Medicare demonstration projects: From idea to implementation

Today's Medicare demonstration project might determine how physicians interact with the program tomorrow, so naturally such projects generate a lot of buzz in the health care community. Lately, some of that buzz has been about how demos might provide part of the solution for Medicare's payment problems.

At a conceptual level, Medicare demonstration projects appear well-suited to the payment reform issue. Federal officials and health professionals agree that the current fee-for-service system based on the sustainable growth rate formula is not working, but there is no consensus on what to do about it. Testing alternatives on a smaller scale could help determine how the system should be replaced without the need to commit to a concrete plan from the outset.

The American Medical Association and other physician organizations have called on lawmakers to prevent annual scheduled pay cuts to allow the Medicare agency time to test possible new payment models. Partial capitation, patient-centered medical homes and private contracting were among several ideas the AMA suggested in May that the Centers for Medicare & Medicaid Services try out.

When appropriate, the Association supports allowing these new ideas to be tested live with physician practices through demonstration projects, said AMA President Peter W. Carmel, MD. The process gives CMS the flexibility to adapt proposals and ensure the best results for patients and physicians.

"New initiatives proposed by CMS must be reviewed and tested to determine how they will affect physicians and patients before they are fully implemented," Dr. Carmel said.

Other organizations have pushed the demonstration process to serve as a payment reform test bed. Rarely, however, does one of Medicare's payment experiments become the next big thing throughout the program. Some demonstrations will offer a new benefit to patients or a new incentive to doctors. But CMS has tested scores of ideas over decades, and even many with promising results survive today as little more than archived links on the agency's website.

"They frequently inform but don't lead to new legislation or direct changes in behavior," said Gail Wilensky, PhD, a former administrator of the Medicare agency.

For a variety of reasons, Medicare's process of testing ideas has not produced the revolutionary change that's needed, Wilensky and other policy experts said. The champion for a project might leave CMS by the time a demo is completed. In other cases, political considerations might stop legislation that would authorize an expansion of a demo throughout the program.

Medicare's participating heart bypass demonstration in the 1990s often is cited as an example of a promising demo that hasn't been implemented more widely. The project bundled fees for bypass surgeries into a single, negotiated sum -- leaving hospitals and physicians free to divide payment as they saw fit and determine the best way to use the funding to care for a patient. All seven participating hospitals were able to shorten hospital stays and saved the program a combined $42 million by the end of the demo in 1996, but the concept was not taken further.

Because some demos end up in limbo, Wilensky is an advocate for the projects having specified trigger mechanisms. If a project is shown to be successful by improving the quality, cost or efficiency of health care, it should be implemented as Medicare policy nationwide, she said.

The legacy of demonstration

Not every Medicare demonstration project fizzles out. The CMS website lists 58 demonstrations, but the Medicare program has launched many more over 40 years. Doctors work daily under the legacy of demonstration projects whether they realize it or not.

The most prominent national payment policy so far that has its roots in a demo is the hospital diagnosis-related group system, said Stuart Guterman, PhD, a health policy analyst at the Commonwealth Fund, a progressive think tank in Washington. The system was adopted by the federal government to pay for Medicare inpatient services starting in 1983.

The DRG system began as an experiment led by New Jersey in the 1980s to create more efficiency in hospitals. The project set payments based on patient diagnoses, and the results impressed officials and lawmakers in Washington. Today all hospitals are paid based on DRGs.

The Premier hospital quality improvement demonstration, which ended in 2006, was a pay-for-performance project that successfully proved that bonuses could prompt health professionals to report quality measures. The effort led to the physician quality reporting system, which has netted physicians hundreds of millions of dollars in bonuses. Quality reporting through the PQRS will be mandatory starting in 2013.

Sometimes demonstrations prove successful from the perspective of program administrators even when they are unpopular with participants. The Medicare recovery audit contractor demonstration that started in 2005, in which auditors scrutinize past claims to identify overpayments, saved the program so much money in its initial three-year run that lawmakers enacted a law expanding the RACs to all 50 states by 2010.

The physician group practice demonstration starting in 2005 has been a lightning rod for debate about whether shared-savings pay models work. In that demo, seven of the 10 participating groups achieved perfect quality reporting rates by the final year of the initial run, with others close behind. But only four groups returned enough savings to Medicare to earn bonuses.

Despite the uneven results, the demonstration was a precursor to Medicare accountable care organizations, a voluntary shared-savings initiative introduced throughout the program by CMS this year, said Vince Kuratis. He's a consultant in Boise, Idaho, who has worked with health professionals on care management issues that often are found in demos.

Federal officials running Medicare demos are looking for a good return on investment, but only a small percentage do very well, Kuratis said. In failing, some of the other efforts push CMS to go in another direction. "They are growing bacteria in a petri dish," he said. "Some work, and some don't show anything."

The beneficiaries of good demos

In 2006, Massachusetts General Hospital and the Massachusetts General Physicians Organization were among those picked to participate in a three-year demonstration project to improve care for high-cost beneficiaries. Patients identified for the program were of poor health, took many medications for their chronic conditions and had frequent hospitalizations.

Under the demo, the Massachusetts collaborative got monthly fees of $120 per patient to coordinate care for about 2,500 high-cost beneficiaries, according to a final report prepared by RTI International, an independent research group.

Physicians would not have been able to coordinate care in this way without the upfront financial and data support from CMS, said Eric Weil, MD, medical director for the care management program. His group hired 12 case managers to work with primary care physicians. The managers developed personal relationships with the patients who agreed to participate.

From the standpoint of both Medicare officials and the collaborative, the project was a success. The patients' hospitalization rate was lowered by 20%, and emergency department visits dropped 13%, according to the RTI report. The program saved Medicare $2.65 for every $1 spent on management fees.

The demo also let the hospital, physicians and care coordinators design a system addressing the particular needs of their community, Dr. Weil said. Such flexibility was not supported by the traditional fee-for-service system, he said.

Dr. Weil cited two examples of how the demonstration improved patient care while saving Medicare money.

An 85-year-old man with brittle diabetes stopped taking his insulin and metformin because of a misunderstanding about his prescription drug coverage. His primary care physician and case manager could not determine the Medicare Part D policy assigned to the patient, so they referred him to a care management pharmacist. The pharmacist identified the plan and provided the patient that day with a three-month supply of his needed medications.

A 91-year-old woman with pneumonia was brought to the hospital for observation and later discharged with orders for antibiotics, nebulizer treatments and visiting nurse services. When the case manager's routine follow-up call 24 hours later failed to reach her, the case manager called a relative, who found the patient at home suffering from lung congestion and fatigue. The process led to an urgent primary care visit and a readmission to the hospital.

"Although no admission was avoided, the close follow-up and expedited triage of the patient hopefully prevented a much more complicated admission," Dr. Weil said.

More physicians soon will have the chance to see if such a demonstration can help their patients and practices. In September, the Center for Medicare and Medicaid Innovation launched the comprehensive primary care initiative, which will offer doctors patient management fees to coordinate their care beginning in 2012.

CMS hopes the innovation center, authorized last year by the health system reform law, provides even more opportunities for doctors to realize the promise of the Medicare demonstration project in finding a better way to pay physicians.

The innovation center can use a $10 billion budget over the next decade to extend the life of several demonstrations as pilot projects and other initiatives, said Bruce Vladeck, PhD, a former Medicare administrator. He sees the new center more closely focusing demonstration efforts on creating new national payment policy.

"That's a much better way to do things and a real change in course from over the last 20 years," Vladeck said.