Tackling the very high costs of big health-care users

Imitation is the highest form of flattery—and it may end up helping greatly reduce the highest source of health-care costs in the U.S.

A new project is aiming to identify the kinds of treatments and models of care that lead to the best results for so-called high-need patients, and then try to replicate those results on a broad scale.( Tweet This)

Such high-need patients, often senior citizens and frequently suffering from multiple chronic health conditions, with multiple medical providers treating them, represent a very large fraction of total health-care spending in the U.S.

Treatment for an estimated 5 percent of the nation's population is responsible for about half—or $1.4 trillion—of total health-care spending annually, according to the National Institute for Health Care Management.

But some of those patients afflicted with heart disease, diabetes, and mental illness and other conditions end up with better health results, and at lower costs, than others. That phenomenon is at the center of the project being funded by the Peterson Center on Healthcare.

"What we want to do is find those programs that do it better, both from an outcome and cost standing ... and spread those features across the country," said Jeffrey Selberg, executive director of the Peterson Center.

Doctor and patient healthcare
Dan Dalton | Getty Images

"We're not interested in just spending less money and sacrificing the outcomes," Selberg said. "We know that there are better outcomes at less money. We like to say that quality and costs are mutually inclusive, that higher quality will lead to lower costs."

The 18-month-long project consists of three grants totaling $2.7 million to the Harvard T.H. Chan School of Public Health, the Institute of Medicine and the Bipartisan Policy Center.

Harvard's work will form the foundation of the project. It will focus on an analysis of medical claims data from high-need patients to figure out differences in how care was provided to them—and "how those differences led to variations in outcomes and costs, and how much of that variation is preventable," the Peterson Center said in a press release detailing the project.

Selberg said that "there's a high degree of variation in quality and cost" in the treatment of high-need patients. That wide variation, he said, suggests there is plenty of room for improving the way high-need patients are treated.

Read MoreOregon dumps cold water on Obamacare rate requests

The Institute of Medicine will then conduct a research review of Harvard's work, looking at the models for providing care to high-need patients, and the challenges in treating them. The Bipartisan Policy Center then will look at the findings of the prior two grants, and determine what current policies may present a barrier to implementing the treatment models that work best on a larger scale across the nation.

The policy-related questions are crucial because some of the high-need patients are among the 9 million people who are eligible both for Medicare and Medicaid, government-run health-care programs whose payment rules are often complicated.

"Dually eligible patients often face more complex health problems and require more care than individuals who qualify for just one of these programs," an issue brief by the Peterson Center noted. "For example, they are significantly likely—in some cases twice as likely—than all other Medicare beneficiaries to have pulmonary disease, stroke, congestive heart failure, diabetes, or some kind of mental or cognitive disorder."

Although the Peterson Center is spending a relatively low amount of money on the project, Selberg said that the savings from reducing health costs that potentially could result from the work "can be very significant."

Persistent problem

The unusually high costs from high-need patients have been known for "probably two decades," said Ashish Jha, director of the Harvard Global Health Institute.

And while there have been prior efforts to reduce the costs associated with them, the problem has persisted, he said.

"A vast majority of interventions have been, I think, very simplistic, and haven't targeted the patient population very carefully," Jha said. "They ended up not saving much money at all."

"What's been interesting is that people assumed this is a very monolithic bunch of patients," Jha said. "What became apparent to me was that this wasn't a single monolithic group of people that required one intervention, and we'd solve the problem."

"It's a complex problem," he said. "It's a really varied group of folks."'

Read MoreCalif. to gather data on its Obamacare customers

Jha said he expects there is going to be "skepticism" among some medical providers when the project identifies treatment models that are doing a better job at others when it comes to caring for high-need patients, and seeks to have the models adopted elsewhere.

"We all think we do our jobs about as well as we can be doing them," he said. "But in the back of our heads, we all know there is something we could be doing better."

Reducing costs for everyone

Jha said the project has the potential for reducing costs and improving care not only for high-need patients but for everyone else in the health-care system.

"If you can fix the system for the neediest, the most vulnerable patients, the really sick people, you're going to make care better for everybody," Jha said.

On the other hand, Selberg said, if the problem of high-need patients isn't addressed with better treatment models being adopted, the effects will be felt by the entire nation.

Read MoreOh, it hurts bad...to schedule a doctor's visit

"The costs will become increasingly unaffordable for individuals, the costs will increasingly eat up our ability as a nation to invest in other things, and the life expectancy of people with multiple chronic conditions will get worse," Selberg said. "There's a lot at stake here."

But "the deal here is we can do it better," Selberg said. "It's being done now."