Where: Baylor University Medical Center, Dallas.
The issue: Reduce readmissions and emergency department visits among elderly patients recently discharged from the hospital.
Personal experience and literature review had shown hospitalist Andrew L. Masica, MD, that certain elderly patients were highly likely to return to Baylor after discharge. Not only was this frustrating for physicians and patients, but it seemed likely to become a problem for the hospital as well.
“Medicare is putting an increasing priority on prevention of readmissions,” said Dr. Masica. Dr. Masica had also noticed that these patients needed special attention in a number of areas—general care coordination, medication reconciliation and counseling, and follow-up after discharge.
“In our normal process, they do have a care coordinator and medication reconciliation is done by nursing. If needed, a clinical pharmacist is available, but their interaction with the patient is limited due to resource constraints. With an elderly patient and a lot of things going on, things sometimes slip through the cracks,” he said.
Dr. Masica, along with colleagues in care coordination, administration and pharmacy, wanted to address several of these problem areas at once. “You can have single elements and maybe get some sort of impact but the results have been inconsistent. What we wanted was to apply all those elements concurrently so in a sense you'd have a safety net,” he said. “We knew that you have to approach these high-risk patients from a multidisciplinary standpoint.”
Their solution was a pilot project to provide a “care bundle” to elderly patients at particularly high risk for returning to the hospital.
How it works
The project targeted patients who were admitted to one of Baylor's hospital medicine groups and who met certain criteria: age 70 or older, used five or more medications, had three or more chronic comorbid conditions, and required assistance with at least one activity of daily living. The patients also had to have been living at home or in assisted living and be admitted for one of 20 medical diagnoses common in Medicare patients.
“Identifying patients who were at risk for readmission was one of the key elements,” said Dr. Masica. “Obviously, you can't do this level of high-intensity intervention to every patient.”
The intervention included daily visits from a care coordinator, who provided education with an emphasis on home self-care, and from a clinical pharmacist, who reconciled and explained medications. “Throughout the hospital stay the [pharmacist] would spend some time with patients if medications were changed, if doses were changed. We were trying to empower patients to understand what they were on and why they were on it,” Dr. Masica noted.
Both the coordinator and the pharmacist also called patients once after discharge, reinforcing the education provided during the hospital stay and helping to resolve any problems or questions that may have arisen after returning home.
The Baylor pilot, results from which were published in the April 2009 Journal of Hospital Medicine, randomized 41 patients who met the set criteria to receive either the care bundle or usual care.
Thirty days after discharge, the intervention group had significantly lower rates of readmission or emergency visits (10% vs. 38%, P=0.04). By 60 days out, the difference between groups had become insignificant (30% vs. 43%, P=0.52). Among the patients who did come back to the hospital during the study follow-up, the time between their discharge and return to the hospital was longer in the intervention group (36.2 days vs. 15.7 days, P=0.05).
“It does show you need to think of patient care as a continuum rather than dividing into inpatient and outpatient settings,” said Dr. Masica. The findings particularly confirmed the importance of communication with patients about medication, he added. “For the patients in the usual care arm, a very common reason for re-presentation to the emergency room or readmission was some issue related to medication. If a hospital needed a place to start, that would be a high-yield area.”
The study design of the intervention posed a challenge when it came to getting patients enrolled in the pilot with a formal informed consent process. “Presenting a multiple-page document that discusses risks to an elderly patient didn't help recruitment,” said Dr. Masica. He would recommend that anyone doing a similar project in the future use either a waiver of informed consent or structure it as a quality improvement project.
Cost might pose another hurdle to hospitalists hoping to follow the Baylor team's example. “To do the intervention that we set up in the study, you might need an additional full-time care coordinator focused on the high-risk patients and also an additional clinical pharmacist,” said Dr. Masica.
His study didn't include any evaluation of cost-effectiveness, but he theorizes that the money spent on coordination and pharmacy would be recouped in overall savings by eliminating later hospital visits.
The decline in the project's effectiveness after 60 days showed the limitations of an effort conducted entirely within the hospital, according to Dr. Masica. “I think that would be expected from an intervention that you just deliver on the inpatient side,” he said. “Linking that to the outpatient care would probably be the next phase.”
For Baylor, the follow-up has been an intervention similar to the one used in the pilot project, but this time targeted at a specific disease. Last August, a transitional care initiative for congestive heart failure (CHF) was launched. The voluntary program is offered to all CHF patients over age 65 and includes additional education (by a nurse practitioner instead of a care coordinator) and enhanced medication reconciliation.
“That's probably how we'll carry this model out down the road,” said Dr. Masica. “This was a pilot study to see if this worked. The long-term focus is on specific disease areas and disease management.”
How patients benefit
The benefits to the patients who avoided readmissions through the project were pretty clear, said Dr. Masica. “It's much better for the patients not to have to come back to the hospital.” The experience of participation was also positive. “Talking with patients and families, they did like that extra education and enhanced discharge planning.”
Words of wisdom
“I wouldn't claim that we have solved the transitions of care problem. We view this as hypothesis-generating,” said Dr. Masica.