Anesthesiologist Claire Rezba started tracking lost health workers almost instinctively. Researchers and industry professionals say the lack of good official data on these deaths is "scandalous" and is putting lives in danger.

This article was published on Sunday, August 2, 2020 in ProPublica.

By Nina Martin

When police discovered the woman, she'd been dead at home for at least 12 hours, alone except for her 4-year-old daughter. The early reports said only that she was 42, a mammogram technician at a hospital southwest of Atlanta and almost certainly a victim of COVID-19. Had her identity been withheld to protect her family's privacy? Her employer's reputation? Anesthesiologist Claire Rezba, scrolling through the news on her phone, was dismayed. "I felt like her sacrifice was really great and her child's sacrifice was really great, and she was just this anonymous woman, you know? It seemed very trivializing." For days, Rezba would click through Google, searching for a name, until in late March, the news stories finally supplied one: Diedre Wilkes. And almost without realizing it, Rezba began to keep count.

The next name on her list was world-famous, at least in medical circles: James Goodrich, a pediatric neurosurgeon in New York City and a pioneer in the separation of twins conjoined at the head. One of his best-known successes happened in 2016, when he led a team of 40 people in a 27-hour procedure to divide the skulls and detach the brains of 13-month-old brothers. Rezba, who'd participated in two conjoined-twins cases during her residency, had been riveted by that saga. Goodrich's death on March 30 was a gut-punch; "it just felt personal." Clearly, the coronavirus was coming for healthcare professionals, from the legends like Goodrich to the ones like Wilkes who toiled out of the spotlight and, Rezba knew, would die there.

At first, seeking out their obituaries was a way to rein in her own fear. At Rezba's hospital in Richmond, Virginia, as at healthcare facilities around the U.S., elective surgeries had been canceled and schedules rearranged, which meant she had long stretches of time to fret. Her husband was also a physician, an orthopedic surgeon at a different hospital. Her sister was a nurse practitioner. Bearing witness to the lives and deaths of people she didn't know helped distract her from the dangers faced by those she loved. "It's a way of coping with my feelings," she acknowledged one recent afternoon. "It helps to put some of those anxieties in order."

On April 14, the Centers for Disease Control and Prevention published its first count of healthcare workers lost to COVID-19: 27 deaths. By then, Rezba's list included many times that number — nurses, drug treatment counselors, medical assistants, orderlies, ER staff, physical therapists, EMTs. "That was upsetting," Rezba said. "I mean, I'm, like, just one person using Google and I had already counted more than 200 people and they're saying 27? That's a big discrepancy."

Rezba's exercise in psychological self-protection evolved into a bona fide mission. Soon she was spending a couple of hours a day scouring the internet for the recently dead; it saddened, then enraged her to see how difficult they were to find, how quickly people who gave their lives in service to others seemed to be forgotten. The more she searched, the more convinced she became that this invisibility was not an accident: "I felt like a lot of these hospitals and nursing homes were trying to hide what was happening."

And instead of acting as watchdogs, public health and government officials were largely silent. As she looked for data and studies, any sign that lessons were being learned from these deaths, what Rezba found instead were men and women who worked two or three jobs but had no insurance; clusters of contagion in families; so many young parents, she wanted to scream. The majority were Black or brown. Many were immigrants. None of them had to die.

The least she could do was force the government, and the public, to see them. "I feel like if they had to look at the faces, and read the stories, if they realized how many there are; if they had to keep scrolling and reading, maybe they would understand."

It's been clear since the beginning of the pandemic that healthcare workers faced unique, sometimes extreme risks from COVID-19. Five months later, the reality is worse than most Americans know. Through the end of July, nearly 120,000 doctors, nurses and other medical personnel had contracted the virus in the U.S., the CDC reported; at least 587 had died.

Even those numbers are almost certainly "a gross underestimate," said Kent Sepkowitz, an infectious disease specialist at Memorial Sloan Kettering Cancer Center in New York City who has studied medical worker deaths from HIV, tuberculosis, hepatitis and flu. Based on state data and past epidemics, Sepkowitz said he'd expect healthcare workers to make up 5% to 15% of all coronavirus infections in the U.S. That would put the number of workers who've contracted the virus at over 200,000, and maybe much higher. "At the front end of any epidemic or pandemic, no one knows what it is," Sepkowitz said. "And so proper precautions aren't taken. That's what we've seen with COVID-19."

Meanwhile, the Centers for Medicare and Medicaid Services reports at least 767 deaths among nursing home staff, making the work "the most dangerous job in America," a Washington Post op-ed declared. National Nurses United, a union with more than 150,000 members nationwide, has counted at least 1,289 deaths among all categories of healthcare professionals, including 169 nurses.

The loss of so many dedicated, deeply experienced professionals in such an urgent crisis is "unfathomable," said Christopher Friese, a professor at the University of Michigan School of Nursing whose areas of study include healthcare worker injuries and illnesses. "Every worker we've lost this year is one less person we have to take care of our loved ones. In addition to the tragic loss of that individual, we've depleted our workforce unnecessarily when we had tools at our disposal" to prevent wide-scale sickness and death.

One of the most potentially powerful tools for battling COVID-19 in the medical workforce has been largely missing, he said: reliable data about infections and deaths. "We don't really have a good understanding of where healthcare workers are at greatest risk," Friese said. "We've had to piece it together. And the fact that we're piecing it together in 2020 is pretty disturbing."

The CDC and the Department of Health and Human Services did not respond to ProPublica's questions for this story.

Learning from the sick and dead ought to be a national priority, both to protect the workforce and to improve care in the pandemic and beyond, said Patricia Davidson, dean of the Johns Hopkins School of Nursing. "It's critically important," she said. "It should be done in real time."Bottom of Form

But data collection and transparency have been among the most glaring weaknesses of the U.S. pandemic response, from blind spots in the public health system's understanding of COVID-19 in pregnancy to the sudden removal of hospital capacity data from the CDC's website, later restored after a public outcry. The Trump administration's sudden announcement in mid-July that it was wresting control over hospital coronavirus data from the CDC has only intensified the concerns.

"We'd be the first to agree that the CDC has been deficient" in its data gathering and deployment, said Jean Ross, a president of National Nurses United. "But it's still the most appropriate federal agency to do this, based on clear subject-matter expertise in infectious diseases response."

The CDC's basic mechanism for collecting information about healthcare worker infections has been the standard two-page coronavirus case report form, mostly filled out by local health departments. The form doesn't request much detail; for example, it doesn't ask for employers' names. Information is coming in delayed or incomplete; the agency doesn't know the occupational status of almost 80% of people infected.

The data about infections and deaths among nursing home staff is more robust, thanks to a rule that went into effect in April that requires facilities to report directly to the CDC. The agency told Kaiser Health News that it is also "conducting a 14-state hospital study and tapping into other infection surveillance methods" to monitor healthcare worker deaths.

Another federal agency, the U.S. Occupational Safety and Health Administration, investigates worker infections and deaths on a complaint basis and has prioritized COVID-related cases about the healthcare industry. But it has suggested that most employers are unlikely to face any penalties and has issued only four citations related to the outbreak, to a Georgia nursing home that delayed reporting the hospitalization of six staffers and three Ohio care centers that violated respiratory protection standards. Of the more than 4,500 complaints OSHA has received about COVID-19-related working conditions in the medical industry, it has closed nearly 3,200, a

The Trump administration told hospitals to stop reporting data to the CDC, and report it to HHS instead. Vice President Mike Pence said the information would continue to be released publicly. It hasn't worked out as promised.

Data problems aren't just a federal issue; many states have fallen short in collecting and reporting information about healthcare workers. Arizona, where cases have been surging, told ProPublica, "We do not currently report data by profession." The same goes for New York state, though a report in early July hinted at just how devastating the numbers there might be: 37,500 nursing home employees, about a quarter of the state's nursing home workforce, were infected with the coronavirus from March through early June. Other states, including Florida, Michigan and New Jersey, provide data about employees at long-term facilities but not about healthcare workers more broadly. "We are not collecting data on healthcare worker infections and/or healthcare worker deaths from COVID-19," a spokesperson for the Michigan Health Department said in an email.

