Medicare readmission penalties examined

Advice is given on how hospitals and clinicians can avoid the pitfalls of the penalty program, and how policymakers could change it to operate more effectively.

On Dec. 6, 2013, the Centers for Medicare and Medicaid Services (CMS) announced successful results from the hospital readmission penalties program begun in 2012: 30-day readmissions had dropped from an average of about 19% in 2007-2011 to 18.5% in 2012 to under 18% in the first half of 2013.

But on the very same day, thought leaders gathered at the National Readmission Summit in Washington, D.C. were critiquing the penalty program. They praised the penalties for effectively focusing attention on the problem of readmissions but expressed concerns about the program's potential effects, from increasing disparities in care to distracting hospitals from needed improvements.

“A 0.6% decline in the national readmission rate meant that 87,000 Medicare patients did not come back to the hospital,” said Nancy Foster, vice president of quality and patient safety policy for the American Hospital Association. “There's still room for improvement, however.”

“CMS, you've done a good thing,” agreed Amy Boutwell, MD, MPP, a hospitalist at Newton-Wellesley Hospital in Newton, Mass., and founder of Collaborative Healthcare Strategies, a company working on improving health care delivery. “However, their good work has created some blinders for us in the field.”

The experts offered their thoughts on how hospitals and clinicians can avoid the pitfalls of the penalty program and how policymakers could change it to operate more effectively.

Expand your focus

One problem with the penalties is that they have focused hospitals on the particular diagnoses chosen by CMS.

“Heart failure, acute myocardial infarction and pneumonia are not the top 3 reasons for readmissions,” said Dr. Boutwell. “I've done hundreds of data analyses for hospitals across the country—large, small, community, rural—and these are on the top 10 list, but they are not the top 3. It's really important to recognize that there are many other important and serious conditions, including issues around social complexity, that need to be attended to to get hospital-wide readmissions down.”

Using data from the Healthcare Cost and Utilization Project, she provided a top 10 list of diagnoses responsible for the most 30-day readmissions to U.S. hospitals (see Table 1) and pointed out some of the issues overlooked by current prevention efforts. “Sepsis is a big deal and we haven't talked about it yet as a care transitions community,” Dr. Boutwell said. The presence of mood disorders and schizophrenia high on the list highlights the importance of behavioral health care, too, she said.

The top 10 is rounded out by some less severe problems. “Urinary tract infections, fluid and electrolytes, renal failure, cellulitis—these are way lower acuity. If you're a readmission discharge advocate running around the hospital looking for red flags and high risk scores, you're not picking up these little old ladies or diabetics with cellulitis,” said Dr. Boutwell.

Analyzing diagnoses by the rate of readmissions also reveals some areas of needed improvement (see Table 2). The top 10 are less common reasons for admissions, but with readmission rates above 25%, they are worth thinking about, Dr. Boutwell noted. “HIV, sickle cell, hepatitis, leukopenia, lupus—these are not on your BOOST tool. They are not on a screener's list,” said Dr. Boutwell.

The lists show why hospitals and hospitalists need to focus on all-cause readmissions, not just those spotlighted by CMS. Cancer didn't make either top 10 list, but it should be on hospitalists' radar, too. “We often categorically exclude cancer,” said Dr. Boutwell. “I can't tell you how many cancer readmissions I have that are not for chemotherapy. They're for pain and constipation and dehydration and things that should be avoided.”

The Medicare penalties have also drawn attention away from the payer that actually has the highest readmission rates—Medicaid. Heart failure readmissions are as much of a problem among Medicaid beneficiaries as Medicare, Dr. Boutwell reported.

The issue of Medicaid readmissions hasn't been publicized or targeted as widely, but many hospitals could still save money by reducing them, she said. “Some states just stop paying at some point and many Medicaid agencies have no-payment policies for the readmission itself. There are a lot of hidden payment penalties already in the Medicaid system,” said Dr. Boutwell. “Because it's such a lower-paying system in general…[hospitals experience] a de facto penalty by not optimizing their care transitions for that population.”

Policy problems

Penalizing hospitals for treating the poor was a major concern of other speakers at the summit. “Hospitals treating disadvantaged communities are more likely to incur [Medicare readmission] penalties,” reported Ms. Foster.

It's well known that socioeconomic factors contribute to readmissions, agreed Mark Miller, PhD, executive director of the Medicare Payment Advisory Commission (MedPAC). “Hospitals and health systems that have to deal with lower-income populations have a tougher road to get to a lower readmission rate,” he said.

The solution proposed by MedPAC is to have these hospitals continue to report their readmission rates like everyone else, so they have an incentive to improve, but only penalize them if they perform worse than their safety-net peers.

The specifics of the readmission penalties will become increasingly important as the CMS program continues, the speakers noted. In 2013, hospitals were subject to at most a 1% penalty on their payments. “The average penalty for a hospital is about $125,000,” said Dr. Miller.

In 2014, the penalty goes up to 2% of payments and then 3% in 2015. If the current trend continues, readmissions will go down over those years. But that decline will actually make it harder for hospitals to avoid penalties, because they'll have to get readmissions below the new average rate.

“The current penalty doesn't decline if readmission rates go down. If the industry begins to reduce their readmissions, there's a multiplier effect that basically makes the penalty stay the same and we think that's the wrong incentive,” said Dr. Miller.

MedPAC has proposed a solution to this issue as well. “We think a clearer, simpler, better incentive would be to have a fixed readmission rate target. Pick a rate that's less than the average and say, if the hospital clears this hurdle, there's no penalty,” he said.

The commission also favors measuring hospitals based on their preventable readmissions for any cause, rather than the specific diagnoses that have been targeted by Medicare thus far. It would make data from small hospitals less subject to random variation and lower the risk that clinicians might try to game the system in their selection of diagnostic codes, Dr. Miller said.

Doc fixes

A broader, all-cause approach to readmission prevention is exactly what Dr. Boutwell called for, too. She noted that this would entail returning to some ideas of the pre-penalty era. “When I started this work at the [Institute for Healthcare Improvement], we were really talking about…what are our processes and how do we improve?” said Dr. Boutwell.

Since the penalties were imposed, many hospitals have directed their efforts and funds toward hiring transition coordinators to target high-risk patients. “If we just start focusing in on heart failure, diabetes, COPD and pneumonia, we are missing all of these extremely needy patients,” she said. “I think we have lost sight of the internal quality improvement that needs to happen in addition to the intensified transitional care that those [new employees] represent.”

If an improvement project focuses just on a small group of patients (such as those with Medicare's 3 diagnoses), models by Dr. Boutwell show it's nearly impossible to significantly reduce overall readmission rates. “We need to have enhanced services for high-risk [patients], but…we need to improve standard hospital-based care for all,” she said.

Her broad-based solutions—for example, better communication with primary care physicians and patients' families—won't come as a surprise to hospitalists, but they are clearly within their purview. So were almost all the diagnoses on those top 10 lists.

Dr. Boutwell urged her colleagues to get to work on these issues. “Please go crunch your numbers for all your patients,” she said. “Do this exercise of saying... ‘What does each component of our strategy reasonably expect to achieve? Are we on the path to reducing our hospital's or our community's readmissions?’”