Nobody wants to see chronically-ill elderly patients making repeated trips to the emergency room. These visits are obviously bad for the patients themselves, who often suffer stress, disorientation, and high risks of infection. They are no good for Medicare, which has to pay the bill: The estimated cost of these readmissions is $17 billion annually. And, despite the common perception, they may not be good for hospitals, which are facing payment cuts for these patients.

Many of the families I met while researching Caring for Our Parents struggled with this problem. Remembering which medicines to take; learning that it is a bad sign when a congestive heart failure patient like my dad gained weight; even something as basic as being able to get to the doctor for a checkup could be the difference between staying at home and ending up in the ER at 3 AM.   

But keeping frail seniors (and others, for that matter) out of the hospital turns out to be hard. One in five Medicare patients is readmitted within 30 days. Although Medicare has been running several demonstration projects aimed at keeping this population out of the  hospital, most have failed. According to a new study by Randall Brown at the highly respected consulting firm Mathematica Policy Research, very few experiments have reduced either hospitalizations or costs.

But the news isn’t all bad. Brown identified three models that seem to be working. Two focus on patients who are already in the hospital. In one, advance practice nurses begin to work with patients before they are discharged and then follow them intensively for 4-6 weeks after they go home. In the second, for 4-7 weeks after they are discharged, patients are enrolled in programs that train them to self-manage their care.  

The third model identifes patients at high-risk for hospitalizaton and intensively monitors their symptoms before they get so sick they end up in the ER. These projects, called the Medicare Coordinated Care Demonstration program, have met with mixed success. Only 3 of 15 have reduced hospitalizations and costs. But Brown says they can work if they carefully target patients, if nurses and social workers spend face time with those in their care, if nurses work closely with primary care doctors and carefully tailor services to these patients’ needs. Two keys: teaching how to manage medications after discharge and helping arrange for non-medical social assistance, such as transportation and adult day care.

The problem is that setting up these programs is not cheap. Massachusetts General Hospital says it spent as much as $9 million to get its model going, although it says it eventually saved between $7 million and $10 million after covering costs.

With Medicare expected to cut hospital payments for patients it believes could have avoided that round-trip to the emergency room, Brown’s lessons will be increasingly important to both hospitals and patients. The trick will be to find ways to not only save money but improve care. If that sounds like a smaller version of the big health reform story, that’s because it is.