This problem is global. Amnesty International, in a July report, pointed to widespread data gaps as part of a broader suppression of information and rights that has left workers in many countries "exposed, silenced [and] attacked." In Britain, where more than 540 medical workers have died in the pandemic, the advocacy group Doctors' Association UK has begun legal action to force a government inquiry into shortages of personal protection equipment in the National Health Service and "social care" facilities such as nursing homes. And in May, more than three months after the first known medical worker's death, the International Council of Nurses called for governments across the world to start keeping accurate data on such cases, and for the records to be centrally held by the World Health Organization. The WHO estimates that about 10% of COVID-19 cases worldwide are among health workers. "We are closely following up (on) these cases through our global networks," a spokesperson said.

"Governments' failure to collect this information in a consistent way" has been "scandalous," said the council's CEO, Howard Catton, and "means we do not have the data that would add to the science that could improve infection control and prevention measures and save the lives of other healthcare workers. … If they continue to turn a blind eye, it sends a message that [those] lives didn't count."

So regular people, like Rezba, have stepped up with their makeshift databases.

Dr. Claire Rezba, an anesthesiologist who tracks the COVID-19 deaths of healthcare workers. (Carlos Bernate for ProPublica)

Rezba, 40, initially wanted a career in public health. While finishing her master's degree at Emory University in Atlanta and for a few months afterward, she worked as a lab tech at the CDC, analyzing nasal swabs to track cases of MRSA, the flesh-eating bacteria. But she decided she cared more about people than bugs, so she headed to Virginia Commonwealth University medical school in Richmond, graduating in 2009 with plans to specialize in the treatment of chronic pain.

During her residency at VCU, her first rotation was in the neonatal intensive care unit. "There was a little baby I helped take care of for three weeks. And the very last day of that rotation, his parents withdrew care. ... He was the first little person I pronounced dead. I went and cried in the stairwell after that." Her next rotation was in the burn unit, then the emergency department. "It seemed like death was just everywhere," Rezba said. Witnessing it "is something very separate from the rest of your life experiences. People look different when they're dying. It's not like TV. They don't look like they're sleeping. CPR is pretty brutal. Codes are pretty brutal."

She began keeping a list as a way to process the grief. "In residency, you record everything — your case logs, the procedures you do. It was just sort of second nature to record their names." Whenever a patient died she would make another entry in her notebook, then "I would kind of perseverate" — ruminate — "over their names." At the end of the year, she took the notebook to church. "I lit candles for them. I prayed. And then I let it go."

A decade later, Rezba was working full time as an anesthesiologist and raising three small children, her list-compiling days long past her, she thought. Then COVID-19 hit. The onetime infectious disease geek became obsessed with the videos leaking out of China — the teams of healthcare workers in full protective gear, the makeshift wards in tents, the ERs in chaos: "I knew early on that this was going to be a big problem." In her job, Rezba was often called upon to do intubations. "The possibility of not having enough PPE caused a lot of anxiety for her," said her husband, Tejas Patel, whom she met in medical school. "She would be the one, if we did hit that level of New York, who could potentially be at risk and bring it home to the kids."

As it turned out, Rezba's hospital wasn't inundated, nor did it experience the PPE shortages that plagued many healthcare facilities. But her anxiety didn't disappear; it just took a new shape. If healthcare workers were front-line heroes, she decided, her role was to search the trenches for the bodies left behind.

Rezba is the first to admit she's not great at technology; she rarely uses a computer at home. Patel discovered what she was doing because their iPhones and iCloud accounts are linked. "Whenever she saves a picture to the phone, I can see it. And I noticed a bunch of pictures of, you know, these strangers." He remembered how, in their student days, Rezba had insisted on humanizing the cadaver in their anatomy lab: "It upset her that it was just this anonymous person. Knowing his birthday and little things like that would make her feel better." Patel figured the photos were part of a similar coping strategy. "It wasn't until much later that I found out she was putting them up on Twitter."

Much of Rezba's digging happens in the middle of the night, when she can't sleep. She usually starts by Googling for local news stories; if she's still not tired, she turns to the obituary site Legacy.com. The hunt for a person's occupation and cause of death invariably takes her to Facebook, where she follows the trail to relatives and co-workers, to vacation slideshows and videos of old men serenading their grandkids on the guitar. Every few days, she checks GoFundMe, where she's recently been struck by the number of people who linger for weeks or months before dying. She's still discovering deaths that occurred in April and May. Anyone under 60 gets special scrutiny. "If the obit says, 'They died surrounded by family,' I usually don't bother trying to find out more, because those people didn't have COVID. The people with COVID are mostly dying alone."

Doctors and nurses are the easiest to find. "If someone worked in the laundry service at the nursing home, the family doesn't put that in," Rebza said. Yet it's the nonmedical staff that she feels a special obligation to uncover — the intake coordinators and supply techs, the food service workers and janitors. "I mean, the hospital's not going to function if there's nobody to take out the trash." Every so often, a news story mentions that several staffers from a particular nursing home or rehab center have died, without mentioning their names, and Rezba feels the rage start to bubble. "What it comes down to is, these are people that are making $12 an hour. And they get treated like they're disposable."

If she can't find someone's identity right away, or if the cause of death isn't clear, she'll wait a couple of days or weeks and try again. Because she comes across them anyway, she's started to keep track of other categories of COVID-19 deaths, like kids and pregnant women, as well as healthcare workers in their 30s and 40s who don't appear to have the virus but suddenly perish from heart attacks or strokes or other mysterious reasons. "I have a lot of those," she said.

Once she's certain she's found someone who belongs on her list, she selects a photo or two and writes a few words in their honor. Sometimes, these read like a scrap of poetry; sometimes, like a howl.

He enjoyed crazy-dancing at home to Bruno Mars, with the moves becoming wilder the more his family laughed.

As a child, she would wrap her clothes around Dove soap so they would smell like America.

This poor baby should have his mother in his arms. Instead he has her in an urn.

A preprint study out of Italy last week hinted at the kind of lessons researchers and policy makers might glean if they had more complete data about healthcare workers in the U.S. The study pooled data from occupational medical centers in six Italian cities, where more than 10,000 doctors, nurses and other providers were tested for coronavirus from March to early May. Along with basic demographic information, the data included job title, the facility and department where the employee worked, the type of PPE used and self-reported COVID-19 symptoms.

The most important findings: Working in a designated COVID-19 ward didn't put workers at greater risk of infection, while wearing a mask "appeared to be the single most effective approach" to keeping them safe.

In the U.S., many medical facilities are similarly monitoring employee infections and deaths and adjusting policies accordingly. But for the most part, that information isn't being made public, which makes it impossible to see the bigger picture, or for systems to learn from each other's experiences, to better protect their workers.

Imagine all of the opportunities it would present if everyone could see the full landscape, said Ivan Oransky, vice president for editorial content at Medscape, where a memorial page to honor global front liners has been one of the site's best-read features. "You could be doing some real great shoe-leather epidemiology. … You could go: 'Wait a second. That hospital has 12 fatalities among healthcare workers. The hospital across town has none. That can't be pure coincidence. What did this one, frankly, do wrong, and what's the other one doing right?'"

To Adia Harvey Wingfield, a sociologist at Washington University and author of "Flatlining: Race, Work, and Healthcare in the New Economy," some of the most pressing questions relate to disparities: "Where is this virus hitting our healthcare workers hardest?" Is the impact falling disproportionately on certain categories of workers — for example, doctors vs. registered nurses vs. nursing aides — on certain types of facilities, or in certain parts of the country? Are providers who serve lower-income communities of color more likely to become ill?

"If we aren't attuned to these issues, that puts everybody at a disadvantage," Wingfield said. "It's hard to identify problems or identify solutions without the data." The answers are especially important in Black and Latino communities that have suffered the highest rates of sickness and death — and where healthcare workers are themselves more likely to be people of color. Without good information to guide current and future policy, she said, "we could potentially be facing long-term catastrophic gaps in care and coverage."

Not only are Hispanics catching coronavirus at higher rates in Texas' largest county, they also suffer some of the worst outcomes.

