Under the current Medicare payment system, Medicare pays different rates to treat similar patients at post-acute care (PAC) settings such as nursing facilities, home health agencies, inpatient rehabilitation hospitals, or long-term care hospitals. The Senate Finance and House Ways & Means committees recently passed bipartisan legislation, the Improving Medicare Post-Acute Care Transformation Act (IMPACT), signed into law by President Obama on October 6, 2014, in an effort to standardize patient assessments and pave the way for effective payment reforms. National health policy expert Dr. Barbara Gage of the Brookings Institution spoke with the Georgetown Public Policy Review about the results of her congressionally mandated Medicare Post-Acute Care Payment Reform Demonstration (PAC-PRD) and the development of the standardized Continuity Assessment Record and Evaluation (CARE) Item Set, which is expected to be used for standardizing data collection following the passage of the IMPACT.
Georgetown Public Policy Review (GPPR): Based on your earlier research, what are some of the issues with the way PAC is currently delivered and paid for?
Dr. Barbara Gage (BG): My earlier research had shown that patients who were discharged from a hospital for say a stroke, might be discharged to a skilled nursing facility (SNF) in one part of the country or a rehabilitation hospital in another part of the country. With the data that were available, you couldn’t dive down deep enough to know whether they were truly the same type of patient in terms of health, functional and cognitive statuses, or whether those factors were associated with the decision to put them in a more intensive rehab hospital for the therapy, or a less intensive rehab setting such as the SNF. We knew we were paying different amounts per day, but not whether that was justified because of differences in complexity.
GPPR: How did the PAC-PRD address this problem?
BG: The PAC-PRD was very important because the goal was to take a currently disparate set of payment systems for post-acute care and collect information in a consistent way so that we could create standardized case mix measurement approaches, and understand whether the Medicare program was paying differently for similar patients who may be treated in alternative types of settings. The standardized measurement item sets were tested using the CARE tool, which took several years of development.
GPPR: What was involved in developing common measures?
BG: We worked with each of the different research, measurement, and clinical communities affiliated with inpatient rehabilitation hospitals, skilled nursing facilities, home health agencies, long term care hospitals, and acute hospitals to gain clinical input on the factors that distinguished cases needing one level of care from another.
Development of the standardized item set involved input from 25 different associations, with each of their clinical representatives identifying the key common concepts and gaining consensus on the best items for measuring those concepts. We needed the different types of disciplines—the nurses, the therapists (physical, occupational, and speech), as well as the different types of physicians, case managers, and social workers. The development of the standardized items was based on consensus from the different disciplines.
So by having standardized information we could identify how complex [the patients] were at the start of their care and at the end of their care, which would give us information on outcomes, which could then be examined in the context of the resources that were applied to that patient. The resource intensity data was collected through staff time studies in each of the settings—we had any staff member associated with that patient’s care tracking their time with the individual patients. These data were used to develop a relative value scale for post-acute care. By applying wage-adjusted time intensity metrics, we could assign a relative value of clinical intensity to the case mix that was identified through the CARE items.
GPPR: So what did the demonstration results show?
BG: The demo results were very interesting because they allowed us to look at the differences in outcomes. We looked at the differences in the probability of the readmission rate within 30 days following discharge. We looked at improvements in function and contributed to some of the current work that’s under way about how best to measure function. Historically, the rehabilitation community had measured changes in motor skills, which was a composite measure of the mobility, the activities of daily living (ADLs), and cognition. But those are three independent factors that, together, may have different impacts on outcomes. So we were able to disaggregate those factors and look at the relationship between the medical complexity, the functional complexity, and the cognitive complexity at admission, to measure the changes in mobility, or the changes in the ADLs, or the changes in the medical complexity, and that was kind of groundbreaking work.
We were also able to look at the relative value of a patient who was treated at different settings, all else equal. Once you control for that case mix complexity using standardized measurement, if you look at the outcomes associated with the patient who goes into a SNF versus the outcomes of a patient who looks exactly like that at admission but was admitted to a home health agency, you can say something about the relative costs of care associated with the outcomes.
GPPR: How can this research be used to help determine which types of facilities to send certain patients and maximize positive health outcomes while reducing costs?
