Healthcare Analytics, Population Health Management, Healthcare Big Data

Population Health News

Will Inadequate Metrics Doom the Accountable Care Organization?

By Jennifer Bresnick

- Accountable care organizations are all about the data.  Detailed quality reporting and cost benchmarking are the hallmarks of this value-based reimbursement model, and the carefully balanced system of risk, savings, performance, and quality has attracted hundreds of provider groups to the Medicare Shared Savings Program (MSSP) with the promise of reaping financial rewards.

Accountable care organizations

But the MSSP isn’t a surefire way to rake in the revenue, as many participants have found out during the first three years of the experimental program. 

ACOs have been having a tough time meeting stringent spending thresholds that seem to punish the highest performers, and some organizations have questioned the benefits of pouring time, technology, and resources into a care delivery model that doesn’t always live up to expectations.

While CMS is quick to point out the fact that its accountable care organizations are helping Medicare slash spending while allowing providers to share in hundreds of millions of dollars in savings, a closer look at the results from the MSSP and Pioneer ACO programs leads to some more questionable conclusions about the sustainability and impact of the accountable care mindset.

Are accountable care organizations worth the trouble?  Are the benchmarks used to pin quality and performance to financial reimbursement up to the task of helping providers manage incredibly complex patient populations?  How can healthcare organizations take advantage of what accountable care has to offer without falling victim to some of the inherent flaws in the structure of current ACO offerings?

READ MORE: CIOs Focus on EHR Optimization for Population Health, Analytics

In 2014, twenty Pioneer ACOs and 333 MSSP participants produced $411 million in combined savings, CMS announced in August, but only half of Pioneers and a third of MSSP organizations actually received shared savings payments for their quality improvement work. 

While these results may not seem particularly stellar, given how hard CMS is pushing to move the majority of Medicare payments into similar quality-based initiatives over the next three to five years, they may not be as dismal as they appear at first glance, either.

“I think that if one takes a step back and looks at the results of the Pioneer and MSSP programs with a little perspective, they may be even better than one might anticipate, given the maturity of the program and parameters that many of the organizations are operating under,” said  Dr. Jonathan Niloff, Chief Medical Officer of McKesson to

“Especially with the MSSP program, many of the organizations are new to the whole business of managing care and accepting risk.  So I’m not at all surprised, given the expertise and infrastructure that's required, that some of the organizations that are new to this endeavor are not generating savings early on.”

Despite the importance of savings to what is primarily a financial arrangement, Niloff points out that an additional 89 MSSP participants improved their quality rankings, even if they didn’t do quite well enough to share in revenue.  That’s an important achievement, and should not be undervalued.

READ MORE: Using Machine Learning to Target Behavioral Health Interventions

“It’s significant that they all demonstrated quality improvements,” he said. “Those are much easier to achieve at the beginning than the direct financial improvements. But I think it's interesting to see that the longer one has participated in the program, the greater the probability that one is accruing the clear financial savings. I think that was clearly demonstrated in the data. Nineteen percent of organizations in their first program year qualified for bonuses. Thirty-seven percent of ACOs in their third performance year received shared savings.”

But the MSSP could do better, argues April Wortham Collins, Senior Analyst at Decision Resources Group, and the skewed benchmarking system may be to blame for the relatively low number of providers accruing meaningful financial gains from the program so far.

“The benchmarks are set at a very high level,” she explains. “CMS looks at all of the beneficiary expenditures for the patient population that is assigned to that particular ACO for Part A and Part B Medicare services over the last three-years. So, for those patients that you are responsible for caring for, they look at how much they've cost Medicare in the last three-years. For year 1 of the ACO agreement, that’s your benchmark.”

“If you can care for this population in such a way that not only improves quality but saves money, and comes in below that benchmark, you can reap some of those savings that they were able to generate, based on the type of agreement you have with CMS,” she continued. 

Medicare allows providers to choose between two tracks within the MSSP.  Track 1 allows participants to share in savings, but doesn’t require them to accept risk.  If an ACO spends more for their attributed patient population under the accountable care arrangement than they did while receiving fee-for-service reimbursement, they don’t get a check from CMS, but they aren’t liable for their losses, either. 