The near-term consequences have also been enormous. The lack of public data about healthcare workers and deaths may have contributed to a dangerous complacency as infections have surged in the South and West, Friese said — for example, the idea that COVID-19 is no more dangerous than other common respiratory viruses. "I've been at this for 23 years. I've never seen so many healthcare workers stricken in my career. This whole idea that it's just like the flu probably set us back quite a way."

He sees similar misconceptions about PPE: "If we had a better understanding of the number of healthcare workers infected, it might help our policymakers recognize the PPE remains inadequate and they need to redouble their efforts. … People are still MacGyvering and wrapping themselves in trash bags. If we're reusing N95 respirators, we haven't solved the problem. And until we solve that, we're going to continue to see the really tragic results that we're seeing."

The misconceptions appeared to stretch to the highest reaches of the federal government, even as infections and deaths started surging again. At a White House event in July focused on reopening schools in the fall, HHS secretary Alex Azar told the people gathered, "healthcare workers … don't get infected because they take appropriate precautions."

Even some medical workers have continued to be in denial. A few days before Azar spoke, Twitter was abuzz over an Alabama nurse who worked the COVID-19 floor at a hospital by day and decompressed at crowded bars by night, where she often went maskless. "I work in the healthcare industry," she was quoted as saying, "so I feel like I probably won't get it if I haven't gotten it by now."

Piercing that sense of invulnerability — making the enormity of the COVID-19 disaster seem real — isn't only Rezba's mission. From The New York Times' iconic front page marking the first 100,000 American deaths to the Guardian/Kaiser Health News project "Lost on the Frontline," news organizations and social media activists have grappled with how to convey the scale of the tragedy when people are distracted by multiple world-shattering crises and the normal rituals for processing grief are largely unavailable.

"The point at which accountability usually happens is when our leaders have to reckon with the families of those who've been lost, and that has not happened," said Alex Goldstein, a Boston-area communications strategist behind the wrenching @FacesOfCOVID Twitter account, which has posted almost 2,000 memorials since March. With COVID-19, "no one has had to look in the eye of a crying parent who wants to show you a picture of their child or listen to someone telling you about who their mom or dad was. There has been no consequence. What would our policy decisions have looked like if [the people making them] had to come face to face with that death and loss in a more visceral way?"

It's a question that weighs especially heavily on healthcare professionals, who have seen, in the most visceral way possible, the worst that COVID-19 can do. Erica Bial, a pain specialist in the neurosurgery department at a Boston-area hospital, fell dangerously ill from COVID-19 in March, her respiratory symptoms lingering for more than six weeks. She lived alone and opted not to go to the hospital, in part because she worried about infecting other people. "At that point [in the outbreak], they would have intubated me, given me hydroxychloroquine and azithromycin and probably killed me." As her recovery dragged on, she wondered how other doctors were faring: "I couldn't believe that I was the only physician I knew who was sick." But as she searched online, "I could not find any data. I just started getting really frustrated at the lack of information and the disinformation. … And then I started thinking about, well, what happens if I die here? Will anybody know?"

Like Rezba, Bial has a background in public health; the Facebook page she created, COVID-19 Physicians Memorial, was an attempt to build "a network where there's accountability. I wasn't necessarily trying to create, you know, reverence or memorialization. I was trying to understand the scope of the problem."

Rezba soon began posting memorials on the page; as it grew to include more than 4,800 members, Bial asked her to help moderate it. Among the things the two women share is a determination to stick to facts. "I didn't want any politics and I didn't want any garbage," Bial said. "(Rezba) was 100% like-minded and trustable." She was also someone Bial could talk to, doctor to doctor, as she recovered. "It wasn't just two people obsessed with something kind of morbid," Bial said. "She was a source of support."

Emergency room doctor Cleavon Gilman also gained a following for his posts on Facebook, a diary about what he witnessed as an ER resident in the NewYork-Presbyterian hospital system, battling the virus as it engulfed Washington Heights. "It was just … overwhelming," he recalled. "We were intubating 20 patients a day. We had hallways filled with COVID patients; there was nowhere to put them." In the space of a few brutal days in late April, three of Gilman's colleagues died, including one by suicide. "When it's a colleague that you're taking care of and you know them as a person you've been on a journey with ... man, that's hard."

A tweet from Rezba's account sharing the work of Dr. Cleavon Gilman, who is also compiling healthcare workers who have died from COVID-19.

Though much of the media focus was on the risks faced by older patients, Gilman was struck by how many of the critically ill were in their 20s, 30s and 40s. In mid-April, his own 27-year-old cousin, a gym teacher at a New Jersey charter school, suddenly died; he went to the ER twice with chest pain but was diagnosed with anxiety and sent home, according to his relatives, only to collapse in his car on the side of the road.

As the crisis in New York City ebbed, Gilman could see trouble ahead in other parts of the country, including in Yuma, Arizona, where he was about to start a new job. It seemed vitally important to help younger people understand the risks they faced — and that they created for others — by not adhering to physical distancing or wearing masks, not to mention the dangers that healthcare workers faced from continuing shortages of PPE. So Gilman began gathering the memorials he saw on Twitter and Facebook, many of them found by Rezba or on @FacesOfCOVID, and organizing the dead on his website in the type of gallery that he knew would pack an emotional wallop. Then he went a step further, making the photos and obituaries — more than 1,000 people — sortable by age and profession.

"You begin to see a pattern here," he said. "When someone says, 'Oh people aren't dying, they're not that [young],' you can come back with actual names, actual articles, quickly. It's more powerful. You have your evidence there."

One of the most overtly political projects is Marked by COVID, formed by Kristin Urquiza in honor of her father, Mark, after her "honest obituary" of him went viral in early July. To Urquiza, who earned her master's in public affairs from the University of California, Berkeley, and works as an environmental advocate in the San Francisco area, "the parallels between the AIDS crisis and what is happening now with COVID are just mind-boggling [in terms of] the inaction by governments and the failure to prioritize public health." She and her partner, Christine Keeves, a longtime LGBTQ activist, hope the project will be both "a platform for people to come forward and share their stories" and the COVID-19 version of the anti-AIDS group Act Up.

They're also raising money on GoFundMe to help other families pay for obituaries; the second honest obit on their site was for a respiratory therapist in Texas named Isabelle Odette Hilton Papadimitriou: "Her undeserving death is due to the carelessness of politicians who undervalue healthcare workers through lack of leadership, refusal to acknowledge the severity of this crisis and unwillingness to give clear and decisive direction to minimize the risks of coronavirus. Isabelle's death was preventable; her children are channeling their grief and anger into ensuring fewer families endure this nightmare."

It's a trend that Rezba supports wholeheartedly. By the end of July, she had posted almost 900 names and faces of U.S. healthcare workers who had perished from COVID-19. She fantasized about what it would be like to leave the counting behind her. "It would be great if I could stop. It would be great if there was nobody else to find." But she had a backlog of dozens of stories to post, and the number of deaths kept climbing.

Ryann Grochowski Jones and Hannah Fresques contributed reporting.

 

 

The Trump administration told hospitals to stop reporting data to the CDC, and report it to HHS instead. Vice President Mike Pence said the information would continue to be released publicly. It hasn't worked out as promised.

This article was first published on Friday, July 31, 2020 in ProPublica.

By Charles Ornstein

In mid-July, the Trump administration instructed hospitals to change the way they reported data on their coronavirus patients, promising the new approach would provide better, more up-to-the-minute information about the virus’s toll and allow resources and supplies to be quickly dispatched across the country.

Utah prioritized the health of businesses over keeping coronavirus case counts down. As case counts rise, the state will now allow indoor gatherings of up to 3,000 people.

This article was first published on Tuesday, July 21, 2020 in ProPublica.

By Lisa Song and Mollie Simon

Back in April, when public health officials were still helping lead Utah's response to the coronavirus, the spread of the disease had slowed, stabilizing at fewer than 200 reported cases a day.

Then came a shift in power, and priorities.

State legislators who felt Gov. Gary Herbert was not moving quickly enough to lift restrictions on businesses created a commission to set guidelines to reopen. "It's not meant to give economic outcomes a higher weight, but it is time to give them some weight," said Sen. Daniel Hemmert, a Republican who sponsored the bill and took his seat on the commission alongside other politicians, bureaucrats and business leaders.

