The poor results from a $1 billion investment remind one former policymaker of a mind-set that undermined an earlier high-profile initiative to improve care delivery.


March 28—Clinical interventions implemented by organizations funded through a high-profile federal healthcare innovation program provided no reductions in the cost of care, according to recent research.

In 2012, the Center for Medicare and Medicaid Innovation (CMMI) began to issue nearly $1 billion in Health Care Innovation Awards to more than 100 organizations nationally to test innovative care delivery approaches for patients in Medicare, Medicaid, and the Children's Health Insurance Program.

In the first broad assessment of outcomes from those participating organizations, a Health Affairs study found that, on average, the delivery innovations did not affect patients’ total cost of care. In fact, the per beneficiary fiscal effects ranged from quarterly savings of nearly $1,000 to increased spending of almost $750. Those figures were relative to mean quarterly spending in a comparison group of $3,458 per beneficiary. Obama administration officials said the first-year awards from the program were expected to save up to $1.9 billion in the first three years, according to published reports.

“The takeaway is this is really hard and we haven’t found any silver bullet,” said Gail Wilensky, former administrator of the predecessor agency to the Centers for Medicare & Medicaid Services (CMS).

Christopher Langston, PhD, vice president of healthcare services at the Aging in New York Fund, agreed that the bright side of the study may be to kill “magic bullet” thinking.

“Some people’s thoughts are just that simplistic: that we’re just going to come up with [a health information technology] magic bullet, or a telehealth magic bullet, or a behavioral health magic bullet; these are very gross categories,” Langston said in an interview.

Previous analyses of the program conducted for CMMI focused on the results of specific participants.

Specific Interventions

Among the 135 unique interventions used by awardee organizations, the researchers examined results from 43 of the most comparable ambulatory care interventions used over one to three years.

“These were a kind of ‘best of breed’ interventions funded all at the same time,” Langston said.

Estimated savings from the use of health IT, community health workers, and interventions focused on “clinically fragile patients” exceeded $150 per beneficiary per quarter, according to the study. Quarterly savings of about $100 were provided by medical homes and behavioral health programs. However, broad confidence intervals meant none of the savings were statistically significant.

“I don’t think this adds anything—we already knew that the only [beneficiaries] you can save money on are predictably expensive,” Langston said. What was needed was insights on matching the interventions with patients who needed them.

Wilensky said in an interview that the lack of statistical significance, owing to the small sample size of some interventions, meant that even interventions showing the greatest savings could actually provide no savings.

“So it’s not shocking, but it’s discouraging,” Wilensky said. “Finding savings even in our system that appears to be so bloated relative to other countries and efficiently run health systems has been mysteriously hard to achieve.”

When asked for comments on these overall results, a CMS spokesperson referred to the agency’s posted case-study-type results for the program.

Meanwhile, the study found spending increased with the use of telemedicine, interventions focused on the “socially fragile,” and new programs that providers launched with the award funding.

The poor financial results from addressing the social determinants of health may reflect less experience in that area or just the costliness of those increasingly popular programs, said Rachel Block, a program officer with the Milbank Fund.

“The thing about social determinants is if you’re going to improve health, if you’re going to improve outcomes, you may have to spend more money in order to organize these services in the community and create those linkages,” Block said in an interview. “It’s likely just a factor that those are bigger up-front investments and they may take longer to show cost savings results.”

Despite finding a lack of financial benefit from community-support services, such as providing housing and air conditioning, the study did not prove that they don’t work, according to Langston.

“What it tells us is, ‘Don’t think it is automatically going to work,’” Langston said. Similarly, “don’t expect just to pay for teleconsultation and get better results.”

The poor cost outcomes for telemedicine, specifically, could reflect the wide variation in how providers are using the technology in their clinical settings, Block said.

“You need to know a little more about how” it’s being used, Block said.

Net savings from more interventions may materialize over a longer time period than the authors studied, Block said.

Previous Disappointments

The disappointing results reminded Wilensky of underwhelming results from other ambitious CMS care delivery initiatives, such as the Physician Group Practice (PGP) demonstration. Of the 10 large, experienced physician groups that opted to participate in the PGP initiative, only one consistently met CMS’s 2 percent annual savings threshold.

“That was really astounding to me,” Wilensky said.

She recalled the explanation of one provider’s CEO, who said his employees were unwilling to implement changes needed to drive quality and savings improvements because they didn’t think the need for them was permanent.

The latest research regarding cost savings “is just a reminder that it sounds easier than it actually is proving to be,” Wilensky said.

Program Design

Some of the poor outcomes of clinical interventions stemmed from the design of the innovation awards program, said Langston, who was an expert reviewer during the program’s second round.

“This kind of analysis should have been done before the healthcare innovation program opened,” Langston said.

Langston cited the “ludicrous” emphasis the Obama administration placed on whether applicants would create new jobs when evaluating whether to provide them an innovation award.

“What should have happened was CMMI, its contractors, and national experts should have put their heads together and said, ‘These are things that we think are actually relevant to better ambulatory care,’” Langston said. “And they should have been more directive to the audience about what kinds of interventions to test and how to put them together—rather than leaving a totally blank slate.”

Langston said the lack of selectivity left awardees to use approaches that “don’t work,” such as telephonic case management and interventions from the “failed” Medicare Health Support Disease-Management demonstration and the earlier Medicare Coordinated Care demonstration.

“I don’t think that the time permitted or the review process was really sufficient to ensure that we were looking at the best of the best and not just trying to retest the flat tire from before,” Langston said.


Rich Daly is a senior writer/editor in HFMA’s Washington, D.C., office. Follow Rich on Twitter: @rdalyhealthcare

Publication Date: Tuesday, March 28, 2017