READ MORE: Unfiltered EHR Data Overloads Physicians, Perpetuates Burnout

The second track, ostensibly designed for more mature accountable care organizations, are eligible for a higher percentage of possible shared savings, but must also take financial responsibility for falling short on quality and spending.

Ninety-nine percent of MSSP ACOs have chosen the one-sided shared savings model.  Only three ACOs are currently participating in the more advanced risk-sharing plan.

That’s because meeting CMS’ benchmarks for shared savings is difficult enough without worrying about revenue losses, Wortham Collins says.  And the process gets even tougher each year that the ACO stays in the program, because the threshold keeps creeping upward based on how well the organization did the previous year.

“If your ACO successfully implements some basic population health management techniques, chances are you’ll be able to generate some savings, and maybe even share in them,” she said.  “However, for each year of the ACO agreement, that benchmark gets adjusted.  The bar gets raised, and it becomes harder to meet the threshold because you’ve already taken care of a lot of the low-hanging fruit.”

“It becomes more difficult in year two, because you have be increasingly innovative with care management and coordination in order to trim even more fat than you did in year one.   It’s even harder for ACOs that started off doing very well.”

“The argument from the ACO community is that if you start off doing a really good job, or you keep doing an extraordinarily good job year after year, you’re effectively being punished for that by the way the benchmarks are set up right now,” she pointed out.

The framework also doesn’t take into account how much providers need to invest in health IT infrastructure in order to participate in the first place, added Timothy ‘Dutch’ Dwight, VP of Business Development at Medullan. 

Even commercial accountable care arrangements, which often include a per-member, per-month payment to entice participation, do not provide enough of an incentive to allow the adoption of the advanced data analytics systems required for meaningful results.

“Nobody is making any money in the process here,” he said. “I would call the ACO an administrative burden on both the health plan as well as the hospital or provider network. It's hard to set up. There are no data standards, so it's very difficult to capture the data you need.  It's very difficult for providers to get a picture of what quality measures really mean.”

“Hospitals and health systems are not able to spend $50 million on EHR infrastructure,” he said.  “They’re looking at revenue versus loss, and spending money on an EHR or data analytics doesn’t directly contribute to revenue right now, because the incentives aren’t aligned yet. What organization in your life has ever said, ‘Let's decrease our revenue – that's a good thing to do’?”

Being averse to risk is understandable for healthcare organizations whose revenue cycles are already balanced on a knife’s edge, and stand to take a beating over the next few years as EHR certification requirements change and the demands of interoperability require reinvestment in health IT infrastructure. 

Add quality-based financial losses to the picture, and it’s no wonder that industry observers have clamored to point out that even the most sophisticated accountable care organizations – those participating in the Pioneer program – have abandoned the pathway in droves.

Called a “crucial milestone” and one of the “centerpieces of the Affordable Care Act,” the Pioneer ACOs have had a troubled history of retention and consistent gains.  Thirty-two Pioneers started out on the accountable care journey, but just nineteen remained in 2014.  Even though they produced more than $120 million in savings during 2014, six received no shared savings payment, and three organizations owed Medicare a total of $9 million in losses.

“There are some unusual features of the Pioneer ACO model, which are built into the way that performance is measured,” said Emily Brower, Vice President of Population Health at Atrius Health, which came out $2.8 million ahead this year.  

“The Medicare fee schedule differs based on your geographical region. The difference between the prices in that region and the national average can help or hurt an ACO regardless of how it’s changing care.  That does lead to some disappointing results for some participants.”

“There is also a threshold of savings that an ACO has to achieve before it shares in savings, and there are different options you can choose,” she added. “An organization that's very risk-averse might choose a higher threshold because that will protect them on the downside, and so they could generate savings but not return those savings to the ACO.  Those are some of the issues that have shaped what we’ve seen this year and in past years.”

But fleeing the Pioneer program doesn’t mean these providers have failed to make accountable care work, Niloff said.  It’s just all about the benchmarks.  “Most of the organizations who left the Pioneer program didn’t abandon risk,” he pointed out. 

“They just left the program because of the way their scores were calculated, and almost all of them went into another type of shared savings program.  So you have to temper how you look at those numbers.”