Email correspondence and interviews with more than a dozen state and local officials in Utah show that the health of the state's businesses was prioritized over the health of the public, as officials stopped slowing the spread of the virus and instead calculated how many sick people its health system could bear.

Dr. Joseph Miner, executive director of the Utah Department of Health, told ProPublica that state leaders originally planned to relax restrictions as cases decreased. But "because of the concern that you can't keep the economy closed this long," they reopened before that happened and shifted their attention to how many cases hospitals and contact tracers could handle.

"We know there's going to be increased cases. We just said, amongst ourselves, this is really what we're addressing: our capacity to respond rather than decreasing numbers."

With key health experts cut out of the decision-making process, including the state epidemiologist and local officials who were stripped of their ability to issue their own restrictions, the governor and the commission quickly swept aside restrictions meant to slow the virus's advance in Utah. You can now hold indoor events with up to 3,000 people and outdoor events with up to 6,000. You can drink at bars, eat in restaurants and go to the movies.

People in Utah are going back to work; new unemployment claims from early July dropped 78% since the peak from early April, when most businesses were closed, and its unemployment rate is just above 5%, which is under half of the national average. Utah was considered one of the states best-prepared to weather the pandemic's economic downturn. But experts say that rising infections could threaten any state's recovery.

And in Utah, infections are rising. The percentage of tests that come back positive is at 10% as of July 13, compared with 3% to 5% in April. On July 14, the state reported its highest number of deaths on a single day since the pandemic began. Since late May, the seven-day average of daily case counts statewide has quadrupled.

Utah's story is mirrored in states across the country, where leaders sidelined public health experts and forged ahead without meeting criteria scientists say are necessary to reopen. In Florida, the governor loosened restrictions as cases rose; Miami is now the national epicenter of the virus. Georgia left it up to businesses to decide how much to scale back service in the middle of a pandemic; South Carolina let the hospitality industry write its own reopening guidelines. The virus is now spreading uncontrolled in those two states, with some of the highest rates in the country. Texas, whose governor took advice on reopening from a panel stacked with corporate executives and business leaders, reopened restaurants and malls after reaching a record in daily deaths; its hospitals are now stretched to the brink.

Herbert's office did not respond to questions about how he's managed the pandemic. In an emailed statement, Jefferson Burton, the commission's co-chair, said the group regularly reviews Utah's reopening guidelines and makes changes based on new data, federal guidance and input from stakeholders. Decisions on when to lift restrictions are based on hospital utilization, infection growth rates and the number of tests performed, among other factors, the statement said. "There are not hard-and-fast data points that automatically trigger a move. … Rather, individual geographic areas are evaluated looking at their specific trends over time."

Dr. Michael Good, CEO of University of Utah Health and the only physician on the commission, said the "goal has always been to slow the spread of the virus." The commission is trying to balance decisions on reopening in a way that keeps infections low while not inadvertently causing other social and economic problems, he said.

Epidemiologists say that one of the biggest risks of reopening prematurely is that it suggests formerly banned activities are now safe. The family of a diabetic man who died after going to a party in Riverside County, California, said he had been careful until the government eased restrictions.

The timing of Utah's spikes in cases clearly overlaps with the loosening of restrictions, said Dr. Emily Spivak, an infectious disease physician at University of Utah Health. People assumed the risk was over. Spivak has seen group barbecues and young people taking party trips to Lake Powell with "full-on 20-year-old summer social behavior." Of course case numbers went up, she added. "It's not rocket science."

A Shift in Power

Initially, Utah acted decisively to try and stop the spread of the coronavirus. In March, working with health advisers, Herbert decided to shut down schools days before New York City and restricted businesses based on their level of risk. Restaurants could not offer dine-in service; gyms and salons were closed.

While Herbert refused to issue a statewide shutdown, saying he didn't want to overrestrict less affected areas, mayors and local health officials wrote their own stay-at-home orders, which came with fines and even criminal penalties for violations. The state was a patchwork of legally binding rules, as The Salt Lake Tribune reported. "We give local control for the regional differences," Herbert said at the time, "and I think we've struck the right balance."

Doctors treating coronavirus patients were supposed to be allowed into the U.S. But hundreds of young doctors have their visas put on hold indefinitely.

Senate President Stuart Adams, a Republican businessman, disagreed. On April 8, he confirmed that lawmakers would convene to address the pandemic and consider limiting the power of local governments, saying their rules were creating confusion. It was time, he told the local paper in Ogden, for "a new phase" focused on the economy.

Sen. Hemmert and Rep. Mike Schultz co-sponsored a bill to establish the Public Health and Economic Emergency Commission, a 10-member team that would guide the reopening and advise Herbert. Staff from the Governor's Office of Management and Budget would provide support.

Apart from two healthcare system CEOs, none of the members had a medical or public health background. Among them: Adams, the two Republican legislators who sponsored the bill, the president and CEO of the Salt Lake Chamber of Commerce and the CEO of the Larry H. Miller Group of Companies, which includes sports teams such as the Utah Jazz as well as a chain of Megaplex Theatres across the state.

By the time the commission assembled, there was a new leader in charge of the state's coronavirus response.

Miner, a physician who had headed the state Health Department since 2015, has a lung condition and had been unable to attend in-person meetings. To be the "boots on the ground" for the the virus response, the governor appointed Burton, a retired military leader with no medical training who had experience in disaster management. Burton once headed Utah's National Guard and is running for a seat in the state House of Representatives. While Miner said he has remained "very much involved," he was not present, even virtually, for key decisions ProPublica asked about. He said he was briefed on outcomes by Burton's deputy.

During the time that Burton has served as co-chair of the commission, the three legislators on the commission have donated (either personally or through their campaigns) a total of $4,000 toward his run for state office, one-fifth of what his campaign has raised. He won the Republican primary in June and has no Democratic challenger in the general election.

The governor's office and a large team of advisers had created a color-coded system for reopening that could be turned up or down like a dial "based on the health risk." The system had four levels that dictated how businesses could operate, labeled red ("high risk"), orange ("moderate risk"), yellow ("low risk") and green ("new normal risk"). All of Utah was under the red designation in late April; each successive level would open more businesses and ease limitations.

The commission took the governor's guidelines and added specific rules for every industry, including restaurants, schools, entertainment venues and religious services. Five days after the group was created, it recommended moving the state to orange.

Herbert accepted the plan and moved Utah to the "moderate risk" level on May 1. Hotels and gyms opened. Groups of 20 could congregate with masks and social distancing.

Another Shift in "Risk"

Guidelines from the Centers for Disease Control and Prevention say one of the key factors for relaxing restrictions is evidence of a 14-day decline in new cases.

By May 12, cities and counties were expressing an interest in moving into the yellow, "low risk" level. Cases had plateaued, but had not dropped. Dr. Angela Dunn, the state epidemiologist, told KSL News Radio that on an "optimistic" timeline, some parts of the state might be ready to move into yellow on June 1.

But on May 14, just 13 days after moving the state into orange, the governor and the commission announced they were moving most of the state into the "low risk" level, on May 16. It was impossible to see, at that point, the full effects of the orange phase after half a month, because of the incubation period of the virus and the lag between symptoms, testing and test results.

Jenny Wilson, the mayor of Salt Lake County, petitioned to keep her county in orange; she noted that the active positive case rate there was several times higher than the state's. She could have kept the restrictions before the bill creating the commission passed, but now she needed the governor's approval. Herbert denied the appeal, but he approved similar petitions from the two largest cities in her county — Salt Lake City and West Valley City — and three other counties.

The rest of the state relaxed further into the phase labeled "low risk." All businesses could resume with certain precautions. Public swimming pools opened; close-contact team sports were allowed with temperature checks. Restaurants could serve buffets.

"Speaking [with] my public health voice, I feel that this is a mistake," Jennifer Dailey-Provost, a Democratic state representative who's pursuing a doctorate in public health, tweeted in reaction. "To say I'm frightened is an understatement. I hope I'm wrong."

