CryptocurrencyLeadershipInvestingClimate ChangeMost Powerful Women

Hospitals have become less safe during the pandemic. So why does the government want to suppress hospital safety data?

June 14, 2022, 4:28 PM UTC
Busy Hospital Emergency Department
A hallway in the emergency department at Mount Sinai South Nassau hospital in Oceanside, N.Y., on April 13, 2020, during an early peak in the COVID crisis. Federal regulators have proposed not publishing some hospital safety data from the pandemic period, arguing that the crisis has skewed the data and made it less useful.
Jeffrey Basinger—Newsday via Getty Images

There’s little question that U.S. hospitals—up against COVID, patient surges, and labor and supply shortages—have become less safe for patients during the pandemic, as preventable events and complications have become more common.

Leaders with the Centers for Disease Control and Prevention (CDC) and the Centers for Medicare and Medicaid Services (CMS) said as much, earlier this year, in an article for the New England Journal of Medicine: “Many indicators make it clear that health care safety has declined,” they wrote, noting, “the fact that the pandemic degraded patient safety so quickly and severely suggests that our health care system lacks a sufficiently resilient safety culture and infrastructure.”

Despite such frank assessments, CMS is now at odds with public safety advocates about whether to make some of the hospital-specific data behind those trends publicly available. The data would capture the rate of potentially preventable adverse events at hospitals during the tumultuous period. Last month, in a shock to many of those advocates, the agency proposed, in the language of federal rules, to “suppress” data from 10 such measures (collectively known as the Patient Safety and Adverse Events Composite, or PSI 90) for the pandemic period, including those which track postoperative complications such as respiratory failure, blood leakage, and sepsis, and events like pressure ulcers (i.e., severe and especially harmful bedsores) and falls resulting in hip fractures at hospitals.  

Those 10 events account for 25,000 preventable deaths and 94,000 incidents of patient harm in the U.S. annually, according to recent analyses. As with other aspects of the health system, the data reflects existing disparities: Studies have found adverse events and surgical complications like those captured by the PSI 90 data disproportionately affect Black patients and other minorities.

For years, the government has collected such events to help determine hospital payments—those with the highest rates of avoidable complications are penalized by CMS, the nation’s largest health care payer. The government also has published the data, to better inform the public about the performance of health providers through its Care Compare tool. The data is used by hospital-rating organizations as well, including the Leapfrog Group, an employer-backed nonprofit which twice annually publishes safety grades for U.S. hospitals. 

“This is really critical data. We can’t get it anywhere else” says Leapfrog Group president and CEO Leah Binder, who notes that while CMS will continue to publish some other patient-safety–related marks for hospitals, these metrics represent a significant part of the picture. Health care consumers “can’t do anything wise about our health if we don’t have the information,” Binder says.

CMS’s proposal to suppress the 10 measures stems from concerns that COVID has distorted the data in a way that could be harmful to hospitals and patients and misinform the public. Dr. Lee Fleisher, chief medical officer and director of CMS’ Center for Clinical Standards and Quality, commented in an email statement provided to Fortune: “CMS’ top priority is to ensuring access to safe, comprehensive health care, and patient safety will always be our primary concern. An important part of CMS’ commitment to patient safety is ensuring public access to the highest quality data.” The agency notes that its measures were never designed to address the circumstances of “a once-in-a-generation event” nor to account for external factors—like changes to clinical practice and unpredictable patient volumes—that came with the public health emergency.

Did COVID skew the data?

The American Hospital Association endorses CMS’s proposal. Akin Demehin, AHA’s senior director of policy, says his organization’s concerns over the measures and how they’re used to characterize hospital performance predates the pandemic. But he argues that COVID, which hit hospitals at different times and with differing intensities, makes them even less reliable as a basis for comparison. The data will be biased against the most hard-hit institutions, he argues, since very sick COVID patients presented particular challenges; for instance, there were risks to repositioning them, a step that is important for preventing pressure ulcers, and they were more likely to require higher-risk interventions, like being intubated or placed on a ventilator for oxygen. Meanwhile, he notes, hospital workforces were under tremendous strain in responding to the public health emergency.

Binder concedes there’s a reasonable debate to be had over whether to adjust hospital payments based on the data, but she sees no logic in keeping such information from the public. “The fact that there was a public health crisis—not only is that not an excuse to suppress data, it’s a reason not to suppress data,” Binder says. 

She continues, “We’ve literally shut down the economy in places in order to protect hospitals from [COVID] surges. We recognize how important hospitals are to this country. As a result, we need absolute transparency about how they’re doing for all their patients.”

The Leapfrog Group is actively campaigning for the public and employers, who purchase the bulk of health care in America, to object to CMS’s proposal during the rule’s public comment period, which is open until this Friday, June 17. “Employers have a right to expect the federal government to give them the information that they have on performance of hospitals,” she says.

Rob Andrews, a former New Jersey congressman who currently heads the Health Transformation Alliance, an organization of 50 of the nation’s largest self-insured employers, says HTA’s member companies typically use such data to help educate their employees, design their health plans (with a goal of disincentivizing use of low-quality providers), and to address hospital performance and safety concerns with insurers and providers. 

“We very much want CMS to reverse this decision. Health care has many problems, but probably the most important one is a lack of transparency in the health marketplace. It’s very difficult to know the quality of what you’re buying and what you’re paying for if you’re a consumer or an employer.” 

Andrews commends the government for its increasing focus on price transparency in the industry in recent years, but he says it makes CMS’s proposal all the more alarming. “Cost data in the absence of quality data are at best, meaningless, and at worst, harmful. We see this limitation on collection and publication of data about these very serious safety issues as a step backward.” 

He acknowledges the importance of accurate, non-skewed reporting, but emphasizes what’s at stake without any transparency. “If your loved one was going into a hospital or a surgical center for procedure, you’d want to know if there was a long history of these kinds of problems,” he says.

Janet Beesting Nelson, a registered nurse and board-certified patient advocate in Florida, helps individuals and families navigate the health care experience, and to do so, relies on CMS’s data. Like Binder, she was astounded that the agency would choose to suppress the 10 adverse event measures now, even as pricing information is beginning to come into focus. 

“Just as we’re making some progress with the price transparency, now you’re going to take away the safety transparency,” she says.

Sign up for the Fortune Features email list so you don’t miss our biggest features, exclusive interviews, and investigations.