The endless war on readmissions
Hospitalists who study readmissions discuss the current state of the evidence and offer their best advice on avoiding patients coming back to the hospital.
On the front lines of the battle against readmissions, it's been hard to gain any ground lately.
In January 2020, the New England Journal of Medicine published an analysis of frequently hospitalized patients in New Jersey. It found that readmission rates were similar (over 60%) in those who received an intervention from the Camden Coalition of Healthcare Providers, a group nationally known for its work in this area, and in patients who got usual care.
More recently, a multidisciplinary, multicomponent transitional program for high-risk patients in Chicago actually found more readmissions in its recipients than controls, according to results published by the Journal of General Internal Medicine (JGIM) on Sept. 1.
Such findings are disappointing, but not surprising, to those working in the field. “Readmissions have been a vexing problem in health care,” said Vincent X. Liu, MD, an intensivist with Kaiser Permanente and a researcher with its division of research in Oakland, Calif. “Everybody's had a swipe at this problem, but I don't know that we've identified, as an entire health care ecosystem, any perfect solutions.”
They're not raising the white flag, yet, though. “It's just a really complex topic,” said Evelyn Chang, MD, a physician and health services researcher with the Veterans Administration in Los Angeles. “The patients are really complex, and our interventions are really complex, so it's hard to have black-and-white answers of what works and what doesn't work. Otherwise, we would have succeeded a long time ago.”
Readmission experts recently walked ACP Hospitalist through some of these complexities, offering their tips for hospitalists in practice and their hopes and predictions for the future of readmission prevention.
The good old days
When hospitalists first began working to reduce readmissions, progress was rapid.
“My work with readmissions started a long time ago, 2009 or 2010, … before the government started their readmission reduction program,” said Nancy Dawson, MD, FACP, a hospitalist and researcher with Mayo Clinic in Jacksonville, Fla. “Initially, we did find some things that did work. For instance, that was back when we put teach-back education in our nursing education program.”
Educating patients about their medications is one of a few interventions that has been definitively shown to reduce readmissions, agreed Valerie Press, MD, MPH, FACP, a hospitalist at the University of Chicago who researches readmission prevention in patients with chronic obstructive pulmonary disease (COPD). “There's the interprofessional team with clearly defined roles, care across settings, … medication reconciliation,” she listed.
Many times a hospital or service has put such pieces together and succeeded in reducing readmissions. But then another facility may try the same thing without any effect. “We often find that the solutions might not be as portable or generalizable as we had hoped,” said Dr. Liu.
It's easy to imagine at least one reason why these interventions don't transfer perfectly, noted Bruce Henschen, MD, FACP, lead author of the JGIM study and assistant professor of medicine at Northwestern Medicine in Chicago. “The needs of patients here may be very different from the needs of patients in a smaller city or a rural area,” he said.
Even at a single center, what worked before to prevent readmissions may stop working, Dr. Dawson noted. “Probably the thing that surprised me the most is it takes multiple continual interventions and changes in those interventions to keep on top of the needs our patients have in order to transition well to home,” she said.
There's also the challenge of figuring out what part of an intervention was responsible for its effects. Dr. Dawson offered the example of her program's recent success reducing readmissions by placing equipment to measure blood pressure, heart rate, and pulse oximetry in patients' homes (published by JGIM in January and covered by ACP Hospitalist in May).
“I'm doing a subgroup analysis of our study right now looking at the patients on telemetry who had a nurse phone call. Were their readmission rates lower than the patients on telemetry who didn't have a nurse phone call?” she said. “Was it that actual contact with a person that made their readmission rate lower?”
Divide and conquer
Researchers have modified and targeted their readmission interventions in all kinds of ways to try to answer these kinds of questions. “We've done it by service line, we've done it by diagnoses, we've done it by the patients who are readmitted in the first seven days,” said Dr. Dawson.
Developing readmission interventions for specific diseases has shown some success. For example, not knowing how to use an inhaler properly has been found to be a common pitfall that can lead patients with COPD to be readmitted, Dr. Press said.
Dr. Henschen's research identified sickle cell disease as a diagnosis worth targeting with its own interventions. “We found that sickle cell patients had a lot of readmissions. Across the American health care system, there's a huge gap in sickle cell care,” he said.
Dr. Dawson had another preferred disease target. “Probably the best example is heart failure, because heart failure usually has specific patient education and requirements for patient care at home,” she said.
