A federal website intended to help people choose a nursing home not only contains inaccurate information, but those inaccuracies appear to be at least partially driven by race, a new study reports.
The U.S. Centers for Medicare and Medicaid Services (CMS) established the Nursing Home Care Compare website in the 1990s to publicly report patient safety indicators for every nursing facility in the nation.
But the site appears to drastically underreport the number and severity of major injury falls and bedsores suffered by Medicare residents in specific nursing homes.
The site relies on self-reported data from nursing homes to track falls and bedsores, but Medicare claims data show that nursing homes are not reporting all of these incidents, said lead researcher Prachi Sanghavi, an assistant professor of public health sciences at the University of Chicago.
About 40% of major injury fall hospitalizations and 32% of severe bedsore hospitalizations found in Medicare claims data did not show up on the CMS website, the researchers found.
"It's not just slightly inaccurate. It's actually very inaccurate," Sanghavi said.
What's more, Sanghavi and her colleagues found that the underreporting varied based on the racial composition of the nursing homes.
Nursing homes with more white residents had higher reporting rates for major injury falls and lower reporting rates for pressure ulcers, and vice versa was true for nursing homes with more Black residents.
"Nursing homes in the U.S., as you probably know, are fairly segregated," Sanghavi said. "Nursing homes that have more white residents have more falls and report falls more accurately, and don't report pressure ulcers as accurately. And the opposite happens with nursing homes that have more Black residents."
The American Health Care Association/National Center for Assisted Living (AHCA/NCAL) — the nation's largest association representing long-term care providers — took exception to the new report.
"Nursing homes are committed to accurate reporting of quality measures, and this flawed study does not represent what's happening today. Additionally, nursing homes have made dramatic improvements on all quality measures over the past decade and continue to focus on improving the quality of life for our residents," the AHCA/NCAL said in a statement.
Race plays a part
"Throughout health care, including long-term care, we all need to address racial and ethnic disparities. A key step policymakers can take is addressing chronic, inadequate funding on Medicaid — a program a disproportionate portion of minority nursing home residents rely on," the statement continued.
For this study, the researchers compared data from the CMS site directly against claims that hospitals filed with Medicare to reimburse treatment for either falls or bedsores, Sanghavi said.
"We basically have a case where we know an individual was discharged from a nursing home and went to a hospital. The hospitalization was for the conditions we're looking at, so falls or pressure ulcers. And then we see that the person is discharged from the hospital and returns to the same nursing home that they first came from," Sanghavi said.
"And so we have this chain put together where we can say, OK, did the nursing homes report that or not?" she continued.
The researchers not only found underreporting, but they found an unusual pattern regarding how underreporting occurred.
"On average, nursing homes that have higher fall reporting rates have lower pressure ulcer reporting rates," Sanghavi said. "The better you report on falls, the worse you report on pressure ulcers."
That confused the researchers, because they expected that nursing homes that were better at reporting falls would probably be better at reporting bedsores.
"We found that in nursing homes, race mix really matters for understanding what's happening," Sanghavi said.
Looking at hospitalization numbers, they found very clear differences in actual events happening to white and Black nursing home residents.
"White residents have basically twice the fall rate of Black residents. And for pressure ulcers, the Black residents have almost twice the rate as white residents," Sanghavi said. "So, the actual hospitalizations for the two events are defined by race."
The researchers speculated that institutional racism might lead nursing homes to treat residents differently if they are Black or white.
Underreporting paints an incomplete picture
"For example, facilities with more white residents may minimize use of restraints, facilitate mobility and provide regular repositioning, all of which could increase the risk of major injury falls while reducing the rate of pressure ulcers," the researchers reported.
It's also possible that health disparities between the races could cause these differences. "For example, Black residents are more frequently diagnosed with obesity, a risk factor for pressure ulcers," Sanghavi said.
The finding were reported online May 23 in JAMA Network Open.
The inaccurate reporting of falls and bedsores "obviously hinders people's ability to assess whether or not a nursing home is a good fit for them or not," said Gretchen Jacobson, vice president of the Medicare program at The Commonwealth Fund, a health care think-tank. "If you don't have that data, and especially if you're missing pieces of information, how can you be expected to make an informed decision?" she said.
"It's also worth noting that once people enter a nursing home, it's not a simple task to change nursing homes," Jacobson added. "So, really having all the information at that time is just critical."
Sanghavi noted that CMS itself appears to have quietly acknowledged the problem, because these self-reported quality measures are weighted less heavily than other data used to craft the agency's overall rating from one to five stars that appears for each nursing home on the Compare site.
"The problem is that they're still on the website," Sanghavi said of the self-reported measures. "The quality measures, because they're posted there, have this sort of legitimacy to them, because they're backed by the federal government."
She said the CMS website could change its methodology to something similar to her team's approach, in which hospitalization claims data would be factored into a nursing home's rating.
"CMS should lean heavily on the claims data that it already owns," Sanghavi said. "They should replace or at the very least supplement the quality measures based on nursing home-reported data with measures from more objective data sources, like Medicare and Medicaid claims from other health care providers who don't really have an interest in the nursing home rating system."