Fewer dollars, more sense
Oregon’s Medicaid experiment looks to stop paying docs per procedure, start tying payment to quality

When it comes to health care, the money trail usually goes like this: You go to the doctor’s office and the physician gets paid a set amount for the exam. If the doctor prescribes a scan, the company that performs it — in some cases, the doctor’s own practice — gets paid for that, too. If the doctor refers you to a specialist, that specialist gets paid some predetermined amount for an exam.

Most medicine in the U.S. is delivered according to this financial model, called “fee for service.”

But for many of the estimated 636,000 Oregonians who will be covered by the Oregon Health Plan on Jan. 1, the money will soon be distributed very differently.

The Oregon Health Plan is our state’s version of Medicaid, a health insurance program for low-income adults and children. And it’s already the subject of an unprecedented experiment in health care reform.

Last year, the federal government agreed to give Oregon $1.9 billion to help it launch a pilot Medicaid reform experiment. The goal? To reduce projected Medicaid spending by $11 billion in 10 years, and slow the rate of spending growth by 2 percentage points in two years. Also as part of the agreement, the state has to improve the quality of health care that it provides.

While Cover Oregon, the state’s health insurance exchange, continues to hog news headlines for its inability to enroll people, Oregon’s other health care experiment — Medicaid reform — remains largely below the radar, even as it enters its second year.

The stakes are high for all Oregonians. If the Medicaid reform effort doesn’t live up to its lofty aims, the state faces penalties of up to $183 million per year — money that would have to be scraped from the state’s already thinly spread budget. If it does meet its goals, the experiment will have produced something that has otherwise eluded experts across the nation: A path toward improving health while curbing the skyrocketing costs of care. If that happens, these new financial models could be adopted by private insurance companies to become the new standard of medicine in Oregon.

Even if it never shifts into the private market, this experiment will have serious repercussions. For some, personal health is at stake. Sixteen percent of the state’s population is covered by the Oregon Health Plan. Half of all babies born in the state are covered by the plan.

For doctors, the risk is also personal. Eighty-five percent of all doctors in the state see Oregon Health Plan patients. For these physicians, reform could mean less money to cover their rent, pay their employees and take home at the end of the day.

And all Oregonians have an interest in seeing the costs of Medicaid contained. The Oregon Health Plan makes up 11 percent of the state’s total budget, and is the fastest growing portion of the state’s expenses.

The bedrock of Oregon’s Medicaid experiment is a system of Coordinated Care Organizations. Across the state, there are 17 certified CCOs, each charged with carrying out the Oregon Health Plan in a particular region.

The state contracts with the CCO — here in Central Oregon, it’s a branch of Pacific­Source, the insurance company — that, in turn, contracts with all of the medical providers in the area. Each CCO gets flexibility in how it works with providers and spends money. But a board of directors — made up of provider representatives as well as local officials and even some OHP members — has to approve the CCO’s decisions. So St. Charles, Bend Memorial Clinic, Deschutes County Mental Health, local dentists and plenty of other caregivers are all active participants in this Medicaid experiment.

The idea behind CCOs was to coordinate care at every level. So far, many of the changes have had little to do with finances. Electronic medical records are being integrated throughout the region, for example. That means when an Oregon Health Plan patient is admitted to the hospital, the care navigator, nurse or primary care physician in charge of that person’s case is automatically notified, to make sure follow-up care is provided and the patient is less likely to be readmitted.

So far, most of the news headlines about CCOs have focused on a strategy of shifting patients from the emergency room to primary care offices. By getting a patient’s asthma under control, for example, the patient will no longer have to rush to the emergency room during an asthma attack, racking up costly hospital bills.

Preliminary data suggests that changes enacted so far are making a difference, albeit a small one. In the first year of the experiment, the state reported that the rate of growth in Medicaid spending — not the amount of Medicaid spending, but the rate of increase — dropped 1 percent.

To make a greater dent in medical costs, however, something seismic has to happen.

The state is requiring CCOs to modify how they pay health care providers. The exact payment models are still up in the air. But one thing is certain: Eventually, Oregon Health Plan providers will not get paid per procedure, as they have in the past.

