Did performance-based nursing home payments meet their match with COVID?

Whether we’re buying groceries, cars, or colonoscopies, quality and cost matter. 

Since 2008, government and private health insurers have moved toward payment systems that purport to reward healthcare providers for the value they deliver — that is, patient health outcomes relative to costs incurred. The idea is that by prioritizing value — with its references to both quality and cost — patients, payers and providers will be well-served. 

But a recent attempt to link emergency funding for nursing homes to their COVID-19 outcomes failed spectacularly, and in ways that surely left public health experts scratching their heads.

Sixty percent of all insurer payments in 2018 were tied to quality or value, and value-based models now account for nearly half of all Medicare payments. Almost all providers currently derive at least some revenue based on their proficiency in hitting quality targets set by insurers or regulators.

When does tying healthcare payments to quality metrics increase value by improving quality, lowering costs, or both? Our recent study examined how the U.S. government provided $9.4 billion of COVID-19 relief funding to nursing homes during 2020.

The first two rounds of government funding distributed $7.4 billion to nursing homes in a non-performance-based manner — a fixed amount per nursing home, plus an additional fixed amount per bed. The Department of Health and Human Services then rolled out its Quality Incentive Payment Program (QIP) to distribute another $2 billion from September through December 2020. Under QIP, a nursing home received funding based primarily on its COVID-19 infection rate relative to its home county infection rate. 

The intent? Let the invisible hand reward high performers, punish low performers and optimally allocate funds to protect nursing home residents.

The results? Over four months, QIP distributed $2 billion to nursing homes that had virtually no COVID-19 infections. Across 15,331 nursing homes and four months, 94 percent of government checks went to nursing homes with no infections; 5.1 percent went to nursing homes with one infection, and 0.9 percent went to nursing homes with two to six infections. The kicker? In those months for which nursing homes did not receive QIP checks, over 300,000 residents contracted COVID-19.

But the dysfunction does not end there. For months when nursing homes did receive QIP payments, the correlation between their infection rates and QIP cash payments was approximately zero. That’s right. Among QIP recipients, a nursing home’s own infection rate had essentially no effect on its payments. In contrast, the correlation between nursing homes’ county infection rates and their QIP payments was 32 percent.

Bottom line: QIP was not “performance-based,” since county infection rates — not facility rates — were the primary driver of payments. Worse still, QIP distributed payments in inverse proportion to need, since nursing homes received nearly $2 billion when they had no infections, but zilch when they had 300,000 infections.

What are the broader lessons? First — in devising payment systems, deal with the hardest question head-on: Are payments warranted at all? MedPAC credits the spike in nursing homes’ profit margins (3.0 percent in 2020, compared with 0.6 percent in 2019) in large part to COVID-19 relief — surely an unsatisfying result if the goal was to combat COVID-19 and not fatten nursing homes’ coffers.

Second — details matter. Performance-based payment systems that link providers’ rewards to invalid or unreliable quality metrics can produce suboptimal (and even deadly) results. Nursing home administrators in particular are well-acquainted with a proliferation of metrics that cycle in and out of quality ratings, but often provide little operational guidance. Value-based payment systems are only as good as their design and implementation.

Third — let’s think twice before incorporating complex, market-based principles into short-term programs intended to combat unpredictable, once-in-a-hundred-years public health crises. HHS’ big whiff with QIP is not a sweeping indictment of all performance-based payment systems. But we should be mindful of the diversity in healthcare’s clinical settings and ecosystem when crafting payment systems. After all, to a bad carpenter with only a hammer in his toolbox, everything looks like a nail.

Elizabeth Plummer, Ph.D., CPA is a professor at Texas Christian University’s (TCU) Neeley School of Business and the TCU School of Medicine. Bill Wempe, Ph.D., is a professor at Texas Christian University’s (TCU) Neeley School of Business.