Two weeks later, on May 27, cases began to surge.

A spokesman from the Salt Lake County Health Department said it's hard to say if the county would have fewer cases today if the entire county (and not just two cities) had stayed in orange. Commuting patterns make it impossible to separate different municipalities, and the piecemeal approach left some streets orange on one side and yellow on the other.

Hospitalization data is important to understanding the coronavirus's spread and impact. But after the Trump administration changed its reporting rules, the CDC removed the data from its site, and only added it back after a public outcry.

County data shows that throughout June, on days with a particularly high number of new cases, more than half came from areas that went yellow before the mayor felt they were ready.

Salt Lake City Mayor Erin Mendenhall said the commission has made it more cumbersome for local leaders to pass their own rules, but she gives the governor credit for allowing her city to remain in orange and approving the county's mandate for wearing face masks in public.

One of the counties Herbert had allowed to stay in orange was Grand County, home to Arches National Park. But that changed on May 28. When local officials asked to maintain extra protections as they moved to yellow — such as leaving hotel rooms empty for 24 hours between bookings — the governor's office denied the request.

The rural county has logged few cases, though it is impossible to tell how many tourists may have caught the virus there before returning home. Bradon Bradford, director of the Southeast Utah Health Department, which covers Grand County, said the local numbers started going up in July.

The commission's top-down approach is "uncharacteristic" of Utah, said Kirk Benge, who leads the Public Health Department for San Juan County in the southeastern corner of the state. "Most politicians claim that they like local authority and they like local decisions. In an emergency, to immediately strip that … I felt, was a mistake at the time."

A Mysterious Decision

Top health officials and other legislators have little insight into the commission's actions. Miner told ProPublica he has never attended a meeting. Nor has Dunn, the state epidemiologist. Both are part of a work group that provides input on what activities are allowed under each phase, but they aren't involved in the final decisions of when and how restrictions are loosened.

Dailey-Provost, the legislator with public health training, said her offers to help and suggest health experts went nowhere. She co-authored a study in April that predicted how coronavirus cases would peak in Utah. Reality has outstripped the study's worst projections.

One of the commission's most significant decisions occurred in late May, when they suggested redefining the yellow phase, which capped gatherings at 50 people, to allow indoor gatherings of 3,000 and outdoor gatherings of 6,000.

The proposal caught Dunn and other health officials off guard. They discussed it at their work group meeting of health and business experts. Burton's deputy, Richard Saunders, who often attends commission meetings, was also present.

Benge, the health officer for San Juan County, said he was "100% against" the change. Many businesses were still transitioning into the yellow phase and hadn't had time to open up, Benge said, so it was too early to loosen the guidelines any further. Dunn and Miner shared similar concerns.

Miner said he recalls the group was being asked to figure out how to implement the change as safely as possible. He said they concluded that if the venues were required to track where everyone sat, at least they could do contact tracing if anyone got sick.

The numbers were headed in the wrong direction. A day earlier, on May 28, the Tribune reported the biggest single-day spike in new cases statewide, as well as an outbreak in a nursing facility that infected more than half the residents. One county in northern Utah saw a 33% increase in cases.

Health officers fretted about the decision in emails released to ProPublica.

"This change happened on a timeline contrary to my recommendations and the recommendations of the state epidemiologist and other health officers," Benge wrote to his county health board.

"I don't think we'll need to move to Green because it's all being phased into Yellow," Bradford wrote to other local health officials.

"So who is going to count to see when they reach 6001 participants?" another officer replied.

Herbert accepted the commission's changes in June.

Jordan Mathis, the officer for the TriCounty Health Department that oversees Daggett, Duchesne and Uintah counties, said the numbers seemed arbitrary. "Why 3,000? Why 6,000? Where'd we pull those from?"

Good, the commission member, said he doesn't recall the exact reasons behind those numbers, but those discussions "were occurring at a time when 999 out of 1,000 Utahns did not have a coronavirus infection," he said. Today, the situation is more serious, with four times as many active cases as there were back then, he said. "Those are not the conversations we would be having today."

Brian Hatch, the health officer for Davis County, sits on a medical team that is supposed to advise the commission. He doesn't recall the commission ever asking for its advice on the phase guidelines. The medical team has focused on recommendations for high-risk residents who are especially vulnerable to COVID-19.

Hatch said the idea for 6,000 people outdoors might have come from his county. When the state moved to yellow, there was no reopening plan for Lagoon Amusement Park, the only amusement park in the state. The park owners worked with Hatch. They settled on 6,000 people, which is 15% of its usual capacity. With social distancing, mandated masks and other precautions like timed tickets, they could operate safely, Hatch said.

Dozens of New York nursing homes didn't see their first COVID-19 case until sick patients were sent there, many under Andrew Cuomo's state policy. To date, 6% of the state's nursing home population, or roughly 6,500 residents, have died.

The park reopened Memorial Day weekend with the governor's approval, a week before the commission meeting on May 29. A spokesman for the park said it reached the 6,000-person limit only a few times since opening. Hatch said there have been no outbreaks connected to the park, and having a large crowd dispersed outdoors is very different from cramming 3,000 spectators indoors. Since the risk of infection is much higher inside, "I'm concerned with 20 people together," he said.

Hatch said he doesn't know where the 3,000 number came from. Health officials couldn't point to any recent events with thousands gathered indoors. A spokeswoman for the Larry H. Miller Group, whose CEO sits on the commission, said the company's movie theaters have kept indoor crowds at 50 and have not hosted any sporting events at their large venues.

The indoor guidelines require assigned seating (to enable contact tracing) and masks whenever social distancing isn't possible.

Dunn and Benge said masks alone aren't enough. Masks need to be used in concert with distancing and hand washing, said Dr. Georges Benjamin, executive director of the American Public Health Association. "(The commission is) using the masks as an excuse to break all the other science rules."

The High Risk of "Low Risk"

Shortly after the May 29 meeting where the work group discussed the changes to the yellow guidelines, the commission held its own meeting.

According to meeting minutes, the commission wanted to determine how and when the state could switch to green, the "new normal." One of the people present was Burton's deputy, Saunders, who had just attended the work group meeting with Dunn and Miner.

Saunders told commission members that the potential impact from recent events — including Memorial Day celebrations and the opening of the amusement park — would become clear in the coming week. He said epidemiologists in the state Health Department advised staying at yellow until June 30.

Adams motioned to move most of the state to green by June 5, as long as the data supported it. The transition meant large crowds could gather without assigned seating. Religious services would no longer require 6 feet of space between families and sports competitions could resume.

The six commission members who were present voted unanimously for the idea. They waited several days to announce the news.

Utah had cycled through two phase changes in one month. Dunn worried residents saw it as permission to abandon precautions like masks. After all, the yellow phase was labeled "low risk."

"I am concerned that we are providing the public with false information regarding their risk for contracting COVID-19," she wrote to Miner, Burton and others on June 1. "Our % positive is at 7%, just a short jump away from the 8% positive at the beginning of this outbreak when we were only testing hospitalized patients. Our growth rate is sharply increasing."

The very next day, the commission said the state was ready to go to green.

When Benge heard the news, he told colleagues he was "concerned that the current 'Phased Health Guidelines' have gradually shifted at the state level, from being focused on protecting health to being more focused on protecting the economy."

The commission acknowledged coronavirus cases "may continue to rise" as restrictions are lifted. But case numbers are "a poor indicator of health risk for all Utahns," as 99% of COVID-19 patients recover, they said in a press release that cited low hospitalization rates, low death rates and increased testing and contact tracing. They called for a "smart" green level where everyone should still wear masks and stay socially distant.

They didn't mention the racial disparities that persisted throughout the pandemic. Fourteen percent of Utah's population is Latino, but they make up 40% of cases. The proportion of patients who die from the virus is three times as high for Native Americans as it is for white residents. Navajo Nation, which extends into Utah, Arizona and New Mexico, recently imposed stay-at-home orders over several weekends because of the worsening situation in surrounding areas. Most of the roads leading to Utah's national parks go through the reservation, said Pete Sands, a spokesman for Utah Navajo Health Systems. So when residents or visitors refuse to take precautions, it directly affects Navajo citizens.

