Home Analysis Closing the loop on readmissions

Closing the loop on readmissions

February 2009

Published in the February 2009 issue of Today’s Hospitalist

A PATIENT WITH DIABETES is discharged on an overly aggressive glycemic regimen, despite the discharging team’s concerns about her fragile home situation and ability to comply with medications. She is readmitted three days later with severe hypoglycemia.

Another patient, first admitted for weight loss and colitis, is readmitted with renal failure. The cause? An inappropriately high medication dose she was given to take at home.

Hospitalists are under a lot of pressure to shut down the discharge-readmission cycle. But a recent study that looked at how often both discharging and readmitting physicians communicate about patients who bounce back “even in the scenarios cited above ” found that in many cases, communication simply does not take place.

Researchers characterized the resulting gap in key patient information as an “open-loop” system in which individual physicians receive only minimal feedback about diagnostic errors or flawed therapies. They also noted that the failure to talk about readmissions may perpetuate medication errors and continuity problems.

“Some readmissions are perfectly reasonable or are for an entirely different medical problem,” says Christopher Roy, MD, the study’s lead author and acting director of the hospitalist service at Brigham and Women’s Hospital. “But there’s almost always something you can learn about the care you provided.”

Curiosity trumps feedback
In the observational study, Dr. Roy’s research team tracked 225 general medicine patients at two academic centers in Boston who were readmitted within 14 days of discharge.

Researchers found that discharging physicians were aware of the readmission in less than one-half (48.5%) of cases. Yet in a follow-up e-mail survey of both discharging and readmitting physicians, researchers learned that nearly all (97%) of the discharging hospitalists and other team members would have appreciated being notified. When communication between the two teams did take place, 61.1% of readmitting physicians said they obtained valuable patient information.

“Most of the time, ‘curiosity’ trumped ‘feedback about my care’ as the reason discharging hospitalists wanted the communication,” says Dr. Roy. “The discharging team almost always had information they wanted to share with readmitters, but the readmitting team wasn’t as interested in having that exchange.”

Why? For one, Dr. Roy explains, readmitting physicians assumed they had all the information they needed from the previous discharge summary. In addition, he says, “People don’t know what they don’t know. They think, ‘This is just another CHF exacerbation, so why do I need to speak with the discharging team?’ ”

According to Dr. Roy, however, even brief communication gives both discharging hospitalists and readmitters “a much more robust understanding of the issues that resulted in the readmission” “information that can also lead hospitalists to use different care strategies when the patient is discharged a second time. When the two teams did communicate, for instance, discharging physicians were able to provide information on psychosocial issues (52.6%), pending tests (34%) and discharge medications (30.9%), in addition to their overall patient assessment.

The study, which was published online in November 2008 in the Journal of General Internal Medicine (and is being published in the March 2009 print edition), was conducted in early 2007 at Brigham and Women’s and Massachusetts General hospitals.

Communication about readmissions would be particularly helpful for residents, says Dr. Roy, because they get “precious little follow-up” on the care they deliver and because they have many fewer opportunities “to see the natural history of a problem or condition.” But study findings would also apply, he says, to experienced physicians and to non-teaching settings.

While Dr. Roy notes that physicians in smaller hospitals may communicate more than their colleagues in large facilities, he adds that hospitalists everywhere face a unique challenge. “When physicians rotate on and off frequently,” he says, “they are going to have communication issues.”

Psychosocial information
What’s behind physicians’ inability to communicate about readmissions? According to Dr. Roy, it’s mainly lack of time. But he also points to physicians’ tendency to focus on the problems at hand rather than revisiting discharge scenarios to uncover possible errors, pending test results or factors that can complicate ongoing treatment.

Dr. Roy also says that the findings support the idea of collecting data on readmissions within hospitals and creating a forum for exchanging information. While he admits that setting up a system to exchange information on all readmitted patients would be too cumbersome, he would like to see hospitals tackle certain patient subsets.

“If there were some way to provide composite data to physicians on patients readmitted with the same diagnosis, that might give helpful feedback,” he says. Hospitalists and other attendings could benefit, for example, from knowing which percentage of their heart failure patients is being readmitted, compared to a certain benchmark. Or, Dr. Roy notes, hospitalists might want to target only high-risk or “frequent flyer” readmissions.

He was particularly surprised by how strongly psychosocial issues “such as substance abuse concerns or problems with difficult family members “affected inpatient and post-discharge care. Those details were often notoriously absent from discharge summaries.

“It was the kind of information that can be helpful to the readmitting team,” he says. While discharging physicians wouldn’t note in their summary that “Mr. X’s daughter ‘tortured us,’ ” he says, “that’s actually important information.”

A vote for automation
As for how discharging and readmitting hospitalists should communicate, Dr. Roy casts his vote for automation.

“We would need an automated feedback system rather than something hospitalists would have to do actively,” he says, “or people just wouldn’t take the time.”

Implementing an automated system should jumpstart a discussion, which would help hospitalists learn “not to make the same mistakes or reinvent the wheel,” he says. Physicians might also, he suggests, incorporate such intra-team communication into morning report or dedicated “readmission rounds” that would take place weekly or monthly.

In a randomized, controlled trial conducted after the study, Brigham and Women’s instituted automated e-mail notification of readmission to both discharging and readmitting physicians, which provided links to each other’s email. While the intervention improved awareness, Dr. Roy says, it didn’t significantly spur more communication.

“The bottom line is that you can lead a horse to water, but you can’t make it drink,” he says. “The main barrier to communication continues to be a lack of time.”

Bonnie Darves is a freelance health care writer based in Lake Oswego, Ore.