Of Medicare’s 43 million beneficiaries in 2006, the 10% who were the sickest and had the most complex health conditions accounted for 67% of annual Medicare spending. Over the past decade, the government has funded dozens of demonstration projects aiming to improve care and reduce the skyrocketing costs for such patients, yet there has been little evidence of improvement. One of the few exceptions is the Mass General Care Management Program, which began in 2006 and was funded by the Centers for Medicare & Medicaid Services. Eric Weil, a primary care physician and the medical director of the program, discusses how it became so successful that it’s being extended—and the major challenges to implementing it in other locations.

Q: Why does it cost so much to treat these patients?
A: Patients in the first three-year phase of the program were, on average, 76 years old, were being hospitalized more than three times a year and were taking more than a dozen medications. More than half had behavioral health problems, and the mortality rate was 16%. Quite simply, they are sick and getting sicker, so they’re likely to require a lot of care.

Q: Do they just need more care than typical 76-year-olds?
A: They need thoughtful attention to how their care is delivered. When I see one of these patients for the first time, I identify his or her key health issues and set up a care plan. I can prescribe all the right medications and make the necessary referrals, but it’s unclear after the patient has left my office that everything will go according to plan. That’s where the nurse care managers come in. I (and the other 200-plus participating physicians) work with them to coordinate patient care. I’m confident that when a patient leaves my office, there’s another person who knows the patient just as well as I do and that a potential gap in care is being addressed.

Q: The nurses make a lot of phone calls to be sure patients are taking their medications?
A: Yes, but they do much more. I depend on care managers to do whatever’s necessary to keep a patient out of the hospital. The care manager not only coordinates a patient’s care with specialists but also works with a team that includes mental health professionals, pharmacists and social workers. For example, I have an elderly female patient who used to be admitted to the hospital every three to four months because of congestive heart failure. Now we have a nurse care manager calling her regularly and potentially even making home visits. She monitors a program we’ve put the patient on and works with a pharmacist to adjust her medications and make sure she’s taking them at the proper times. The patient hasn’t been back in the hospital for a year—amazing.

Q: With all that costly follow-up, resulting in improved patient outcomes, how did the program also achieve financial benefits?
A: Better care results in less need for expensive hospital services. There was a 20% decline in hospital admissions and a 13% drop in emergency department admissions during the program’s first three years. CMS invested more than $8 million in the program; the savings from keeping patients out of the hospital have covered that amount and saved potentially an additional $7 million to $10 million. Essentially, for every dollar spent, the program saved at least $2.65.

Q: Were there disappointments?
A: We were less successful in reducing the rate of readmissions. We make every effort to address gaps in care, improve quality and maximize coordination. But this is a very ill population of patients. Despite all of our good work, we can’t halt the progression of illness.

Q: Can this be a model for the entire country?
A: The model is doable in most places; it would just need to be adjusted for every institution. However, if this program were in a remote location with small practices, which might require that nurse care managers work for (and drive to and from) several practices rather than be embedded in one, and with no electronic health record network, that would introduce a lot more challenges in coordinating care.

Q: How might such programs be funded eventually?
A: The MGH demonstration project experimented with modifying the traditional fee-for-service model by adding a monthly management fee to support the additional care that we deliver to this population, paid by Medicare. Ideally, health reform will support the development of models to encourage care coordination.