Here in Central Oregon, the contracts between PacificSource and many local providers are now being negotiated in closed door meetings. By July 2014, Central Oregon’s CCO must show the state that it has made progress toward shifting away from fee-for-service payments. By 2015, the new payment models have to be embedded in the CCO’s contracts with local clinics.

Maryclair Jorgensen, director of payer relations and contracting for St. Charles, said at a presentation in Portland this fall that, from the beginning, Oregon’s Medicaid reform experiment has balanced on two pillars: transforming care and transforming payment.

“You can’t have one without the other,” she said.

Along with changing how providers get paid, CCOs are also being required to meet new standards in health care quality — or, at least, the data that purportedly supports it. The state has developed a new set of benchmarks aimed at gauging the quality of care. Some of the measurements are to be gleaned from patient satisfaction surveys, others from medical data — the percentage of Oregon Health Plan adolescents who go in for annual well-care visits, for example, or the rate of pregnant women who receive timely prenatal care. While the effectiveness of this measurement system is still an open question, Medicaid providers must soon begin submitting data to show how their treatment of patients measures up against the state standards.

Early adopter: Mosaic Medical

Mosaic Medical, with clinics in Bend, Redmond, Madras and Prineville, specializes in providing coordinated care for low-income patients. This year, 41 percent of the practice’s patients were members of the Oregon Health Plan, and another 34 percent were uninsured, said Elaine Knobbs, a spokeswoman for the clinic.

For its Medicaid patients, the clinic receives some money from PacificSource, the local CCO. In addition, Mosaic Medical also receives Medicaid money directly from the state. Since February 2012, the clinic has been using an alternative payment model for the money that comes through PacificSource. In March, the clinic began using an alternative payment model for the state Medicaid money, too.

Instead of waiting for a Medicaid patient to come in for an office visit, then submitting a claim and waiting nine months for reimbursement, the clinic now receives a set sum per-member, per-month.

This hasn’t altered the clinic’s annual financial picture, according to chief financial officer Diane Kayser. But it has given the clinic greater flexibility in how it delivers care to patients.

For example, the clinic now operates a secure online patient portal, called MyChart. Patients who don’t need to visit the clinic can go online to type questions to a nurse or doctor, request prescription refills or access their medical records.

“In the past, the only way we got paid for serving those patients was if the patients came in,” Knobbs said.

The electronic system saves patients a trip to the doctor’s office — which often means avoiding time off and lost wages — and it saves space in the clinic for patients who do need to come in.

The problem with ‘fee for service’

Michael Bailit, a health care economics consultant from Massachusetts, said he’s been in the business long enough to see all kinds of fads come and go. Competition in the marketplace, regulation of the industry, competition and regulation, HMOs, preventive care, wellness programs and a long list of other approaches have, at one time or another, been touted as the solution to rapidly rising costs.

“There’s not one answer” to all of our health care woes, he said. “But addressing the payment model (should be) part of the solution because our traditional payment model is certainly part of the problem.”

The problem with fee for service, he said, is that it encourages care that doesn’t necessarily benefit health. Instead, practitioners have a financial incentive to deliver more services, especially services with a high profit margin.

Paying providers per unit of service is “inherently inflationary,” Bailit said. “And it doesn’t reward desired behaviors.”

The standard model doesn’t encourage practitioners to work together to create a multidisciplinary plan for patient health, for example. It doesn’t foster a patient-centered approach to health. And it doesn’t reduce the rampant overuse and misuse of medical treatments. According to federal estimates, errors and waste account for 30 to 40 percent of all health care spending.

The vast majority of health care in the U.S. is paid using a fee-for-service model.

“And not only that, it’s fee for service without any other incentives or disincentives tied to performance,” Bailit said.

The reason for reforming the payment model for health care is not only to reduce costs, according to Bailit. The idea is to design a payment model in which the financial incentives to providers align with the goal of delivering high quality care.

According to Bailit, there are four alternative financial models:

Pay for performance

A pay-for-performance model offers two types of financial bonuses to medical practitioners. One bonus is tied to performance, or the quality of care delivered. So once the state’s health benchmarks are set, data is collected from insurance claims or a clinical audit and providers receive bonus payments for meeting or exceeding the state benchmarks.

Another bonus is tied to services performed, similar to fee for service. So each time a desired service is delivered — a well-check for a child or adolescent, for example — the provider receives a set payment.