BG: This is the same question the Hill asked a couple months ago, and as I told them when they asked for my input on the issue, the standardized information is necessary to understand the individual case. The determination of what setting somebody belongs in is really a clinical decision, but that decision gets made on the basis of how medically, functionally, and cognitively complex they are. So having these standardized items collected at the hospital will allow rational, logical decisions about what is the most cost-effective setting for that case.
It’s very consistent with the whole area of case mix measurement, which is what underlies the hospital Inpatient Prospective Payment System (IPPS). The IPPS is just the logic tree based on the medical condition that the patient at the hospital has, stratified by the complexity of that condition. This takes that same logic and takes it outside of the hospital. We have a logic tree for each of the PAC settings. However, each PAC PPS uses different items in describing patient complexity. If a patient has a pressure ulcer it doesn’t matter what setting they’re in; they should be measured using the expert’s language in pressure ulcer measurement, not the nursing facility’s language or the hospital’s language. So for the CARE items, we turn to groups like the national pressure ulcer advisory panel council for their input on defining the pressure ulcer items, just as we turn to the other communities for the best items in their clinical wheelhouse
GPPR: The Affordable Care Act deals with the inefficiencies of the fee-for-service system by creating bundled payments for hospitalizations and post-acute care. How can the new payment structures lead to improvements in care and lower health care costs?
BG: I think the bundled payments systemis one of the most important initiatives that the Center for Medicare & Medicaid Services (CMS) has established. While CMS has tried many efforts under the health reform umbrella, the bundled payments have really targeted the post-acute populations, which are the high-cost Medicare populations with the greatest variability in service use. By bundling the payments, CMS is giving the bundle-taker the incentive to manage the cost of those types of cases, but they are also holding them responsible for the outcomes. The bundler, or the awardee (such as a hospital), has a stake in outcomes as well as in cost-containment. By giving somebody responsibility for the entire episode of care, you’re really broadening the discussions from a silo-based, fee-for-service, volume-based payment system to an episode-based system with accountability across entities for the better outcomes for that patient.
GPPR: How would initial fees for the bundles be set, and in what way does existing research inform these decisions?
We just had a piece come out recently in the Medicare and Medicaid Research Review, the Medicare journal, that shows the distribution of costs associated with different episodes of care for different case mix groups. If you look at the MS-DRG (Medicare Severity Diagnosis Related Group) and the reason that a patient was discharged or treated at a hospital, you can see how the costs and the probability of use differ across the subsequent services, and it forms a baseline for understanding where to set the rates. The way the bundled program runs is—they’re sharing savings that are achieved relative to the average cost for that type of case. So what types of arrangements they are making to distribute risk among their partners is up to the awardee. It really opens the field for innovations for each of the provider communities and the insurer communities to try something different.
GPPR: In what way would you consider this to be a compromise between regulations and letting the organizations in the private sector do what they think is right in order to increase efficiency and improve care?
The whole goal of the innovation center is to let the private sector come up with ideas about better ways to manage the cost. Regulation and insurance rules are just two labels for the same thing, but because it’s public insurance, it’s considered regulation. All they are doing is setting insurance rules, and in order to insure somebody you have to have parameters around the expected costs; there have to be rules regarding who you cover, what services you pay for, and for what durations.
About Dr. Barbara Gage
Dr. Barbara Gage is a Fellow and Managing Director of the Engelberg Center for Health Care Reform at the Brookings Institution. Dr. Gage is a national expert in Medicare Post-Acute Care policy issues, including bundled payments, episodes of care, and case mix research. Dr. Gage has been monitoring and evaluating the impact that Medicare payment systems have on access and quality of post-acute care, services that patients receive at nursing facilities, home health agencies, inpatient rehabilitation hospitals, or long-term care hospitals.
Dan Kinber interviewed Dr. Barbara Gage in person on January 31, 2014.
Feature Photo: Flickr/(Myfuture.com)
Dan Kinber received a B.A. in Computer Science and French Studies at Wesleyan University (2001), an M.A. in Cinema Studies at New York University (2005), and a Master of Public Policy at Georgetown University (2015). Dan has been conducting interviews with health care providers, analysts, and beneficiaries of the Affordable Care Act, researching its impact on health care quality and costs. He founded and developed mustbeheard.org, a website for communicating the findings. Dan interned at the U.S. Department of Health and Human Services, and recently evaluated the Hospital Readmissions Reduction Program. Currently, he works as a Health Research Analyst at Mathematica Policy Research. Contact: dk764@georgetown.edu.