Keeping patients with chronic obstructive pulmonary disease (COPD) out of the hospital is difficult, experts agreed at the annual meeting of the American Thoracic Society (ATS), held in Dallas in May.
Yet the financial incentives of bundled payments and readmission penalty programs have made it a major goal for hospitals and hospitalists. The expert speakers offered their experiences and ideas related to this challenge.
“Unlike other major diseases, for which hospitalization has gone down, COPD has remained pretty steady for the past 20 years or so. This would be relatively good news but for the fact that each of these hospitalizations is getting more and more expensive,” said Surya Bhatt, MD, an associate professor in pulmonary and critical care at the University of Alabama in Birmingham.
An initial hospitalization also tends to be followed by additional stays, with 30-day readmission rates for COPD hovering around 20%. “The financial and public health importance of reducing readmissions for COPD is substantial,” said Laura Feemster, MD, MS, an associate professor in the division of pulmonary, critical care, and sleep medicine at the University of Washington in Seattle.
In 2014, CMS gave hospitals an incentive to tackle this challenge when it added COPD to the list of conditions targeted by the Hospital Readmissions Reduction Program (HRRP). Subsequent bundled payments programs have provided additional impetus. Clinicians and researchers speaking at the conference described some of the resulting readmission-reduction research and outcomes.
Searching for a solution
Dr. Bhatt's hospital became an early entrant into the struggle by volunteering for CMS’ Bundled Payments for Care Improvement Initiative, which paid hospitals a preset amount for care surrounding a COPD admission, including readmissions.
“When we entered into an agreement with CMS to initiate this bundled program, we started looking at the literature to see what is it we can do to try and reduce readmissions,” said Dr. Bhatt. “We looked at these studies and most of them were not successful, [with] very heterogeneous interventions.”
Despite these discouraging indicators, his hospital applied a broad array of clinicians and strategies to the problem.
“We put together a multidisciplinary team. In addition to pulmonologists, we have a nurse practitioner and dedicated nurses as part of our team. And we also have care managers, home health, hospice and palliative care, and pulmonary rehab, and smoking cessation counselors. We also brought on social workers,” Dr. Bhatt said.
Every morning, the inpatient team would be alerted to the presence of any patients with admitted with COPD (or asthma, which was included by CMS to avoid gaming of the system by miscoding of COPD patients).
“We cast a wide net so we would not miss anybody,” said Dr. Bhatt. The identified COPD patients would then receive standardized care for an exacerbation, including steroids.
They were also interviewed by the team and provided with smoking cessation counseling, social services, comorbidity assessment, discharge planning, pulmonary rehabilitation, and palliative and supportive care as needed. For two weeks after discharge, the patients were called every day and then once a week for the following 10 weeks.
Dr. Bhatt and his colleagues published the outcomes of 78 initial COPD patients treated under this system in the May 1, 2017, Annals of the American Thoracic Society. “A lot of things that we value in our patients went up,” he said. Compared to a control group, the patients were more likely to get pneumococcal and influenza vaccines, home health care, durable medical equipment, pulmonary rehabilitation, and a visit with the pulmonary clinic.
However, one key metric didn't move much. “Despite all these interventions, our readmission rate went from 17.4% to 15.4%, which was not statistically significant,” said Dr. Bhatt. “For this specific purpose of trying to reduce costs for this hospital, either the volume of admissions was not big enough or the intervention did not work.”
The University of Chicago was another hospital that joined the first round of bundled payments for COPD, explained fellow speaker Valerie Press, MD, FACP, who is an assistant professor of medicine and pediatrics there.
“At the time, there was a lot of belief that bundled care and these kind of value-based purchasing options were going to expand and [our hospital administrators] wanted to learn how to provide this type of care for chronic disease,” said Dr. Press.
So her hospital, too, developed a multidisciplinary, multimodal intervention, which included physicians, respiratory care therapists, and dedicated nurse practitioners who provided inpatient care and outpatient follow-up, among others. “We tried to use evidence as much as possible,” Dr. Press said.
After implementation of the program, 30-day readmissions of COPD patients went down. “If you just look at that downward slope, you'd say, ‘Wow, that is really successful.’ But what we know nationally is before the penalty went into place at all, let alone for COPD, readmission rates were already declining nationally. And so it was likely some part of this was part of the national trend,” she said.
Based on the success of this initial trial, though, the hospital entered the bundled payment program, which included 90 days after a hospitalization. “The evidence was very limited on what to do beyond 30 days. . . . How often should we call a patient? How often should we bring them back in the clinic?” said Dr. Press. “One thing we were worried about is that maybe we had gotten much better at 30 days, but we were just going to push all the readmissions to 60 or 90 days.”
