Practices hit Medicare P4P quality targets, but bonuses still fall short

New results from the second year of a Medicare pay-for-performance demonstration project for large physician groups are both raising hopes and casting doubts about the concept's feasibility.

Ten groups are participating in the Physician Group Practice Demonstration, a four-year effort that began in 2005. The practices implemented care coordination initiatives aimed at selected beneficiaries. For each year of the project, the groups can receive up to 80% of the savings they generate for Medicare by preventing complications and hospitalizations. Practices' bonus payments depend on these savings and their quality scores.

All 10 groups hit performance targets on at least 25 of the 27 quality measures included in the project's second year, which ran from April 2006 through March 2007, the Centers for Medicare & Medicaid Services announced Aug. 14. Five practices achieved the goals on all 27 measures, which relate to treatment of heart failure, coronary artery disease and diabetes mellitus.

However, only four groups qualified for performance bonuses, which totaled $13.8 million. After the first year of the demo, only two of the 10 practices did well enough to get back any money from Medicare.

Still, several of the participants said the demo is worth it.

"Even though we continue to lose revenue and incur additional costs, we think this project is one of the most successful undertaken by Medicare," said Samuel Carlson, MD, senior vice president and chief medical officer at Park Nicollet Health Services in St. Louis Park, Minn. "It showed that if you align the incentives properly, you can improve care." Park Nicollet did not receive a performance payment in the project.

The demo's weakness, the groups said, is in its "efficiency" component. CMS originally said practices could share in any of the savings generated. Shortly after selecting the 10 groups, however, the agency said the savings on care for enrollees assigned by Medicare to the project would have to exceed 2%, compared with a community control group, before any payout occurs.

In the project's second year, the groups saved Medicare a total of $34 million on assigned beneficiaries, but because of the threshold, CMS recognized only $17 million, said John Pilotte, senior research analyst at the agency's Medicare Demonstration Program Group.

The practices also extended care coordination to Medicare enrollees outside of the project but did not receive credit for any savings generated through higher quality care of those patients, several participants said. So the true Medicare savings were much higher than the CMS figures.

"The problem is we have to take all the financial risk," said F. Douglas Carr, MD, medical director, education and system initiatives, at the Billings Clinic in Montana. The clinic did not receive a performance payment.

The Billings Clinic has spent about $700,000 per year in direct costs during the demo and has lost potential revenue from hospitalizations and other complex care. Likewise, Park Nicollet is spending about $750,000 per year on efforts to support the program, as well as forgoing about $5 million worth of hospital services and other care, Dr. Carlson said.

Even at Marshfield (Wis.) Clinic, which received a nearly $5.8 million bonus in the project's second year, the costs outweigh the performance pay, said Theodore A. Praxel, MD, the group's medical director of quality improvement and care management.

CMS officials, however, are encouraged by the results. Pilotte, the demo's project director, pointed out that twice as many practices received bonuses this year and said the overall results are "proof of concept."

Only four groups received performance payments in the project's second year. However, all 10 groups received some money back because they agreed to put at risk their potential bonuses in the separate Medicare Physician Quality Reporting Initiative. While other physicians only had to meet CMS reporting requirements to qualify for PQRI payouts, the demo participants had to hit quality targets to do so. All of the groups received at least 96% of their potential PQRI payments.

Can bonuses work?

Robert Bennett, government affairs representative at the Medical Group Management Assn., said results raise concerns about whether the demo could be broadened to the wider physician community.

"It's taking a lot of money for groups to participate, and they're not getting it back," he said. "If the large guys can't do it, it makes you wonder if the smaller guys can."

The project shows that receiving a performance payment might be difficult if the government sets the bar too high, said Gail Wilensky, PhD, a former Medicare administrator. She predicted any more extensive Medicare physician P4P program would be "slower and gentler" than the demo.

The AMA has expressed concern that the opportunity for payments in the current project is based too much on savings and not enough on quality improvement. The AMA has developed extensive principles for all pay-for-performance programs.

While acknowledging the project's financial downsides, physicians were quick to point out its quality upsides.

"We learned a lot about how to be cost efficient and deliver high quality," said James Lee, MD, an internist who directed the Everett (Wash.) Clinic's participation in the effort. "It's an opportunity we wouldn't have had otherwise."

Everett Clinic added a nurse "coach" to ease patients' transitions out of the hospital and prevent re-hospitalization by going over discharge instructions with them in person and making follow-up doctor's appointments. It also partnered with a local hospice to provide outpatient palliative care.

The Billings clinic developed a report card shared with diabetes patients showing therapy goals for each quality measure and whether they had been met. The aim is to hold doctors and nurses accountable and get patients involved in care.

"When the patient has it at the point of care, it's very powerful," Dr. Carr said.