Too many states are letting their cases grow based only on healthcare capacity, said Benjamin, the American Public Health Association director. "Why open the economy and allow needless death and illness and disability … [when] this single-minded strategy is eroding the economy in the long term?" he said. "Sick people can't work. People who are afraid to go out and shop [or] eat aren't going to go out."

Experts say there's more at play than ensuring hospital beds; workers face personal protective equipment shortages and burnout. There's no surefire way to prevent COVID-19 deaths, and many survivors are left with heart damage, scarred lungs, neurological problems and other long-term effects doctors are just beginning to understand.

A day after the commission's announcement on June 2, Dunn told reporters that no community in Utah was ready for green.

Nearly 8,000 ventilators are destined for foreign countries as part of Trump's plan to make the U.S. "king of ventilators." But public health experts worry the machines are crowding out more urgently needed aid.

Herbert waited until June 12 before moving one county to green. A week later, he approved requests from nine other counties to do the same.

As Herbert finalized the partial shift to green, Dunn sent an urgent memo to state and local health officials, which the Tribune published several days later on June 22.

"If we do not reach a rolling 7-day average of 200 [cases] per day by July 1, we need to move the entire state to orange," Dunn wrote. "This will send the message to Utahns that this outbreak continues to be a serious problem."

"This might be our last chance for course correction," she warned. "Contact tracing and testing alone will not control this outbreak."

Dunn recommended a statewide requirement for face masks. If that wasn't possible, she wrote, "we need to be clear with the public about why decisions are being made lessening restrictions — economic, not health."

Herbert said he appreciated Dunn's analysis but would not close down the economy. The seven-day average stood at 485 cases per day.

Beyond Capacity

By mid-July, the seven-day average had reached 650 cases per day.

A growing number of health and business interests want a statewide mask mandate. That includes the Salt Lake Chamber of Commerce and the Larry H. Miller Group, whose leaders sit on the commission. Adams has come out against the idea.

Herbert has required masks in state buildings and in K-12 schools starting this fall but stopped short of a blanket rule, citing individual freedom and local control. The issue has become so politicized in Utah that one county commissioner compared the idea of a mask mandate to Nazism. In Provo, residents crowded into a county commission meeting to protest the school mask requirement.

Darin Mellott, a real estate executive who serves on a separate economic task force on the pandemic, describes himself as an establishment Republican. But he personally feels masks are an easy way to stem the tide. "I think future generations, if we do nothing, are going to look back and say, 'Why did you subject so many people … to this threat, because of some imagined threat to our liberty?'"

Mellott said Herbert has a tough balancing act — "I think we would be in a much worse situation if it weren't for him" — but that the governor and other state leaders need to give more health professionals a seat at the table. "This is a war against the virus, and the medical professionals are the generals," he said. "So listen to the generals."

In a recent press conference, Herbert acknowledged that labeling the different phases based on risk might have given the public a false sense of security. He challenged residents to voluntarily wear masks and set a goal to keep average new cases below 500 a day by August 1. Herbert cited 800 a day as the absolute maximum the state can handle.

Dunn told ProPublica she suggested 200 cases a day because it allows the state to do contact tracing within 24 hours. Because of the lag time between infections and hospitalizations, any preventative measures taken today won't have an effect for another two to four weeks, so there's no time to lose. "You can't wait until you're already underwater," she said.

The state is hiring more contact tracers, but can now handle only 300 cases a day.

For the world's best-known corporate-management consultants, helping tackle the pandemic has been a bonanza. It's not clear what the government has gotten in return.

This article was first published on Wednesday, July 15, 2020 in ProPublica.

By Ian MacDougall for ProPublica

In the middle of March, as the coronavirus pandemic was shutting down the country, McKinsey & Co., the giant management consulting firm, saw opportunity. The firm sprang into sales mode, deploying its partners across the country to seek contracts with federal agencies, state governments and city halls. Government organizations had been caught unprepared by the virus, and there was a lot of money to be made advising them on how to address it.

That month, a partner in McKinsey's Washington, D.C., office, Scott Blackburn, got in touch with an old colleague. Deb Kramer had just been promoted to become an acting assistant undersecretary at the Department of Veterans Affairs, where Blackburn, whom McKinsey declined to make available for an interview, had held senior roles between 2014 and 2018. During that period, the two had overseen a major overhaul of the agency called "MyVA," a project McKinsey had worked on as well. Blackburn had worked at McKinsey before going to the VA, and he returned to the firm afterward. He and Kramer were in touch repeatedly in the middle of March, according to a person familiar with the exchanges.

On March 19, Kramer made a highly unusual request: The VA, she said, needed to hire McKinsey within 24 hours. The VA runs a sprawling health care system that serves 9 million veterans, many of them older and plagued by chronic health problems, and typically takes many months to solicit and accept bids and vet bidders for a contract. The health system's leadership wanted to sign a multimillion-dollar contract with McKinsey to spend up to a year consulting on "all aspects" of the system's operations during the COVID-19 pandemic, Kramer told a VA contracting officer, Nathan Pennington. Pennington memorialized parts of the exchange in a public contracting document.

"There is no time to spare," the contracting document stated, "every day wasted by a lack of situational awareness down to the community level, and the inability to model scenarios and test alternative courses of action, increases the risk to the citizens of this nation, to include Veterans and our own employees." The VA, the document observed, needed help with "life-and-death decision-making today."

The exigent circumstances left no time to seek competing bids or to fully vet McKinsey's proposal, Kramer argued. It was the only contractor she and her colleagues were aware of that could provide the required services without needing "ramp-up" time the VA couldn't afford. Pennington conducted no market research and only a minimal review of the cost, "as there was no time," he wrote. Kramer approved the $12 million price tag. The contract was signed on March 20.

That would turn out to be just a down payment for McKinsey. Ten days later, the Defense Health Agency was added to the VA contract, upping its value to $22.5 million, and the week after, the Air Force hired McKinsey — also with a no-bid contract. The firm's assignment for the Air Force was to serve on a task force developing a strategy to get defense contractors, many of them McKinsey clients, to produce medical supplies during the pandemic. To justify the $12 million value of that contract, an Air Force contracting document cited what the VA had agreed to pay McKinsey. It did not mention the VA price tag's largely unvetted nature. Finally, in early May, the DHA expanded the scope of McKinsey's work, signing an additional contract worth up to $6.1 million.

A VA spokeswoman, Christina Noel, said that the agency "adhered to all federal contracting laws" in hiring McKinsey and that "no-bid contracts can help provide VA the flexibility needed during this national emergency to deliver the services required to support clinical needs and save lives." A DHA spokesman, Richard Breen, said McKinsey had "expertise needed and an existing contract with [the VA] for COVID-19 modeling support with a separate and distinct scope." An Air Force spokesperson did not respond to emailed questions.

In a matter of weeks, McKinsey had extracted a total of $40.6 million in no-bid contracts out of its initial agreement with one federal agency. The firm has continued to scoop up COVID-19-related contracts for various governments since then. Altogether, in the four months since the pandemic started, the firm has been awarded work for state, city and federal agencies worth well over $100 million — and counting.

Many of the most prominent government pandemic efforts have been staffed with battalions of McKinsey's trademark dark-suited young MBAs. The joint coronavirus task force operating out of the Federal Emergency Management Agency and Department of Health and Human Services enlisted McKinsey, on a pro bono basis, to help obtain medical supplies. New York Gov. Andrew Cuomo's team hired McKinsey to draw on existing epidemiological models to project hospital capacity and medical supply needs. The Food and Drug Administration retained the firm to do data analysis.

Among states, California, Illinois, Massachusetts, New Jersey, Tennessee and Virginia have worked with McKinsey. (Comprehensive contracting data is not yet available, and only some states have revealed the dollar value of their contracts. For example, the contracts for New York and New Jersey represent a combined $18 million in revenue for the firm.) Cities including Atlanta, Chicago, Los Angeles, New Orleans and St. Louis have also used the firm during the pandemic.

