overall effect of all the programs has been to push the market toward value-based care. By one metric or another, it has definitely encouraged and brought needed change to the market. It has pushed the commercialization of ACO-like contracts throughout the healthcare market.
Looking back to the managed care era from two decades ago, it’s interesting to note similarities and differences. A few years ago, some people were saying accountable care had been tried and we couldn’t do it again. But there are a lot of differences this time around. We don’t have the same sort of data infrastructure now that we did in the 1990s. Everything was on paper then and it couldn’t be analyzed the way it can be now. Our ability to predict risk and cost is vastly improved. Risk adjustment wasn’t handled very well back then, causing unfair financial burdens to be placed on provider groups. All those lessons learned have created an environment where we have a chance for value-based care to succeed.
Accountable care programs are here to stay, thanks in part to the support of the ACA. Over the next 3 to 5 years, we will see a tipping point occur in value-based payment. The proliferation of data and the tools to analyze and exchange that data are critical to the success of this transition. Once this tipping point is reached, providers and health systems will more fully bear the financial risk of the populations they are accountable for, and by that time they will hopefully have the skills and analytical tools in place to successfully manage their ACO.
Would you like to use or share these concepts? Download this presentation highlighting the key main points.