In April, the Centers for Medicare and Medicaid Services (CMS) proposed a rule that included in it an important change in how hospitals would be scored within the Hospital Readmissions Reduction Program (HRRP).
The rule, which would take effect in FY 2019, would require Medicare to consider social risk factors when calculating hospital penalties under the HRRP, as mandated by the 21st Century Cures Act. This modification was something that healthcare stakeholders have long been calling for, as they attest that the HRRP disproportionately penalizes institutions that serve low-income and clinically complex patient populations.
Historically, HRRP, since it began in 2013, has assessed readmissions rates for certain conditions and has penalized those hospitals whose patients had higher readmission rates than would be expected. But, the readmission rates used in the program have not accounted for the demographics of a hospital's patient population. Then came the Cures Act, which mandated that HRRP begin to account for patients' socioeconomic status in its assessment of readmissions performance.
According to an Advisory Board analysis of the April CMS rule, the federal agency “intends to sort hospitals into five peer groups according to their dual-eligible inpatient stay ratio. Within each peer group, hospitals' excess readmission ratio for each of the program's six conditions will be compared to the group's median excess readmission ratio for that condition. The proposed change, which is designed to be budget-neutral for Medicare, would shake up the magnitude of HRRP penalties, especially for some safety-net hospitals.”
Indeed, as further noted in a rule analysis from The BDO Center for Healthcare Excellence & Innovation, a consulting firm dedicated to helping healthcare organizations achieve optimal clinical and financial performance, “Dual-eligible patients are historically more expensive for hospitals, skewing the readmissions figures for safety-net hospitals, leading to unfair penalties against them and inhibiting their ability to provide the best quality of care to populations that need it the most.”
But, according to the center’s officials, the suggestions put forth in the CMS rule would allow safety-net hospitals to control for variables often out of their control as they relate to dual-eligibility. However, they stated in their analysis, when it comes to high readmission rates and the associated penalties, other population factors are often also at play regardless of a hospital’s quality scores including: the proportion of patients who are linguistic minorities; the proportion of patients who have behavioral health diagnoses; and the proportion of patients with minimal social support.
Recently, two senior leaders at the center—Bill Bithoney, M.D., senior fellow at The BDO Center for Healthcare Excellence & Innovation, and Patrick Pilch, national co-leader at the firm—spoke with Healthcare Informatics about their high-level takeaways from the rule, the impact that it might have on the HRRP, and what they would like to see further changed in the final rule, expected to drop this fall. Below are excerpts of that interview.
What were your main takeaways from the rule, as it relates to the Hospital Readmissions Reduction Program?
Bithoney: We saw it as a leap forward and that CMS is taking into account dual-eligible patients, as they typically cost three to four times as much [as non-dual-eligibles], largely because they have great needs but also because of their social, behavioral and demographic issues. So we felt that it was a great leap forward to take into account these sociodemographic factors. However, one could also take into account other things—some people have recommended strongly that we look at census tracts and the poverty within the census tract that the hospital operates. So let’s look at the population in general, and those social demographic issues that are associated with that poverty, such as the presence of isolated linguistic minorities and the increased prevalence of behavioral issues and opioid issues.
Bill Bithoney, M.D. & Patrick Pilch
I would also like to see them take into account the proportion of patients included in the hospital sample that have three or four serious medical diagnoses. It’s well-known that you are much more likely to be readmitted if you have three or four of these [conditions compared to one]. Medicare can take this into account; it’s not that hard. The data capacity of Medicare is incredible and has been burgeoning; they can calculate a risk score for [all of the lives that they cover]. So you can weigh the hospital’s readmission rate by any of these factors. These risk scores are all weighted based on how many diagnoses a patient has, and you get a medical risk score. The only place CMS does social weighting of risk in PACE [Programs of All-Inclusive Care for the Elderly], where they have a RAPS [Risk Adjustment Processing System] score that begins by taking into account the Medicaid population of all patients, not just dual-eligibles. So there’s a mechanism that exists in CMS to do just that—weighing hospitals by if their patients have three or four diagnoses and if they are poor.
Past CMS administrators have said they don’t want to take into account poverty because it’s prejudicial and they want to treat everyone the same. But we think that’s a cop out—sure that sounds good, but we know these risk factors impact readmission rates. There are 5,000 hospitals in the U.S., so these mathematics and data rates are available and routinely calculated.
Pilch: if you look at the continuum of care, especially for people who have comorbidities and chronic illnesses—and we do a lot with various Medicaid reform plans around the country—the hospitals are typically the most expensive providers of care for chronic illnesses. So in respect to the PACE program, for instance, we believe that this is something that CMS wants to continue to foster, and to the extent that this program is run effectively, a lot of the care can be provided outside the hospital. So it does beg the opportunity to look at what programs are in the market that hospitals can partner with so that when someone goes emergent, the hospital is there, but for other [problems], there are ways in which you can be more coordinated. Overall, the programs that are risk scored appropriately have been more successful.
Bithoney: Having risk scores for everyone will certainly be a major benefit. If those scores include social issues, which is easy to do, that will make life so much better. When you have risk scores, people who run ACOs [accountable care organizations] and Medicare Advantage [MA] programs, for instance, have the benefit of knowing those scores. When I was a hospital CEO, in our MA program, 3 percent of our patients accounted for 49 percent of the cost of care in that program. [Ratios like these] are found continuously. So if CMS would provide hospitals with risk scores for all their patients, they can target those 5 percent and do outreach into the community, which would make big inroads. We were able to intervene with those 3 percent prospectively, and cut 8 percent of the cost of care. These programs run billions of dollars so an 8 percent cut can be impressive. And the risk scores can be a guideline for prevention of illness in a value-based world.
How will these proposed changes impact future HRRP penalties?
Bithoney: What people believe will happen is that the penalties for hospitals that reside in poorer areas will not be so severe. It also may well be that some of the hospitals that are not being penalized now and are being rewarded find out that they aren’t doing as well as they thought.
How might hospitals which serve a higher proportion of dual-eligibles need to rethink how they are managing the patients who are coming in?
Pilch: I believe that for the safety-net hospitals, this rule doesn’t give them an out, but more of an opportunity. We talk to a number of safety-net hospitals, and we recently had a conversations with a CFO of a large safety-net hospital in the Midwest in which we asked how many dual-eligibles come into your hospital every day. The CFO responded that he wasn’t sure. But now they will really have to know. Then you can make sure what you are representing in terms of the volume coming in is actually the true volume coming in. Because you do not want to inadvertently under report or not report. If you do, it will come back at you in the end.
Do you think using readmission rates as a measure of hospital quality is fair?
Bithoney: I think the way it’s been taken into account has not been fair, but everyone does have a different definition of “fair.” CMS is acknowledging there are risk factors beyond the control of the hospital system such as taking care of linguistic minorities. Some cultures have very different approaches. So this is an admission from CMS that they believe it’s not fair at this point in time and they want to make it fairer.
What would you like to see changed for the final version of this rule as it relates to the HRRP?
Bithoney: In the three-hospital health system that I managed in Massachusetts, we gave those patients extra services and intensive services, and we cut the Medicare readmission rate for the entire population in the MA program from 20 percent to 9.8 percent. What it showed me was that such a large proportion of the hospital readmissions were sick and poor people. When you combine the people who have three or more diagnoses that resulted in having a high risk score with the people who are in the poor census tract, you have great opportunities to use that risk scoring to target the sickest patients prospectively.