It's too early to fully judge these engagements, but a preliminary assessment shows mixed results. After early stumbles, New York state and New Jersey are doing relatively well. On the other hand, the FEMA/HHS task force has faced harsh criticism for a slow and dysfunctional effort to procure supplies. And project documents and interviews show that, at the VA, consultants have been slow to deliver urgently needed data. In other places, officials have denigrated McKinsey's contributions. "Basically, they are compiling data for us," a top official in Florida's Miami-Dade County wrote in an internal email obtained by ProPublica. "And putting it in pretty formats." That contract was for up to a month of work, with a price tag of up to $568,000, and the confusing set of reopening guidelines that emerged with McKinsey's help has been widely panned.

McKinsey defended its work in a written statement the firm sent in the name of Liz Hilton Segel, its managing partner for North America operations: "Like many other companies, we chose to engage and do our part in helping governments fight this pandemic." The statement noted that McKinsey "has the capabilities to support leaders and public servants who are navigating this humanitarian and economic crisis. … We are proud of the support we have provided to public sector leaders, front line staff and those engaged in fighting this pandemic."

Given that McKinsey consultants operate as advisers, with government officials charged with making final decisions, it can be hard to identify the firm's responsibility for any given decision. But the firm's government work has been steadily rising in the wake of a multidecade hollowing out of government (a trend McKinsey has promoted and ridden). Today, that increasingly means that if you examine the government's response to the pandemic, you're likely to find McKinsey's fingerprints.

Like countless organizations, McKinsey has been buffeted by the pandemic, encountering turbulence and uncertainty. For starters, the sputtering economy put many of its corporate clients under duress.

Then there was a bevy of self-inflicted problems.

McKinsey's bankruptcy practice, which would normally thrive during hard times, has been under a "black cloud," as a lawyer representing McKinsey put it in a court hearing in April. The firm's practice has been dogged by a federal investigation into potentially criminal self-dealing and tied up in litigation with the founder of a rival firm over whether McKinsey properly disclosed possible conflicts of interest. ("McKinsey's bankruptcy disclosure practices have always complied with the law," Gary Pinkus, chairman of the firm's North America operations, said in a written statement.)

More broadly McKinsey has seen its long-gilded reputation tarnished in recent years as government projects come under critical scrutiny. Media outlets, such as ProPublica, The New York Times and The Wall Street Journal, have investigated a variety of ethically and legally dubious actions. That included helping the Trump administration execute exclusionary immigration policies, corruption allegations against local companies McKinsey worked alongside in Mongolia and South Africa, and a pattern of hiring the children of high-ranking officials in Saudi Arabia. In each instance, McKinsey has denied wrongdoing. With $10 billion in annual revenues, the firm is now as big or bigger than many of its clients and has developed a culture that resists oversight.

Newly uncovered documents show the consulting giant helped ICE find "detention savings opportunities" — including some that the agency's staff viewed as too harsh on immigrants.

In April, McKinsey was penalized for running afoul of the federal government. The General Services Administration, which oversees federal procurement, canceled two government-wide contracts, one of which had earned the firm nearly $1 billion between 2006 and 2019. In a report issued earlier by GSA's internal watchdog, investigators revealed that McKinsey had refused to comply with an audit. Instead, the firm went over the head of a contracting officer and found a GSA supervisor who was willing to accommodate the firm. That supervisor worked with McKinsey to improperly inflate the contract prices, the investigators found, part of a troubling pattern of favoritism the supervisor showed toward McKinsey.

For nine months, GSA negotiated with McKinsey to lower its rates. McKinsey's intransigence ultimately led officials to see cancellation as the best option, according to a statement from a senior GSA official, Julie Dunne. (DJ Carella, a spokesman for McKinsey, which denied wrongdoing, said in a statement, "We are disappointed with GSA's decision and look forward to potentially returning to the GSA schedule in the future.")

The fallout devastated the firm's U.S. public-sector practice, current and former consultants say. "The public-sector practice was already underutilized after the [GSA] report," one of them said. "And then it just stopped."

For McKinsey, the pandemic provided a new opportunity to regain its foothold in the federal government. Coordination from the White House was inconsistent at best, and many state, federal and city agencies were already short staffed. Years of budget cuts and anti-big government policies had left them dependent on outside contractors even in ordinary circumstances.

So dependent that McKinsey consultants on the FEMA-HHS task force ended up working in the procurement process, according to a federal official briefed on the task force's work. That was unfamiliar terrain for the McKinseyites. Crucial medical supplies from surgical masks to ventilators were scarce and the government had solicited offers from any vendors claiming to have access to the necessary supplies. As the offers came in, McKinsey consultants were among the task force members assigned to help vet them before forwarding them to federal procurement officers, the federal official said. Carella, the McKinsey spokesman, said the firm "did not 'vet' offers" of "PPE, ventilators or medical supplies," but rather "helped the client assess the availability of life-saving equipment," without making any decisions about what to pass on to contracting officers.

A FEMA spokeswoman, Janet Montesi, put it differently. The task force "vetted hundreds of leads for PPE that were passed along to FEMA and HHS," she said in a statement. Responding to questions about McKinsey's pro bono work for the task force, Montesi added that "the volunteers played an important role," but career contracting officers followed legally required processes before entering into any contracts.

McKinsey consultants struggled to understand the complicated government procurement rules, according to the federal official. The career procurement officers found themselves rejecting what seemed like every other offer forwarded by the consultants and other task force members, because they ran afoul of various rules. "Even though a career employee can spot the problems quickly, you still have to stop doing your regular procurement work," the official said. "That delays the whole process."

In the weeks just after McKinsey signed its $12 million contract with the VA, Richard Stone, who runs the agency's health care system, and his aides crowed about hiring the consultancy, according to federal officials who spoke with them at the time. Stone, a medical doctor who had worked as a consultant at Booz Allen Hamilton, seemed to lack faith in his own staff. The VA, he told one of the federal officials, was now "better prepared because we have private-sector capability."

His enthusiasm didn't last long. Kramer had insisted that only McKinsey could meet the VA's needs immediately. Yet more than three weeks after the consultants started, they still hadn't provided data and analysis on key parts of the VA health system. One of McKinsey's daily PowerPoint updates for VA officials, dated from mid-April and obtained by ProPublica, shows that the consultants had yet to analyze the pandemic's effect on two types of health care facilities most vulnerable to the coronavirus: VA-run nursing homes and VA facilities in rural parts of the country. The rural facilities alone serve 2.7 million veterans — more than a quarter of the veterans enrolled in the VA — and half of them are over 65 years old. These analyses, notes one of McKinsey's slides, would be added "in the coming days."

The federal government is spending billions of dollars to combat the coronavirus, and spending shows no sign of slowing down. Explore who the U.S. is buying from, what it's buying and how much it's paying.

Also absent was data on medical supply capacity across the VA health system. Slides assessing conditions in numerous regions — including those facing some of the country's worst outbreaks at the time, like New York City, Detroit and New Orleans — contained a placeholder for the missing data: "To be incorporated over the next few days." (Hilton Segel, the McKinsey partner, seemed to blame the VA. "We conducted the analyses as the data was available so the client could make decisions in real time with the best available information," she said in a written statement.)

Meanwhile, VA nurses and other front-line care providers were sounding the alarm about widespread shortfalls in personal protective equipment and other materials. With a complete void in McKinsey's slides where the relevant data should've been, VA officials disputed that there was a supply shortfall, delaying the agency's response. It wasn't until late April that Stone acknowledged in an interview with The Washington Post that medical supplies were at "austerity levels" at some VA hospitals.

Noel, the VA spokeswoman, defended the consultants' work. "McKinsey & Company is fully fulfilling the terms of its contract, providing timely, critical intelligence about capacity and utilization rates of non-VA health care facilities and other analytical services during VA's response to COVID-19. This is a capability that VA does not have, and these services are vital."

At congressional oversight hearings in recent weeks, questions about the sufficiency of the VA's medical supply stores have persisted. Meanwhile, infections have been rising sharply in its facilities. A month ago, there were about 1,685 VA patients and employees with active COVID-19 cases. As of July 14, that number had jumped to 5,887.