The benefit of this system, according to Bailit, is that it’s relatively simple to understand and administer, requiring minimal changes to clinics that currently operate under a fee-for-service model. This is important because Oregon Health Plan members make up just 10 percent of total patients, on average, at Oregon medical clinics. It can be costly to operate one small slice of the business according to a totally different set of rules and practices.

But this financial model has its drawbacks, too. In most cases, Bailit said, it hasn’t been shown to actually lower costs. It can even raise costs by providing bonuses for services that don’t help improve health quality.

And, as with fee for service, the incentives are tied to patient volume. In other words, clinics have a financial incentive to see as many patients in as little time as possible. Bailit said it’s nearly impossible to create this kind of bonus system and have it overcome the simple economics of volume.

Infrastructure payments

Another alternative involves upfront payments to hire new staff or develop new programs that help improve health quality and reduce waste.

One benefit of this system is that it addresses some of the early costs involved in shifting the strategy of a medical practice. A primary care office could use the money to hire a nurse, for example, to coordinate care for patients with especially complex cases. This could reduce waste and lessen the workload for expensive — and often overbooked ­— physicians.

This system is also pretty simple for a clinic to administer alongside a traditional fee-for-service portion of the practice, Bailit said.

But it assumes that the funds will be well-invested by the practice and will produce positive, long-lasting outcomes. In reality, Bailit said, that doesn’t always happen.

Episode-of-care payment

Sometimes called a bundled-care payment, this is a shared payment across multiple care providers for an isolated episode of care. In the case of a hip replacement, for example, one pot of money would be paid to the hospital, physicians and physical therapists.

For chronic conditions such as congestive heart failure, a shared payment would be made — to primary care providers and cardiologists — over a certain period of time.

This system, according to Bailit, provides incentives to reduce errors and waste and to coordinate with other providers. It rewards caregivers for finding effective, efficient treatments.

But it doesn’t encourage providers to decrease the incidence of episodes. In that sense, it’s very similar to the old fee-for-service model of care.

In addition, this system is complicated to administer. Worse, it could actually discourage providers from accepting high-risk patients or patients with complicated medical conditions.

“Ideally, you would build something into this system to deter cherry-picking (of patients) and spread out the risk,” Bailit said.

Population-based payment

There are two versions of the population payment model, which differ only in terms of how the financial risk is distributed.

Both versions begin with a payer — PacificSource, for example — giving to a provider — St. Charles, perhaps — some “expected” spending amount for a population of patients.

According to one version of this model, sometimes called “up-side risk,” if spending falls below the expected amount, both parties share the savings. Health quality data is usually entered into some predetermined formula to determine how to divvy up the money.

This model creates financial incentive for the parties to work together and cut costs. By strategically incorporating the right data into the savings-sharing formula, it can balance the incentive to save money with an incentive to improve quality.

But in this version, neither party faces any real negative risk, according to Bailit. The best incentive to save money, he said, is to face the prospect of losing money.

This model is extremely difficult to implement on a small scale, Bailit added. With a patient population of less that 5,000, it’s nearly impossible to pull off. To calculate expected payments and collect health quality data for all of the patients is complex and costly.

The “shared-risk” version of the population payment model, sometimes called “global payment,” is similar to above except that providers lose money if they exceed the expected spending amount for a defined population.

According to Bailit, this model creates the strongest incentives for care providers to reduce costs. But providers need quite a bit of infrastructure to be able to manage care of a population within a budget — they need sophisticated data analysis methods, and high-level managers to oversee this type of clinical practice. If a medical clinic doesn’t have these capabilities, the financial losses could put the clinic out of business, which would harm all kinds of patients, including those not on the Oregon Health Plan.

Choosing a model

For each model, Bailit said, there are real life examples of success and failure, so it’s difficult to generalize about which one is best.

In every case, he added, details matter. The best payment model is a means to an end — better, more efficient care. And it doesn’t help to change incentives if the provider can’t respond to it, or doesn’t know how.

“The design is only half of the issue,” Bailit said. “The other half is execution.”

It’s important to consider a payment method all the way down to the individual clinician, he said. If a CCO uses an alternate payment methodology with a physician group, does it translate to individual physician compensation? In most cases, the answer is no, he said.