Their fears were not confirmed—readmissions at all three timepoints dropped. Unfortunately, that was not enough to make the program an unqualified success. “We spent more than we saved,” said Dr. Press. “If we just had those data, we'd say we failed to meet the cost objective.”
However, the program had a lot of additional data, including a natural experiment when the program's nurse practitioner role was under- and unstaffed.
“If you look at when we were fully running the program compared to that quarter when we were only partially running the program, we spent six times as much per episode. So it actually cost us more to not have the program running fully. Really impressively, when we weren't running the program at all, it was 15 times as much,” said Dr. Press.
The mixed results led to some discussion of whether the hospital should continue with the program and the payment model, but the decision was made to join another CMS initiative that will expand the number of outcomes that are assessed.
“Now they're going to be looking at how many of our patients have advanced care planning and other sorts of details. One of the main differences is instead of being compared to ourselves, we're actually being compared regionally,” said Dr. Press.
That poses the challenge of figuring out how to even measure all the costs of a hospitalization in real time; as a solution, the team developed a new cost-gathering dashboard. “We're optimistic that we're hopefully on the right track,” she said.
Identifying the culprits
So why do some hospitals manage to overcome these challenges and reduce readmissions, while others try and fail? Seppo Rinne, MD, PhD, has worked to answer that question, and he offered his findings to the conference attendees.
“We sent a survey to all the VA [Veterans Affairs] hospitals in the United States and asked, ‘What are the organizational practices?’ Some of them were organizational practices, some of them were structures, structures being like a COPD-specific clinic or a COPD-specific pulmonologist. We also looked at who had alerts, performance measures, incentives,” said Dr. Rinne, an assistant professor of medicine at Boston University.
When he and colleagues compared these aspects and practices of the hospitals to their COPD readmission rates, they found no clear connections—the associations turned out just as they would by chance, according to results published in the May 1, 2017, American Journal of Respiratory and Critical Care Medicine.
“We're not seeing a lot that's actually driving success. I had a little bit of an existential crisis at this point,” said Dr. Rinne.
But then with additional research, published in the July 1, 2017, American Journal of Respiratory and Critical Care Medicine, they found that readmissions of COPD patients were associated with readmissions for other medical conditions and inversely with patient experience measures. “To me that means that there may be some underlying organizational behaviors,” he said.
Additional hospital comparisons, presented at the 2018 ATS meeting, identified some of those behaviors. “Practices were not different, but what we did find was that there were differences in relational coordination. We found more shared goals between providers, we found more shared knowledge, we found more mutual respect, we found more high-quality communication,” said Dr. Rinne.
These differences were uncovered through interviews with hospital staff, and the variations in hospital culture were tangible, according to Dr. Rinne. “At sites that had low rates of readmission, you heard people talk about each other like friends, like colleagues. At sites with higher readmissions, you heard tension, even animosity,” he said.
That might sound hard to change, but another study, published in the Feb. 19, 2018, BMJ Quality & Safety, showed that hospitals can change their culture and see associated improvements in quality outcomes, including patient mortality.
“We think that we are going to be these lone rangers who implement care of high quality, but the reality is we are members of a health care team that operates within an organization and the quality that we deliver is a factor of the organization's behaviors and practices,” said Dr. Rinne.
The challenge is then how to encourage those optimal behaviors, practices, and culture. Dr. Feemster described how the HRRP, while developed with the laudable goal of reducing avoidable readmissions, has had some negative effects.
“There are definitely some reasons to look at this as a glass half-empty,” she said. “There are some unintended effects that have been realized. Our group and others have criticized the HRRP from the outset for the fact that it does fail some basic criteria for an accountability measure.”
Problems include the particular penalization of safety-net hospitals and recent research indications that the push to avoid readmissions might lead to increases in ED visits and observation stays, Dr. Feemster said.
In addition, studies have demonstrated increases in postdischarge mortality among patients with heart failure, but not acute myocardial infarction, under the HRRP raising the possibility that the policy is better suited to acute, rather than chronic, conditions, she noted.
“Whether we like it or not, the HRRP measure is here and it's not going to go away, so we can definitely improve it,” she added. Potential fixes include comparing hospitals to similar peers (which has already begun with the passing of the 21st Century Cures Act) and changing outcome measures to include returns to observation or the ED and mortality after discharge.
“Measurement alone does not improve the value of care,” she noted.
It does draw attention to it, though, Dr. Press pointed out. “Regardless of your feelings about the 30-day readmission metric and the poor level of evidence that we have to actually do anything about it, COPD is actually getting some attention and I'm happy about that. That's my silver lining,” she said.