McKinsey's success cultivating government clients during the pandemic is, in many respects, the realization of a 70-year mission. The 1950s were when the firm began pushing the view that businessmen should supplant civil servants, particularly in the management positions tasked with putting policy into practice. The pitch was self-interested but well calibrated to appeal during the Cold War: The "free enterprise society" of the U.S. "dictates that industry should be given as extensive a role as possible," McKinsey wrote in a 1960 report to the fledgling National Aeronautics and Space Administration.

NASA soon relied almost exclusively on outside contractors. By 1961, almost $850 million of the agency's $1 billion budget went to aerospace contractors. NASA would become the template for "the emerging 'hollowed-out' structure of the contractor state," the historian Christopher McKenna wrote in "The World's Newest Profession," his 2006 book on the consulting industry.

Over decades, McKinsey's approach became self-reinforcing. As successive administrations chipped away at the civil service, politicians who advocate small government got the dysfunctional bureaucracy they had complained about all along, which helped them justify dismantling it further.

The upshot of this process can be seen throughout McKinsey's coronavirus consulting. Stone, the VA health system head, thought a consulting firm made the VA better prepared. The Defense Health Agency provided "no staff support" for the head of its COVID-19 task force, according to a contracting document, prompting the agency to outsource that work to McKinsey. A senior official in Miami-Dade County had a more jaundiced view. She wrote in an email to a colleague that the firm's consultants were merely "doing the research I am too burned out at this point to do" — adding that she was "quite flattered" that it took an entire team of high-priced consultants to replace her.

McKinsey's business model also generates a second round of revenue from its government work: The firm effectively sells data it obtains from one government project to other agencies. McKinsey generally retains in its central databases anonymized work product from its engagements, so future consulting teams can get a head start on similar projects. Ordinarily, the federal government might be expected to put together that type of clearinghouse and share it with state and city governments free of charge. But in the absence of such a clearinghouse, McKinsey has something state, city and federal government agencies need, and access to government data has formed a core part of McKinsey's COVID-19 pitch.

McKinsey's data was one of the factors cited by VA officials to justify hiring the firm within 24 hours. As a contracting document explained: The firm "already possessed an immense amount of both global and community epidemiological data on COVID-19" the VA didn't otherwise have access to. McKinsey customers pay not only in cash but by adding new data that the firm will be able to sell to the next customer.

Hiring McKinsey is a famously expensive proposition, even when compared with its leading competitors. A single junior consultant — typically a recent college or business school graduate — runs clients $67,500 per week, or $3.5 million annually. For $160,000 per week, you get two consultants, the second one mid-level.

To alleviate the sticker shock, McKinsey has lately offered a coronavirus discount. In project proposals, the firm branded these COVID-19 rates "philanthropic prices." The reduced rates ranged from $125,000 per week (for the two-consultant package) to $178,000 (for five). In a separate column, a McKinsey pricing sheet played up pandemic-only add-ons, whose language read like action-figure packaging: "COVID team includes COVID analytics and best practices."

In her statement, Hilton Segal, the McKinsey partner, noted that, when the pandemic began, the firm reduced its fees to the public sector "as part of our commitment to help. We made our intellectual property and capabilities available widely, including by setting up a COVID Response Center that provides free public access to insights from our research. Thousands of McKinsey colleagues stepped up to help — through client work, pro bono service, developing and publishing insights on the pandemic, and more."

The firm's work for Miami-Dade County suggests some clients get little in return, according to interviews, as well as emails and project documents obtained by ProPublica through a public records request. On a Friday in late April, Jennifer Moon received an email from a senior McKinsey partner. Moon is the budget director and a deputy mayor for Miami-Dade, and her boss, the mayor, wanted to hire McKinsey to help finalize guidelines for reopening the county's economy, which he had shuttered over a month earlier. It fell to Moon to hammer out the details.

In his email, the McKinsey partner, Andre Dua, directed Moon's attention to a pricing sheet listing what he called "our special Covid 19 pricing" and outlined the anticipated scope of the firm's work. The proposal consisted largely of consulting buzzwords: variations on the word "analysis"; offers of "best practices," "perspectives" and "decision-support."

Moon could read between the lines, and she discerned a familiar set of tasks. "Here's what the consultants will be doing," she wrote, forwarding one of Dua's emails to the county lawyer reviewing McKinsey's contract. "Apparently, it takes 5 people with staff support to do what I've been doing myself." The $142,000 per week it would cost was more than the combined annual salaries of the two staffers who had been helping Moon prepare the reopening plan — very ably, she noted in an email.

The reopening plan's publication was imminent, and Moon didn't expect the project to last longer than a month. But as contract negotiations unfolded, Dua and Geoff Bradford, a McKinsey contract manager, resisted the county's attempts to limit the contract's duration to four weeks. They wanted to keep the agreement as open-ended as possible to "provide flexibility." What they meant by that, they explained in a series of emails, was the flexibility to expand the scope of the project and keep McKinsey's consultants around longer.

Dua and Bradford also resisted county officials' efforts to be transparent. McKinsey has a long-standing policy of refusing to reveal the names of its clients and demanding that clients likewise not reveal that they've retained McKinsey, unless they're legally obligated to. Amid the pandemic, the firm has taken confidentiality a step further. McKinsey's COVID-19 contracts still require that clients not disclose that they've hired the consultancy. But many of them now allow McKinsey to unilaterally "disclose that we have been retained by the Client and a general description of the Services." The firm has taken advantage of that clause to market its government work online.

When the officials in Miami-Dade objected to the confidentiality clause, Dua took a firm line: removing it "will be a show stopper on our end." Eventually Bradford allowed that the firm might budge, but only a little. McKinsey would consider letting county officials share its work product "with specific entities," he wrote in a markup of McKinsey's draft contract. But only "if we're able to define those entities and can attach guardrails to such disclosure." McKinsey insisted, in other words, that it should decide what the government could say — and to whom — about the advice it had been given.

As the project progressed, Moon's initial skepticism was borne out. There were non-intuitive and innovative recommendations among the "best practices" for reopening the county. But many of them were obvious or had already been suggested by county staffers: "install Plexiglass barriers between cashier and customer," for example. At times, it seemed as if the consultants were picking best practices at random: a slide on reopening construction sites recommended grouping workers into teams that not only work and eat together but "live" and "travel" together.

Myriam Marquez, communications director of the mayor of Miami-Dade County, didn't respond to a request for comment. McKinsey's statement noted that the firm's "work with Miami-Dade County was focused on sharing insights and observed practices from governments and businesses around the world, as leaders navigated the uncertainty of reopening major sectors of their economy."

In mid-May, Miami-Dade County issued its reopening plan, "The New Normal." The document, which stretched to 175 pages, was widely panned for its needless complexity, which sewed confusion among the public. The plan structured reopening around two overlapping schemes. Five different colored "flags" represented different phases of reopening. Separately, five "archetypes" — a McKinsey innovation, emails and project documents show — grouped industry sectors and public spaces by how much human interaction they required: "can be performed remotely," "lower proximity" and so on. Yet the five flags and the five archetypes didn't align; one flag might cover two archetypes and vice versa. Imagine a stoplight where shapes have been added to the usual three-color scheme — a yellow square means something different from a yellow circle, which means something different from a red circle — and you'll start to get a sense for the confusion the systems provoked. As the Miami Herald put it, echoing the local reaction in Florida-appropriate terms: "They're just like the flags that lifeguards fly on their stands at the beach. Except more confusing."

Xavier Suarez, a longtime county commissioner and former mayor of Miami, didn't see what had been gained by pulling more than half a million dollars out of the county's pandemic-hit budget to hire McKinsey. "It just strikes me as a colossal waste of money," he told me. The firm's "archetypes" certainly hadn't added anything. "I remember reading about archetypes in psychology when we covered Carl Jung," Suarez said. "It sounds like just the right kind of" — and here he paused for effect — "sesquipedalian word a consultant would come up with to try and sound smarter than you."

Carlos Giménez, the mayor of Miami-Dade County and a political foe of Suarez, evidently came to agree. On June 19, his administration quietly posted a revised version of it. Notably scrubbed from the new "New Normal": McKinsey's archetypes. As of mid-July, COVID-19 cases continued to rise in Miami-Dade.