Even within a diagnosis, though, patients may have very different needs, Dr. Liu noted. “As clinicians, we recognize that, right?” he said. “They really need one very specific type of intervention, whether that be patient education about weight gain and salty foods … some kind of physical intervention, mobilization … [or] to be able to pick up the phone and speak to someone.”
Interventions that are disease-specific can also backfire. “If you're too focused on one disease, you may not understand how you're affecting other diseases,” said Dr. Press. For example, suppose a COPD-specific intervention leads to prescribing a patient steroids to reduce the risk of readmission for COPD— that could worsen control of diabetes.
Such comorbidity is common among the frequently admitted patient population. “A lot of them have nonspecific diseases or multiple diseases at once that interact with each other, and those are the more difficult patients to come up with one or two things you can intervene on,” said Dr. Dawson.
There's also the issue that the cause of many patients' readmissions may have less to do with their illnesses than other aspects of their lives. “We do such a great job in the American health care system of categorizing people's medical needs, but the social needs are often more difficult to capture,” said Dr. Henschen.
Patients' social needs can also be more difficult to fix. “How do you right-size, create the best way to tackle things such as unstable housing, lack of social support, isolation, inability to get transportation?” he asked.
Tackling problems that a clinician or hospital is likely to be able to fix is a critical requirement for successful readmission prevention, the experts noted. “Focusing on what we can actually identify as a modifiable factor might be all we can really do,” said Dr. Chang. “Everything else might be trying to streamline our efforts.”
As researchers work toward that goal, they have debated how to use the available data predicting patients' risk of readmission. “The question is, do you target the high-risk patients, or is it better to target the intermediate-risk patients, because you might be able to make a difference?” Dr. Dawson said.
The highest-risk patients are the source of the most readmissions, of course, but they may not have as much in the way of modifiable risks as less sick patients, Dr. Chang noted. “It could be that the only thing we can offer them at that point is advanced care planning,” she said.
Ideally, interventions should target patients earlier in their illness trajectory, before they have already had multiple admissions. “It's easy to find someone who's definitely going to come back,” said Dr. Henschen. “The challenge is trying to identify the patient who maybe presents to the hospital for the first time and is at risk of coming back.”
Dr. Liu is optimistic about the potential of computers to help with this problem. “This is where we believe in the promise of big data, of predictive analytics, of machine learning and artificial intelligence,” he said. “These types of technologies excel at finding data-driven differences within patients and then trying to maximally target the right intervention to the right patient.”
This strategy will only work if the right data are collected, noted Dr. Chang. “These models might be a little better if we had more social determinants of health risk factors included in our electronic health record,” she said. “Generally we're not able to capture some of these things.”
Steps to take now
Hospitalists don't have to wait for perfect data to be collected and interventions to be developed in order to combat readmissions, though. The experts offered their advice on what individual clinicians can do in their care for patients.
These include many of the evidence-based interventions, such as high-quality medication reconciliation (“not just check the box,” said Dr. Press), which addresses any obstacles to medication adherence and access.
Research has also shown a follow-up visit soon after discharge to be a critical step, one that some hospitals have had to take over if patients don't have good access to primary or specialty care. “We'll often provide at least that visit, and then we can also help to see them for a couple of more visits until we can get them transitioned back to their care team,” said Dr. Press.
Dr. Henschen's top tip was for hospitalists to talk explicitly about the issue of readmission with patients. “Patients know that they've been in the hospital a lot and are willing to talk about it, but oftentimes it's not brought up,” he said. “So my first piece of advice would be just to ask, and think about that stigma.”
The conversation may identify steps that can be taken to reduce the patient's likelihood of readmission (“Think about out-of-the-box solutions,” advised Dr. Henschen) but even if it doesn't, the interaction may help to build the patient's trust in the health care system.
“Because they're being taken care of by so many different people and maybe get different messages from everyone, there's often a lack of trust,” said Dr. Henschen. “Work on trust building with patients.”
Trust isn't built in a minute, and constraints on hospitalists' time are a major challenge with readmission prevention, Dr. Dawson noted. “As a physician at the bedside of the patient, I think I have the most effect on the day of discharge,” she said. “Unfortunately, so many times a hospitalist is rushed through that, and that's the worst time to be rushed, honestly.”
She offered a few examples of the topics it's critical to slow down and cover. “Make sure the patient understands their disease and understands what it would take to get them to be successful at home … and know who to contact if they have questions.”
That contact may prove to be the most important readmission intervention of all, Dr. Dawson speculated. “We're seeing more and more patients just need that continual contact and reassurance, and I think that's where most programs are going,” she said.