ACO Success Comes With Risk. Get Over It.

For some organizations, forming or joining an ACO is a tentative first step toward taking on risk. For others, it’s just one milestone in a journey toward a full-risk integrated delivery system. In both cases, the advice is consistent: Dive in.

This article first appeared in the April 2016 issue of HealthLeaders magazine.

There was a time when many, if not most, hospitals and health systems doubted that they would ever have to veer from fee-for-service healthcare and all its simplicity despite its misaligned incentives for patient health. That time seems over.

Even in states where Medicaid has not expanded, and even where employers and commercial payers aren't pushing value-based contracting, hospitals and health systems are embracing and experimenting with new models to improve outcomes.

Whether those take the form of commercial or government payer accountable care organizations, they're putting the pieces together.

But how best to undertake the journey? What are some lessons learned from those at various places on the path toward taking full premium risk, or capitation? Two organizations that have been through some challenges provide a variety of perspectives through which others can pattern their journey.

Longevity equals leadership?
If experience with value-based care equals credibility, Montefiore Health System, a six-hospital health system in the Bronx, New York, can lead others. It traces its roots in the risk business to the late '90s, when it had plenty of company in health systems that took on risk through their own health plans, often to their ultimate regret.

"We're in our 20th year of managing populations at financial risk," says Stephen Rosenthal, the health system's senior vice president of population health management. "California excepted, many organizations left the risk business in the late '90s, but we stuck it out."

It helped that they didn't really have a choice.

"When you're managing a government programs population, you don't have a lot of room for mistakes, so you spend a lot of time understanding what that population needs, particularly those who have emotional and behavioral health challenges," says Rosenthal.

"We're in our 20th year of managing populations at financial risk," says Stephen Rosenthal, the health system's senior vice president of population health management. "California excepted, many organizations left the risk business in the late '90s, but we stuck it out."

Asked to pick out a couple of factors that have led to Montefiore's success with risk, Rosenthal starts with leadership supporting the move, but the second factor is much more difficult.

"There needs to be a delivery system that recognizes the value of providing quality services and addresses some issues that occur outside the hospital as opposed to using the hospital as the pivot point," he says.

Montefiore uses its physician network as the pivot point, utilizing the power of 5,000 frontline clinical providers, about one-third of whom Montefiore directly employs.

Rosenthal says conveying a message of trust and one that emphasizes that new tools and processes are about improving patient care for a group with more than 3,000 independent physicians is a challenge, but over time, through investments in technology and by helping them understand what quality means by the metrics, physicians begin to understand there's real value both for them and their patients in being part of the network.

"Building a strong population strategy takes time because you're changing culture," says Rosenthal. "These people have been in a fee-for-service world forever, and when you think about it as a physician, it's much easier to do business on a fee-for-service basis."

Montefiore can take on global capitation, but to get to that point, it began developing customer service, provider relations, case management for complex illness, as well as the ability to manage individuals who have social challenges. None of those efforts came about at the flip of a switch, and those are just the beginning.

The key is knowing the population through social workers and community organizations who are medically skilled to manage individuals at a certain level of care, Rosenthal says.

"If I know the population is 100,000 people, I can use data to drill in to look at subsets that need our attention," Rosenthal says. "I can find pockets of chronic illness—end-stage renal disease, for example. Many of the problems in these populations can be managed quickly on an ambulatory basis if they can be identified quickly."

Montefiore actively consults with other health systems to help scale population health programs.

Rosenthal says it brings significant experience tailoring interventions based on the known characteristics of certain populations. Interventions and skills that work for one group may not work for another, so capabilities have to be scaled based not only on the size of the population but the kinds of nonacute care interventions that will likely be needed by subgroups of that population.

Those needs are discoverable based on what Montefiore has learned over 20 years managing very different populations.

"There is no one answer," Rosenthal says. "A Medicaid population is very different from a Medicare or a commercial population. You build different components to your infrastructure based on the population you're serving."

For a Medicaid population, you might allocate more ancillary services, more behavioral health providers and case workers. You might have social workers in a centralized structure that does telephonic outreach. Through Medicaid, Montefiore has a large mental health population under risk, so they place behavioral health providers directly in primary care practices so the physician has that support at the point of care.

With Medicare, by contrast, the population is elderly, likely with multiple comorbid conditions and more frailties, and may need different kinds of support as opposed to social work—more durable medical equipment or more transportation, for example.

"Of course we use technology to build workflows and identify individuals who would benefit the most from our interventions," Rosenthal says, adding that the system has built a formula to help approach each individual patient for whom it is responsible.

"Care guidance is what we call our process. It means care managers have a fairly clear didactic approach to managing the individual so one case manager has the same tools as another," he says. "That way we're not purely relying on a skill set of an individual, but on a pathway of milestones and quality targets that's guiding them."

Systemically, Montefiore is not quite at its goal even now, which would mean taking full delegated risk through the government or the health plan, under which Montefiore would receive the majority of the premium. But even at that goal, the journey is not complete, says Rosenthal, acknowledging that there is a point of diminishing returns.

"As people live longer and as we do a better job of caring for them, costs will continue to rise," he says. "The way to best manage that is to continue to grow."

Montefiore manages about 450,000 lives at risk now, but Rosenthal says they need 1 million—the kind of scale where diminishing returns become less important.

