Medicare ACOs can learn lessons from earlier demo project

The precursor to Medicare's large and growing accountable care program ran five years and ended the month after the Affordable Care Act became law. The results, summarized in a new report, have proved prescient and may inform policy changes the CMS is expected to issue soon.

The final analysis of that experiment neatly summarizes data from other papers, all of which show Medicare saved a little money without compromising quality, albeit on a limited set of measures.

Policymakers widely cite the Medicare Physician Group Practice Demonstration, which ended in March 2010, as the forerunner to Medicare's accountable care initiatives under Obamacare.

Financial success varied widely across 10 medical groups in the demonstration. One saved an average of $818 per person, per year, while another saw per person, per year spending increase an average of $323, according to the analysis conducted for the CMS by not-for-profit research firm RTI International.

“We did find savings, which was one of the goals, but they were modest,” said Gregory Pope, a program director for RTI International who co-authored the report. “Not huge, but not trivial.”

On average, the 10 medical groups reduced Medicare spending by 2% per person, per year ($171) during the demonstration, which split the savings between the doctors ($102) and Medicare ($69, or 0.8% less than the program would have spent otherwise).

The results look familiar when compared with the early performance of more than 100 Medicare ACOs launched in 2012 under the healthcare reform law's shared-savings program. Their efforts also yielded mixed results and minimal overall savings.

Meanwhile, the demonstration program, like its ACO offspring, required participants to meet quality targets before they received bonuses. The quality targets seek to dissuade doctors and hospitals from cutting costs at the expense of good medical care.

And quality overall improved under the demonstration. Researchers found that to be the case even among those that saved the most. “It did not seem like the source of the cost savings was a decline in quality,” Pope said. (The quality results the CMS has published so far for ACOs in the shared-savings program are limited and mixed.)

Medical groups in the early demonstration focused their cost-saving efforts on the most expensive place to treat patients (the hospital) and the most expensive patients (complex, chronically ill and vulnerable). It worked. Spending for hospitalized patients dropped an average of $228 per patient, per year. Spending for patients at high risk of costly care declined $1,922 per patient, per year.

“It's not surprising,” Pope said. “That's kind of where the money is: people who are using services and are at risk of adverse health events and expenditures and high utilization.”

Researchers identified four strategies that, though not directly linked to savings, were widely adopted across the 10 demonstration medical groups: encouraging patients to do more to manage their medical care and health; establishing programs for doctors and nurses to more closely monitor and manage patients' care; using better coordination among hospitals, clinics and nursing homes as patients travel from one to another; and encouraging greater use of providers other than doctors.

Information technology also played a significant role. Even though one participant entered without an electronic health-record system and all but a few of the others lacked fully developed systems, the use of information technology was widespread, the report said. Medical groups developed and used patient registries and automated clinical alerts within their EHRs to track patients and monitor the quality of care.

The analysis of the group-practice demonstration reinforces that modifications to the shared-savings program could allow those strategies to flourish, said Stephen Shortell, director of the University of California at Berkeley Center for Healthcare Organizational and Innovation Research.

Medicare officials have been vocal in their desire to see the model continue to expand and are expected to soon release proposed rules amending the program.

For example, Shortell said, the CMS may do more to let providers and patients know who will be getting care under an ACO so doctors can do more to educate and encourage patients, known within the industry as patient engagement.

“That is huge,” he said. “It's just beginning to occur systematically across the country.”

As it stands, Medicare notifies doctors which patients were included at the end of the year, when performance is evaluated.

Shortell said Medicare might also raise the financial incentives for investment in care coordination and the use of providers who are not doctors. Success, however, will depend on development of multiple strategies. “There's no silver bullet here.”

Follow Melanie Evans on Twitter: @MHmevans