“The Pioneer ACOs are folks that really were doing this sort of thing prior to there being a formal program for it,” added Wortham Collins.  “And in fact, CMS really looked to them as an example when they were modeling the MSSP.  The reason why some providers have dropped out of the Pioneer program has everything to do with risk.  The Pioneers had to go full-tilt into accepting risk, and some of them have had to pull out because they’ve found that it’s simply not viable for them to do that.”

“One of the challenges is that CMS has tied its program parameters to the level of risk that an ACO assumes,” agreed Niloff. “And I can understand why CMS is trying to do that as an incentive. But what it's doing at the same time is handicapping ACOs that are trying to come up the learning curve, because it's forcing them to try and generate savings when they’re still figuring out how to manage patients appropriately.”

Some erstwhile Pioneers are looking into joining CMS’ Next Generation ACO program, which hopes to solve some of the financial problems inherent in earlier iterations of the initiative.  But it also demands more from providers in the way of technology. 

“ACOs in the Next Generation ACO Model will take on greater financial risk than those in current Medicare ACO initiatives, while also potentially sharing in a greater portion of savings,” said Patrick Conway, MD, Deputy Administrator for Innovation and Quality and Chief Medical Officer at CMS in a March blog post announcing the new option.

“To support increased risk, ACOs will have a stable, predictable benchmark and flexible payment options that support ACO investments in care improvement infrastructure to provide high quality care to patients,” he said. “These changes are responsive to feedback from external stakeholders.”

Next Generation ACOs will also be required to boost their patient engagement and population health management technologies, which will include telehealth options, home monitoring, and more robust skilled nursing coverage.  During the second year of participation, they will be eligible to take part in a capitated payment mechanism that may offset some of their costs.

“What people will find most attractive about the Next Generation ACO is the ability to start to create a narrow network,” Niloff observed. “They can start to designate preferred providers and have more control over where patients are getting their care, so they actually have the opportunity to coordinate that care.”

Effective population health management is a basic required competency for accountable care organizations, but immature technology and workflow roadblocks make truly coordinated, managed care a difficult proposition for most providers, Dwight said.

“It will take quite a few years for the data analytics to come up to speed and alignment with the financial landscape,” he predicted.  “Providers are trying to implement care coordination strategies without very good systems to support them.  EHRs really don't do care management.”

And the sheer number of concerned entities trying to reduce waste, coordinate services, and ensure appropriate care delivery may causing more problems than they’re solving.  “You have everyone trying to manage patients, and they’re not necessarily communicating well with each other,” Dwight said.

“If I check in to a hospital today, I might be visited by a care manager from my health plan. I'm going to be visited by a care management person from the hospital. If I'm older and in the Medicare environment, they're going to assign a nurse to me, and that person is going to tell me that they are my point-person for the future.  So as a customer, I’ve got three different people trying to help me, but I don’t know which one I’m supposed to talk to.”

“If you consider the workflow of each one of those organizations trying to do the same thing, you've got a significant overlap, and a big opportunity not only to cut costs, but to streamline the patient experience and make care management actually useful to them.”

The process is even more challenging for most Medicare populations.  Elderly patients are often trying to cope with multiple chronic diseases or major acute events that are not only expensive, but require detailed and constant supervision from a coalition of providers across multiple care settings and specialties.

For accountable care organizations looking to cut spending in the Medicare setting, success once again comes down to having the right data – and a thorough understanding of what it means.

“You really need to understand your patients,” stated Wortham Collins.  “No two ACOs are alike. Even though they all have to follow the same rules and meet the same quality measures, the fact is that accountable care is very much a local game.”

“The Medicare patients that are going to be attributed to you can vary very widely from one market to another, or one state to another.  The makeup of that population will go a long way towards determining how difficult it is to manage those patients and to share in savings.”

Some providers may be starting off with patients who have low levels of medication adherence, or poor computer literacy that limits engagement, or few community transportation options that make it difficult to obtain preventative care. 

Providers may be able to overcome some of these obstacles by implementing outreach plans or using technology tools to close care gaps, “but it’s a completely different story if your patients are already doing all the things they should be doing, and they’re still very sick and costing a lot of money,” Wortham Collins says. 

“The fact is that Medicare patients tend to have a lot of chronic disease, and a lot of high-cost needs.  As they get older, there sometimes aren’t too many costs you can cut while still keeping quality high.”

