Updated

An ambitious effort to cut costs and keep aging, sick Medicare patients out of the hospital mostly didn't work, a government-contracted study found. The disappointing results show how tough it is to manage older patients with chronic diseases, who often take multiple prescriptions, see many different doctors and sometimes get conflicting medical advice.

The study showed just how hard it is to change the habits of older patients and their sometimes inflexible doctors. And it points up the challenges the Obama administration will face in trying to reform health care for an aging nation.

Most of the patients had serious, but common, age-related illnesses including diabetes, heart disease and lung disease. Programs were set up at 15 centers around the country. Only two cut the number of times these patients were hospitalized, and those are still in operation. None saved Medicare any money.

The authors of the study called the results "underwhelming." An editorial in the Journal of the American Medical Association, where the study appears Wednesday, used the term "sobering."

"The only way you can really do it is by changing patients' behavior and by changing physicians' behavior, and both things are really hard to do," said study author Randall Brown, a researcher at Mathematica Policy Research Inc., in Princeton, N.J., which was hired to evaluate the programs.

Often, these patients need to stop smoking, or lose weight, exercise more, eat healthier foods — a challenge even for generally healthy people. Those changes are especially tough for sick, older patients who often are set in their ways.

"The same thing with physicians," Brown said. "A lot of them feel like they know how to take care of patients, so why do they need a nurse calling up and asking them why the patient isn't on some certain medication?"

Many patients in the study had more than one chronic disease, a common Medicare scenario. In 2002 alone, half of Medicare patients had been treated for five or more ailments, and they accounted for 75 percent of Medicare spending, the study authors noted.

Seeking ways to reduce those costs and improve care, the Centers for Medicare & Medicaid Services selected 15 proposals for test-site programs in 2002. The sites developed their own programs, enrolling a total of 18,309 fee-for-service Medicare patients through 2006.

About half got the patients got the usual care. The others got more intensive, coordinated care. That often involved nurses who acted as go-betweens, helping doctors give patients clear, appropriate advice; counseling patients on changing bad habits and recognizing worrisome symptoms. The nurses were available on a regular basis by phone or in person to answer patients' questions.

Jim Reid, a 74-year-old retired Pennsylvania welder, was among study patients who got coordinated care.

When he enrolled in 2002 in a test program run by Health Quality Partners, a nonprofit group in Doylestown, Pa., he was obese, had high blood pressure, high cholesterol and pre-diabetes.

But Reid was a rare success story.

He actually took the advice offered in group sessions run by nurses. He learned how to read food labels and avoid salty, calorie-laden foods. He also started exercising, walking with a pedometer and building up to a few miles daily.

Now, he breakfasts on oatmeal or vegetable omelets instead of coffee and doughnuts He's lost almost 60 pounds. His blood pressure and cholesterol have greatly improved and his pre-diabetes is gone.

Sticking with the program "is hard," he acknowledged. "As you get older, you don't want to do it." But he said it has "put an extra 10 years in my life."

Reid credits his success to the personal attention of a nurse coordinator.

"I have to have somebody to own up to," he said.

That close, in-person contact with nurses was also a feature of the project's other more successful, still-operating program, at Mercy Medical Center-North Iowa in Mason City, Iowa.

In both programs, each patient had face-to-face contact an average of about once a month with a nurse. That was far more frequent personal contact than in other programs. Both reduced hospitalizations — 17 percent yearly compared with usual-care patients at Mercy, and by about 20 percent in the Pennsylvania program, but only among its sicker patients. That program worked with Doylestown Hospital and recruited patients from area physicians' offices.

Targeting sicker patients and providing frequent in-person contact show the approach has some benefits and that success with future reform efforts "is possible, but it's not easy," Brown said.

Peter Ashkenaz, a spokesman for the Centers for Medicare & Medicaid Services, said the agency is evaluating the Iowa and Pennsylvania programs to see if their positive results persist.

He said there are other approaches being tested, some that offer incentives to doctors who meet quality benchmarks, or who use electronic health records to improve quality.

But so far, Ashkenaz said, "as the study shows, we have not yet found broad success."