WASHINGTON -- Patients being discharged from the hospital to post-acute care facilities need more information about the facilities' quality, but how that should be provided is hard to say, according to members of the Medicare Payment Advisory Commission (MedPAC).
"In the hospital world, there are thousands of quality measures," said David Nerenz, PhD, of Henry Ford Health System in Detroit. "Maybe [post-acute care facilities like] skilled nursing facilities are simpler, but you still would have hundreds of measures, and every star rating system [for quality] takes a subset of those. The second key thing is that the measures are not correlated with each other, so any subset you take will have no predictive power" for how the other measures will turn out.
"So if you're a beneficiary and you care about things outside of the star rating, that rating is not useful to you ... it may even be misleading."
Commission members were discussing a proposal developed by commission staff to provide patients going through the discharge planning process with more quality information about home health agencies, skilled nursing facilities, and other post-discharge institutions, as a way to encourage them to select high-quality providers. The issue has arisen because although Medicare spends billions of dollars on post-acute care, beneficiaries don't always choose the highest-quality care provider.
For example, the MedPAC staff found that 84.3% of beneficiaries who were discharged to a particular skilled nursing facility had at least one other higher-quality facility nearby, and 46.8% had five or more. Those higher-quality providers were better in significant ways, such as having a lower rehospitalization rate. However, hospital discharge planners currently are not permitted to recommend a specific post-acute care provider.
MedPAC staff member Evan Christman outlined three possible approaches:
- Flexible: Hospitals define their own quality measures and levels of performance for the facilities, and generate a list of high-quality providers to be shared with patients; hospitals would be required to collect and review performance data on the post-acute care providers, and maintain a formal record of the process
- Prescriptive: Hospitals must use Medicare-defined quality measures and performance levels; the Centers for Medicare & Medicaid Services would notify hospitals and beneficiaries of qualifying post-acute care providers
- Revised prescriptive: Medicare would account for variations in post-acute provider quality across markets, and could include specific data on how a provider stands up against competitors in a given geographic area
Commissioners were divided about which approach to use. "I'm not a big plan of a flexible approach," said David Grabowski, PhD, of Harvard Medical School, in Boston. "I think we'll end up with business as usual. I would much prefer a 'revised prescriptive approach,' where we're trying to tailor this to particular markets."
"This information could be a floor not a ceiling," he added. "If hospitals want to provide additional information, I would be fine with that ... I don't think we should limit the information set, but we should have a core set of measures and tailor it by market."
Paul Ginsburg, PhD, of the Brookings Institution, a left-leaning think tank here, said he wasn't comfortable with a prescriptive approach. "In many other areas of healthcare, our quality measurement is very primitive ... but the way this [proposal] was written, it's as if it's perfect information," he said. Using a prescriptive approach is "almost making the hospitals into Medicare's agents ... I'd like to hear more about a flexible approach."
Brian DeBusk, PhD, of DeRoyal Industries in Powell, TN, disagreed. "I'm really not comfortable with the flexible approach; the idea that you're going to choose your own quality measures feels like PQRS all over again," he said, referring to the Physician Quality Reporting System that Medicare formerly required doctors to use to report quality data. Under PQRS, doctors were able to choose which quality measures they would report.
Instead, he said, "I would recommend we take their standardized measures and run them through standardized peer grouping methodology ... What you could end up with is a prescriptive approach where results are stratified based on sociodemographic status, and results could be used for caregiving [and reporting]."
Practical issues need to be considered too, said Jack Hoadley, PhD, of Georgetown University's Health Policy Institute here. "A list of [facilities] could be created, but the reality is, when you get down to the actual picking, who's got a bed today? There are very practical things about convenience and location, and [maybe] this patient's got a ventilator ... We ought to be thinking about whether there are ways to study the process of discharge planning."
"I'm surprised and confused by this discussion," said Craig Samitt, MD, MBA, of health insurer Anthem, in Indianapolis. "I'm very much in favor of the prescriptive approach. Beneficiaries are hungry for this information, and we've provided no information.
"Is the information perfect? Likely not, but we shouldn't let the perfect be the enemy of the good."
Based on the discussion, "the solution is one on which we don't have a consensus," said commission chair Francis Crosson, MD, of Palo Alto, CA. He suggested including information on the issue in the commission's annual report to Congress in June, while continuing to work to reach agreement.