Hard data lacking on postdischarge follow-up

Effectiveness of hospital clinics and phone calls examined.


Don't call Jeffrey Greenwald, MD, “the readmissions guy.”

“That sits like a lump in my throat,” he said at Internal Medicine Meeting 2016. “I'm not a readmissions guy. I'm a care transitions guy.”

To that end, at his session, “The Evolving Role of Hospitalists in Care Transitions,” Dr. Greenwald, who is an associate professor of medicine at Harvard Medical School and Massachusetts General Hospital in Boston, offered attendees a hard look at the data on a specific area of care transitions: postdischarge follow-up.

He first addressed postdischarge clinics, which he said have historically developed largely due to problems of access to primary care. For example, “You call up and after you listen to 10 minutes of Muzak they say, ‘How's 2 weeks from 2019 for an appointment?’” Dr. Greenwald said. “This isn't Primary Care Practice Bashing 101...but it is the reality most of us face.”

Jeffrey Greenwald MD Photo by Kevin Berne
Jeffrey Greenwald, MD. Photo by Kevin Berne

Postdischarge clinics can also be used to improve continuity of care with the hospital care team, Dr. Greenwald said, for example if a clinician feels a patient needs to be seen 1 more time because of concerns about asthma. But, he said, “It raises the question of whether or not these clinics are really just a Band-Aid on a broken primary care system. And I will tell you as a bit of a spoiler alert, that's in fact what a lot of the practices that have opened postdischarge clinics have discovered.”

Dr. Greenwald gave a brief overview of the existing literature on postdischarge clinics, first pointing to a study published in the Journal of General Internal Medicine in 1996 finding a statistically significant difference in ED visits between patients who visited a postdischarge clinic and those who did not, but no significant differences in 30-day readmissions, length of stay, or mortality.

A more recent study published by the Journal of Hospital Medicine in 2014 looked at a primary outcome that was a composite of readmissions, ED visits, and mortality in patients who had a postdischarge visit at an urgent care center, their primary care physician's office, or a hospitalist-run postdischarge clinic. No differences were seen between the postdischarge clinic patients and those who saw their primary care physicians or between the postdischarge clinic patients and those who went to an urgent care center. When each of the variables was looked at individually, readmissions were highest in the postdischarge clinic group, and no differences were seen in ED visits or mortality.

“Their primary outcome was negative. They couldn't demonstrate a benefit,” Dr. Greenwald said. “At the best, there was sort of a wash.”

Dr. Greenwald pointed out that the literature on postdischarge clinics has traditionally been complicated because patients are not randomly assigned; they go there because they can't get into a primary care clinic or there is no primary care clinic available to them. “They're not random. They may be sicker, they may be older, they may be poorer, they may be more complicated, they may be from some other socioeconomic area where they're underserved....Whatever the issue is, these are not apples-to-apples comparisons,” he said.

He listed the results of his own nonscientific survey of postdischarge clinics, noting that of the 15 he was familiar with, 40% have shut down. “They've shut down because they're not financially viable, they didn't have enough patients to go to them, there were some conflicts with primary care practices around them, etc.,” Dr. Greenwald said. “The point is at the end of the day a lot of these, which sounded initially like really good options, failed.”

Regarding postdischarge appointments, “How many of you were told by your resident that this patient with heart failure, when you were an intern, has to see Dr. Smith by next Thursday or his head will explode?” he asked the audience. “We had to have an appointment on the books, and it had to be soon, darn it, because otherwise you just know this patient's going to fall apart. Well, the problem is the literature doesn't really corroborate that experience.”

He discussed a study published in the Journal of the American Medical Association in 2010 that looked at the relationship between early physician follow-up and 30-day readmission among approximately 300,000 Medicare beneficiaries hospitalized for heart failure. The study assessed hospital-level follow-up rates for 7-day postdischarge appointments by quartile and found that while the lowest-performing quartile had the highest readmission rate, there was no difference among the other 3 quartiles.

“So as long as you're not in the bottom quartile, it's OK to be in the second worst. Well, that's not very reassuring, is it?” Dr. Greenwald asked. Yet another study published in Medicare Medicaid Research Review in 2014 found no effect of postdischarge visits on 30-day readmissions among approximately 500,000 Medicare patients with acute myocardial infarction, congestive heart failure, and pneumonia.

Some evidence does suggest that medically complex younger patients may benefit from postdischarge visits, however, Dr. Greenwald said. In a study published in Annals of Family Medicine in 2015, researchers stratified 44,000 younger Medicaid patients (mean age, 26.5 years) by comorbidities and calculated a presumptive risk of readmission. In low-risk patients, expedited follow-up within 7 days made no difference in outcomes. In the higher-risk patients, however, expedited follow-up did seem to have an effect on readmissions, he said.

Postdischarge follow-up phone calls are also lacking definitive evidence support, according to Dr. Greenwald. He pointed to a review done in 2009 by the Cochrane Collaboration looking at hospital-based telephone follow-up after discharge in 33 studies involving 5,000 patients.

“They couldn't find any evidence of a readmission reduction. But is anybody surprised?” Dr. Greenwald said. “There are almost no studies where you do 1 thing and it changes readmission rates. You have to bubble wrap patients in lots of interventions if we're going to change readmission rates.”

A retrospective study published in Population Health Management in 2011, meanwhile, looked at the impact of a postdischarge telephone follow-up call on 30-day readmission rates. Among 30,000 patients who were part of a disease management program, almost 7,000 patients received a phone call from a nurse in the first 14 days after discharge. The peak readmission day was day 2 or 3 after discharge, and Dr. Greenwald noted that this has been seen in other trials as well. The study found that patients who didn't receive a postdischarge call had a 30% higher readmission rate.

This is at best an association, since causality can't be proved in this type of trial, Dr. Greenwald stressed, noting that his key takeaway from the study was the peak day for readmissions and what that could mean for the optimal timing of postdischarge calls.

“If you're waiting until day 14 to make your phone calls... you've already missed a lot of patients. And if you're going to intervene with phone calls, they shouldn't be 2 weeks out. Maybe they should be on day 2 if you hope to see an impact,” he said. “Don't wait until day 14. The horse is out of the barn.”

Dr. Greenwald stressed that he does not advocate abandoning all of these postdischarge interventions since they may directly help patients who experience other adverse events related to care transitions, such as medication errors and misunderstanding of treatment plans.

“No single intervention will likely dramatically move the readmissions needle alone,” he said. “These types of transitional care interventions should fit into a bigger menu of interventions you offer patients if you want to work toward more comprehensive transitions improvements, including reducing readmissions.”