That means the incentive disappears as soon as the exam room door is closed. And that’s a problem, he added, because it means even the most well-designed incentive is unlikely to change behavior at the clinical level.

So what will Central Oregon’s CCO do? How will local practices be paid in the coming years for seeing Oregon Health Plan patients?

It’s too soon to say. Representatives of PacificSource, St. Charles, local medical practices and other health care groups declined to talk on the record until contracts are signed.

Dan Stevens, senior vice president of PacificSource and head of the local CCO, said the group is trying to develop financial systems that align with the community’s health-related goals.

“All the parties are nervous,” said Jorgensen, of St. Charles, during a presentation at a CCO conference in Portland this fall. “There’s not enough money to go around, it’s an underfunded program. But we understand that if we don’t make a transformation in payment, the system is going to fail … We’re all still at the table.”

St. Charles has four “medical homes” — and a fifth opening in Madras next year — that provide coordinated primary care for about 6,000 Oregon Health Plan members. Next year, that number will likely rise to about 8,000.

During her presentation, Jorgensen said that one financial option under consideration involves giving to St. Charles all of the per-patient Medicaid money for its 8,000-or-so patients. As a full-service organization, offering everything from primary care clinics to a hospital, the organization could then decide how to divvy up the money among its own departments. If this were to happen, St. Charles would take responsibility for solving the financial puzzle of its own piece of the pie while still meeting the health-related and budgetary benchmarks passed down from the state.

If this option is selected, it opens the door for other financial models to be experimented with next year, too, for the thousands of Central Oregonians who belong to the Oregon Health Plan but don’t go to St. Charles for primary care. So the region could soon end up with several experimental models in the works.

‘Not fun and not easy’

Last year, a dispute between a Salem hospital and its local CCO threatened to undermine the entire Medicaid transformation effort. For months, Willamette Valley Community Health and Salem Hospital argued over how much the hospital should be reimbursed for its Medicaid patients.

“Every CCO has a ‘How do you split up the pie?’ discussion,” said Dean Andretta, executive director of the CCO. “They are not fun and not easy.”

In Salem, it resulted in a lawsuit and a new bill being introduced. Officials from CCOs all over the state warned lawmakers that the legislation, which would have forced the hospital to drop the lawsuit, could dissuade hospitals from participating in the transformation effort, unraveling the entire reform. In the end, the lawsuit was dropped and the bill died.

“It was a snag we hoped to avoid,” said Ruth Rogers Bauman, Willamette Valley CCO chairwoman.

“We’re talking about fairly major changes, it takes time to build the trust,” Bauman added. “And that has to happen first before you can make changes. Now, we’re there.”

Bauman said although Salem’s spat was out in the public, deciding how to divvy up the funds inherently causes tension at CCOs all over the state.

One major challenge, Dr. Bud Pierce, a Salem oncologist and past president of the Oregon Medical Association said, is finding out the real cost of doing business.

Before this experiment, hospitals essentially negotiated Medicaid reimbursement rates, Pierce said.

“Until now, everyone who wants to get paid for Medicaid would use their lobbyist and influence to get a favorable rate,” Pierce said.

With regionalized budgets, hospitals and CCOs need to find the true profit margin and what it really takes to remain solvent.

And it’s imperative to change the model, because resources aren’t unlimited.

“We’re being given a chance,” Pierce said. “And it’s very important for the medical community to realize it. Because if we can’t control our spending, economics will limit you.”

Whatever combination of models get selected, the whole Medicaid experiment is about to become even more complicated. One provision of the Affordable Care Act, also called Obamacare, calls for the expansion of Medicaid to cover more low-income adults. Children who don’t have other health care coverage are already eligible. Beginning Jan. 1, adults who earn less than 138 percent of the federal poverty level — about $15,800 a year for a single person or $32,500 a year for a family of four — will be eligible for the Oregon Health Plan. Previously, the program was limited to those earning less than 100 percent of the federal poverty level, or $11,490 for a single person or $23,550 for a family of four.

That means another 160,000 Oregonians — including about 9,000 in Central Oregon — are about to join in this experiment.

— Reporter: 541-410-9207; lraff@bendbulletin.com

Lauren Dake contributed to this report.

Click window to continue