"Price pressure is really affecting everyone," says Rosenthal. "Employers are pushing on insurance companies, and there have not been large increases in Medicare or Medicaid in the past several years. So as these pressures continue, the only way out is this model."

Ready, but few takers
Tricia Nguyen, MD, MBA, concurs with that sentiment.

However, the market she's in has yet to catch up with her health system. Fortunately, Texas Health Resources, her employer, sees a future where the ability to take full risk for populations, up to owning its own health plan, is critical to survival.

Nguyen is the system's executive vice president for population health and the president of its Texas Health Population Health, Education & Innovation Center. After experiences bringing risk contracting to Banner Health, and at two health plans prior to that, she came to Texas Health in 2013 charged with making the 24-hospital health system ready for risk.

Her early goals were modest, with each milestone intended to build upon the other.

"At the minimum, I was ensuring we would be able to manage the top 5% of the population whose risks were driving 50% of the costs," she says. "Every ACO should, at minimum, be starting at that point."

Upon assessment, Nguyen says Texas Health was already managing that cohort for its Medicare risk population, so there was some infrastructure in place for managing risk contracts from a complex care management approach, using the system's current suite of documentation systems from Epic, Allscripts, and Salesforce.

Over time, she says the organization will probably need a more robust tool to manage more than just the 5% and scale it, but is waiting to make that decision until the completion of the organization's clinically integrated network, which will include physicians and other clinicians at the hospital campus of UT Southwestern Medical Center in Dallas.

The joint operating company will put more than 2,700 providers under one network, called, in an amalgamation of both organizations' names, Southwestern Health Resources.

Asked to evaluate where Texas Health stands on its final goal of the ability to assume full delegated risk, Nguyen says the organization is at a 7 on the scale of 1–10, with 1 being completely new to risk-based contracting, and 10 the ability to take full risk.

"We have the assets to deliver on delegated risk minus a few assets like claims capabilities and actuarial analysis to do true product pricing to ensure we can manage to an allocated budget from the payers," Nguyen says.

"We believe that managing a percentage of health insurance premiums allows us to deliver high-quality care and services of high value, while keeping healthcare affordable."

Payers may be another story, and employers are perhaps even more reluctant for differing reasons. Nguyen says that some of the market's biggest payers seem to want to work directly with physician groups on risk contracting, and other payers with less market penetration are ready to delegate full risk to clinically integrated networks.

So payers are interested, but maybe not to the degree the organization had hoped, she says, and employers are even farther behind. Nguyen has to manage the balancing act between investing for readiness for risk, while not getting too far out in front of the trend such that the organization is delivering quality and outcomes benefits without being paid for the extra investment needed to meet those goals.

One of the biggest of those investments is in clinical nurse leaders who were introduced into Texas Health's workflow redesign for acute care management, which Nguyen has co-led since her arrival. Traditionally, case managers had been tasked with ensuring that the health system was utilizing resources efficiently and coordinating with the care teams and transitions of care.

In the past, those nurse managers were responsible for doing both tasks. In the redesign, 75%–80% of that team was redeployed to focus just on transitions of care—starting when a patient first enters a hospital—in order to better tackle the readmission risk.

"That was a $10 million investment all by itself for a new set of masters-prepared clinical nurse leaders," Nguyen says.

Texas Health has been part of UT Southwestern's Medicare Shared Savings ACO for a year, with about 45,000 lives attributed just to its physicians. Another 20,000 are attributed to UT Southwestern.

In commercial ACOs and its Medicare Advantage plan, Texas Health has full risk for about 30,000 lives, and Nguyen says the health system is considering applying for the Next Generation ACO program through the Centers for Medicare & Medicaid Services for 2018. She says she has seen a substantial decline in readmission rates and a slight decline in admission rates.

Nguyen says other health systems have to answer a chicken-or-egg question, as Texas Health has, regarding their approach to risk and the investments that are required to take risk.

"Do I invest in care management before I get risk contracts or the other way around? There are some simple things you can do that do not cost a lot of money," she says. "You do not need to invest in the sexy IT systems.

You don't need a high-tech Cadillac of workflow programs to manage populations. Have a conversation with your payer, whether it's insurance companies or employers, which I would encourage, to build something together."

Texas Health is not immune to the chicken-or-egg quandary. Because of reluctance from employers to narrow networks, one promising risk program has had to be shelved, at least temporarily. Forward Health Partners, an ACO launched with four other health systems to great fanfare in early 2015, is on hold.

"The reason it's on hold is we find employers in this market are not quite ready to adopt narrow network products," says Nguyen. "To take it off the shelf, employers have to engage and accept that it's not just a discount off of current negotiated rates. It has to support clinical interventions that will mitigate inappropriate use of resources. They're just not ready."

Until Texas Health can make greater strides in attracting full risk from employers and other payers, it will have to slow-play some of its clinical efforts on some populations.

"Clinically and ethically we don't treat everyone differently, but where interventions do differ is on wraparound services," she says.

"For Medicare beneficiaries who we're at full risk for, because there are dedicated care managers on that cohort, because there's a budget to do that, they call and do home visits. Otherwise they don't. The basic core of how we take care of patients is the same. However, resources are limited."