Once again, this is where Medicare ACOs tend to run into trouble.  Providers who are already doing a very good job of managing their patients don’t have much wiggle room for improvement, and those who have ground to make up in the patient management arena are being held to the standard of those high-performing organizations due to the lack of specificity in the benchmarking scheme.

“One of the big problems for a lot of ACOs is that the historical benchmark is created from data at a national level,” Wortham Collins notes.  “But the actual costs of managing a population could be very different for Arizona versus Colorado.”

“In Florida, for example, the average Medicare beneficiary costs are increasing year by year, because they have a very high concentration of Medicare beneficiaries, and more elderly people are retiring to the state all the time.  If you look at that situation versus somewhere like Colorado, which has a younger population, a one-size-fits all benchmark is very difficult to work with.”

Providers who wish to be successful as a Medicare ACO have to be willing to make major changes very quickly, says Niloff, even if they aren’t technically accepting financial risk.

“There’s certainly a growing understanding that to be successful, you need to change the care model,” he said.

“You need an IT infrastructure; you need to align incentives. I think one of the challenges, especially early on in the program, was just making that cultural and organizational transition.”

“It’s a lot of work to change internal incentive models, implement IT infrastructure, and revamp the care processes required for success in an at-risk model or an ACO model.  It's really the whole concept of implementing a population health management program.”

“When some people started to do this – and if they only had a shared-savings ACO for instance – that program only comprised a relatively small proportion of their total population.  It was difficult to martial the resources and get the buy-in if their at-risk patients were only four or five percent of their total patient population.”

Brower agrees, and attributes some of Atrius Health’s success as a Pioneer ACO to its willingness to jump into accountable care with both feet.  “It’s very, very difficult to operate in both the fee-for-service and the global payment or value-based payment world, because those incentives are not aligned,” she said.  “The more payers in your market that embrace value-based payment, the easier it will be.”

“It can seem somewhat paradoxical, because it can feel that it's very risky to move so fast, especially for systems that haven’t done much value-based work before, but I think it's actually less risky. I think it's harder to get results if you've only got one set of patients or one program in that model, and it actually becomes easier and less risky if you put all your eggs in one basket.”

As CMS and private payers all invest more heavily in value-based reimbursement programs that mimic formal ACO agreements, providers will need to get used to making population health management an everyday activity, Niloff predicts.  The more resources a provider can put behind its population health strategy, the more successful they will be throughout the transition away from fee-for-service reimbursement.

And in the end, overcoming the inherent challenges of ACO benchmarks will all come back to the data, Wortham Collins concludes.  The initial experiment may not be an unqualified success, but ACOs can certainly thrive if they stay the course and make the necessary changes to their patient management strategies.  However, they must have the ability to extract actionable insights from their own reporting first.

“The data is definitely there,” she said.  “It's just a matter of packaging it and delivering it to these ACOs in a way that's useable. And then the ACOs must have the capability to take that data from CMS and use it to gain insights. Make sure that you have that technology infrastructure that enables you to capture and analyze data, and then to use it in a way that you can effectively impact the health of your patient population.”

A strong health IT infrastructure with robust population health management capabilities will allow providers to trim necessary costs while making the most of Medicare’s benchmarks, even if they aren’t quite perfect just yet.  CMS is trying to be responsive to industry feedback about the highlights and challenges of their accountable care programs, and the introduction of the Next Generation program appears to be proof that the model will continue to evolve for the foreseeable future. 

What providers make of the opportunities in front of them remains to be seen, but it is clear that the accountable care organization is very far from failure.  In fact, what Brower calls “a wonderful moment in healthcare” is only beginning to take shape as intrepid ACOs embark on their second three-year MSSP contracts.

“I'll be interested to see when CMS releases the next round of MSSP participants that have reapplied for the program, and how many brand new entries there are for 2016,” said Wortham Collins.

“It'll be really interesting to see how many folks come back. It will also be interesting to see if some of the folks who have been sitting on the sidelines the last three years, watching and waiting to see if this program would really take off, are encouraged by these performance results, or whether or not they see this as a warning.”


Join 25,000 of your peers

Register for free to get access to all our articles, webcasts, white papers and exclusive interviews.

Our privacy policy

no, thanks

Continue to site...