Intervention lowers hospital readmissions

Coaches empower patients in self-care

Medicare spends about $17 billion a year on hospital readmissions that could have been prevented, experts say.

About one in five Medicare fee-for-service patients are rehospitalized within 30 days of being discharged from the hospital. In the best-functioning hospitals, about 5% to 8% are readmitted; in some hospitals, nearly one-third of Medicare patients have a readmission within 30 days post-discharge, says Stefan Gravenstein, MD, PhD, clinical director of Quality Partners of Rhode Island and professor of medicine and community health at Brown University, both in Providence, RI. Quality Partners is a nonprofit organization established to be a Medicare improvement organization for the state of Rhode Island.

Gravenstein and other researchers decided to test an intervention to reduce 30-day hospital readmissions among this population. It is based on the transitional care work and randomized controlled trial by Eric Coleman, MD, MPH, professor of medicine with the Divisions of Health Care Policy and Research and Geriatric Medicine with the University of Colorado, Denver.

The intervention involves having coaches meet with and call patients to empower them to access community providers when their symptoms begin to show trouble, rather than waiting until they are very sick and need to be hospitalized.

"Eric Coleman, who is a geriatrician and science professor, demonstrated a few years ago that if you taught patients basic skills, they could self-manage and speak up for health care when they needed it," Gravenstein says. "He called it the care transitions intervention and demonstrated that it resulted in more than a 30% reduction in readmissions among older patients."1

The logical follow-up to Coleman's work was to try the intervention in a real-world setting, which is what Gravenstein and co-investigators did in a new study that found a significant reduction in 30-day readmission rates for patients ages 65 years or older. Individuals who received the intervention had a 30-day readmission rate of 12.8%; those who did not receive any part of the intervention had a 20% readmission rate. And an internal control group of people, who declined to participate or who were lost to follow-up before having a home visit, had a readmission rate of 18.6%.2

In addition to educating patients and empowering them to be more proactive with their health care needs, the system needs providers to be ready to see patients immediately, in order to work, according to Gravenstein.

"The backdrop system has to be ready so when doctors get the phone call from patients, they can say, 'Yeah, we have a spot for you,'" Gravenstein says. "Hospitals have to notify primary care physicians and give them information that supports successful coaching."

"As a real-world intervention, we wanted to offer this to as many people as possible, given our resources," says Rachel Voss, MPH, program coordinator of Quality Partners of Rhode Island.

"We did find similar to Coleman's results a 36% reduction in the readmission rate when compared to people we had never approached about this intervention," Voss says.

The study selected a random sample of the targeted population, but was not designed as a randomized controlled trial.

"As a Medicare-funded pilot program, we hired coaches to work with six hospitals and work with any patient who was cognitively intact and discharged from the hospital to the community," says Rosa Baier, MPH, senior scientist at Quality Partners and a teaching associate at Brown Medical School.

The coaching intervention was based on the four pillars of Coleman's model: medication management, a patient-centered record that the patient maintains for transferring information to various providers, timely follow-up appointments, and watching for and responding to red flags or warning signs and symptoms.1

"The coaches do not do the work for the patient, but empower patients to take care of themselves," Voss says. "They guide people at home through medication reconciliation, and they teach them to reach out to their physicians so they can self-manage properly."

The intervention spans a 30-day period and has the coach make a hospital visit, a home visit within three days of discharge, and two follow-up telephone calls within the first week and the first four weeks post-discharge. Patients receive a booklet for recording their personal health record, including their main health problems, medications, and questions for their doctors.1

Coaches are not the same as nurses, Gravenstein notes.

"When you send a nurse to the home, the nurse may notice the patient has swollen legs, and the nurse might have the patient increase the water pill," he says.

"Coaches, instead, help patients recognize when something is going wrong and how to reach into the provider system to get the help they need," he adds. "So if the coach sees the patient, and the patient says, 'My legs are swollen; what do I do?' The coach helps them reason through that problem and realize that it's okay to call the doctor and arrange for an appointment in the next couple of days."

Often Medicare patients will ignore their symptoms or put off a doctor's appointment until the problem is exacerbated and requires an emergency room or urgent care facility visit, Gravenstein says.

"They say, 'I can just wait,'" he explains. "For an 80-year-old with heart failure, that means a 911 call."

For purposes of the study, the coaches were nurses and social workers, Voss says.

"In theory, they don't need to be nurses because they're not supposed to use nursing skills," she adds. "We have other projects similar in style to this intervention where coaches are not nurses; they have some familiarity with the health care system, but they don't have the level of background as nurses, and they're still as effective at the intervention of empowering patients."

Coaches teach patients to use their common sense, Gravenstein says.

"If your toe is swollen, you don't need a nurse to tell you that your toe is swollen and somebody should take a look at it," he says. "You need common sense to say, 'I need to have someone take a look at this.'"

People who are health literate already have these self-empowerment skills, he adds.

"You want the patient to own these decisions about when to generate an encounter with the doctor's office," he says. "You're teaching them to fish rather than giving them a fish."

Outreach is a main skill patients are taught. Also, coaches push patients to make those appointments sooner rather than later, Gravenstein says.

"The coach's job is not to catch every problem that arises, but to teach the patient to recognize and do outreach for help when it arises," Baier says.

Patients receive the personal health record in the hospital where they are taught how to understand their medication list, Voss says.

"Patients write down the medication information in their personal health record, so when they get home they have a written record in their own handwriting that they can match up with all the bottles on their table and in their cupboards, as well as with their discharge instructions," she explains.

Patients can share their personal care journals with their community providers. This helps facilitate information transfer across health settings, so that providers have the right information about a patient's health status and recent hospitalization, Baier says.

They also can list their symptoms and warning signs in the record.

"It's a touch point for communication for health care providers," Gravenstein says. "Our role is to engage these various providers to make sure these things happen."


  1. Coleman EA, Parry C, Chalmers S, et al. The Care Transitions Intervention: results of a randomized controlled trial. Arch Intern Med. 2006;166(17):1822-1828
  2. Voss R, Gardner R, Baier R, et al. The Care Transitions Intervention: translating from efficacy to effectiveness. Arch Intern Med. 2011;171(14):1232-1237.