As a Congressional advisory board grapples with how to bring all post-acute providers under a single payment system, the use of self-reported data continues to raise doubts about the ability to accurately measure patient outcomes.
Members of the Medicare Payment Advisory Commission on Thursday directed staff to continue evaluating a proposed Prospective Payment System with an eye toward improving how they will measure patient function.
It was one of several post-acute topics in a two-day MedPAC meeting; members also discussed skilled nursing staffing minimums and the reliability of that data in an afternoon session (see below). But much of the morning was dedicated to the ongoing development of the PPS model.
“Beyond just harmonizing payments, I think there’s real value in this exercise in harmonizing cost sharing, harmonizing quality measures, harmonizing the regulations across these different PAC settings. I think that’s going to have value wherever this kind of model ends up,” said MedPAC member and Harvard healthcare policy expert David Graboswki, PhD.
But differences in how functional status and outcomes are assessed continue to make it hard to compare settings and the care delivered there “apples-to-apples,” Grabowski added.
He and several other commissioners expressed specific concerns with rating patient function. Early research into the advent of the Patient Driven Payment Model showed that nursing homes significantly decreased therapy minutes, while their assessment of individual patient needs, or characteristics, “went way up,” Grabowski said.
“How much of that is real? We think very little based on the hospital claims. We think a lot of that is upcoding,” he said. “So as we rely on coding from the different post-acute providers, let’s make certain that it’s accurate. That’s really the tension: How much we can get from the hospital claims, and how much we have to rely on them telling us what the characteristics are. I’m very suspicious based on what’s happened in PDPM that we’re going to get back accurate information.”
MedPAC is following through on a 2014 requirement to design a unified payment system that would pay for and assess care provided by skilled nursing, inpatient rehabilitation facilities, long-term care hospitals, and home health agencies.
Staff is currently working toward its next report, due in June 2023, in which it is evaluating a system proposed by the Secretary of Health and Human Services earlier this year.
That work looks to update analysis and design features, especially in light of changes in provider and beneficiary behaviors during the pandemic; compare the prototype to features MedPAC previously said it would prefer; assess if additional diagnostics are needed; and analyze how the system could be implemented, and whether levels of payment of a transition period would be needed.
MedPAC will review the entire report for the first time in March and vote on the final draft in April.
Carol Carter, a principal analyst for the commission, said researchers were looking to better understand how and when function is recorded in various PAC settings, particularly when it is tied to payment.
“There are incentives to assess patients as lower than they actually are to set the payment at a higher rate, so we’re worried about the quality of the information as much as we are missing data,” she said.
Member Dana Gelg Safran of the National Quality Forum pointed out that nursing homes weren’t the only setting that might produce biased information, adding that home health patients could give inaccurate information about themselves because of cognitive issues.
Member Robert Cherry, MD, chief medical and quality officer at UCLA Health, echoed Grabowski’s concerns about measuring functional status across settings. He wants regulators to eventually make functional outcomes a “mandatory report out” so that nursing homes and other settings have to provide information using a common methodology that is less prone to upcoding.
“Those functional outcomes are really critical in determining whether or not the payment is appropriate,” he said. “At the end of the day, it’s really important to understand what we’re paying for and what we’re getting.”
Incentivize higher staffing?
Also on Thursday, MedPAC members received a cursory analysis of Payroll Based Journal staffing data. In light of the Centers for Medicare & Medicaid Services’ study of a potential staffing minimum, commission staff wanted to know how members might want to use that data to inform future policy efforts.
Among key findings was the fact that contract labor nearly tripled during the pandemic, from 3% in the first quarter of 2019 to 8.4% in the fourth quarter of 2021 (the most recent data included).
Members noted that the overall number of staff hours fell in that same time period, creating a narrower gap between available workers and resident needs. While that might not have looked like a crisis to his fellow MedPAC members, Grabowski underscored that the fall of staff hours occurred even as the residents themselves presented as more medically complex, demanding more care.
Noting how much better the objective payroll-based data was than previous, self-reported metrics, he encouraged MedPAC to pursue a position that would better incentivize higher staffing levels.
“They have this great measure now, and they should weight that more heavily (on Care Compare),” he said. “I would be very much in favor of recommending that to CMS.”
Grabowski also said MedPAC should evaluate the hours physical therapists spend on resident care, while another member encouraged staff to determine how frequently physicians are in nursing homes and how that might be connected to quality.
