When Pfizer representatives met with senior U.S. government health officials on July 12, they laid out why they thought booster shots would soon be necessary in the United States. Data from Israel showed the vaccine’s effectiveness waned over time, especially in older and immunocompromised people.
But officials from the Centers for Disease Control and Prevention disagreed, saying their own data showed something quite different, according to four people with direct knowledge of the meeting who requested anonymity to speak candidly.
Other senior health officials in the meeting were stunned. Why hadn’t the CDC looped other government officials on the data? Could the agency share it – at least with the Food and Drug Administration, which was responsible for deciding whether booster shots were necessary? But CDC officials demurred, saying they planned to publish it soon.
That episode, say senior administration officials and outside experts, illustrates the growing frustration with the CDC’s slow and siloed approach to sharing data, which prevented officials across the government from getting real-time information about how the delta variant was bearing down on the United States and behaving with greater ferocity than earlier variants – an information gap they say stymied the response.
CDC officials declined to identify what data they had cited at that meeting with Pfizer officials. But on Wednesday, more than a month later, the agency published a trio of reports showing similar trendlines as the earlier Israeli studies – declining protection against the delta variant as it became dominant in the United States.
“The moment there’s something really problematic, it should be shared,” said Eric Topol, professor of molecular medicine at Scripps Research. “In the time it takes to get out an MMWR report [a weekly scientific digest], too many people have gotten infected, too many people have gotten long covid, too many people among them have gotten very sick, some even getting hospitalized.”
The CDC’s fumbles on the delta variant, following a year when its missteps were often attributed to Trump administration meddling, tell a more complicated story – that the once-storied agency faces other challenges that have hampered an agile response to the pandemic. Critics lament that the most up-to-date data about the delta variant has come from other countries, such as Israel, Great Britain and Singapore. And they say the CDC’s inability to share real-time information led top administration officials, including the president himself, to offer overly rosy assessments of the vaccines’ effectiveness against delta that may have lulled Americans into a false sense of security, even as a more wily and formidable variant was taking hold.
Some inside the agency share those criticisms.
“It’s not acceptable how long it takes for this data to be made available,” said a senior CDC official, who spoke on the condition of anonymity to discuss internal matters. “It’s done in a very academic way. Cross every ‘t,’ and dot every ‘i,’ and unfortunately, we don’t have that luxury in a global pandemic. There’s going to be a need to have a significant cultural shift in the agency.”
There are signs of change: On Wednesday, CDC Director Rochelle Walensky announced plans to develop a new forecasting and outbreak analytics center to analyze data in real time to better predict disease threats. She said it would be the country’s first government-wide forecasting center. The leadership team includes well-respected epidemiologists, among them, Marc Lipsitch from Harvard University and Caitlin Rivers from the Johns Hopkins University Center for Health Security.
Walensky was unavailable for comment on this story. But CDC spokeswoman Kristen Nordlund responded to criticisms of the agency’s slowness to share data, saying that researchers across the country have “worked tirelessly every day throughout this pandemic to gather and analyze data so that we could make the best possible decisions for public health.”
Gathering and analyzing data – especially in a crisis – needs to balance accuracy and speed, Nordlund said. The pandemic has underscored the continued and substantial need for modernizing public health data infrastructure at all levels, she said, and Walensky has put initiatives in place to improve the agency’s ability to provide “relevant, current data” to inform policy decisions.
“Just this week, CDC was able to collect, analyze and publish 10-day old data . . . that resulted in actionable policy decisions to work toward covid-19 booster shots for all Americans,” Nordlund said, referring to the administration’s recommendation Wednesday that millions of vaccinated Americans get booster shots. “This included analyzing 85,593 weekly reports from 14,917 nursing homes that were as recent as Aug. 1, 2021. This kind of analysis can often take a year considering the studies involve tens of thousands of participants.”
– – –
There were clues beginning in May and June that the delta variant might upend the nation’s progress against the coronavirus, even as the CDC told vaccinated people on May 13 they could remove their masks, and President Joe Biden vowed a return to normalcy around July Fourth if enough Americans got the shots.
Reports from other countries battling the delta variant, with similar or higher vaccination rates to the United States – including Singapore, the United Kingdom and Israel – began to show the vaccines appeared less effective at preventing mild to moderate infections, although they were still offering robust protection against severe illness. The Singapore Ministry of Health published data daily from contact tracing investigations beginning in May, which showed that vaccinated people were not only getting infected with the delta variant, but transmitting the virus.
But the CDC was not releasing real-time data. Public confusion began to grow as high-profile sports figures, celebrities and others reported contracting mild cases of the virus despite being fully vaccinated, which increased pressure on the CDC to provide more information.
“They try very hard to get things right and they keep things very closely held until it pops . . . and it’s not as real time as people would like,” said one senior administration official who requested anonymity to speak candidly. “They’re not meeting a need. Why are we dependent on Israeli public health data?”
Nearly seven months into her tenure as the agency’s director, Walensky is among those pressing officials to speed up publication of data and guidance, and to share more frequent updates with the public. As coronavirus cases began to rip through low-vaccination states in June, she asked for data about how the virus was affecting unvaccinated people.
“Preliminary data from a collection of states over the last six months,” she stated at a White House covid-19 briefing on July 1, “suggest 99.5% of deaths from covid-19 in these states have occurred in unvaccinated people.” She also said that unvaccinated Americans accounted for more than 97% of those hospitalized.
Other officials cited those rates in a constant refrain throughout July to encourage vaccinations. In response to a reporter’s question Aug. 5, Walensky acknowledged they were based on data from the first six months of the year. As a result, they almost certainly overstated vaccine effectiveness against the delta variant since the time period included several months when most Americans were unvaccinated and before delta had taken hold in the United States.
Walensky said then the agency was updating the numbers, and she was confident they would continue to show that unvaccinated Americans account for the overwhelming majority of those severely ill and dying of covid-19.
Another data decision provoked more bafflement – the agency’s announcement in May that it would track only the most severe breakthrough infections in vaccinated individuals that result in hospitalizations and deaths. Some critics argued that would make it more difficult to gain visibility into real-time changes in the pandemic.
“It’s quite clear in my opinion that their [decision] backing off monitoring breakthrough infections pre-delta was a mistake,” said Ezekiel Emanuel, a bioethicist at the University of Pennsylvania who was part Biden’s covid-19 transition task force. “I’ve been arguing to do more monitoring, not less, and the CDC has been going in the wrong direction.”
CDC officials and other experts argue that a better approach is studies the agency is conducting that track tens of thousands of people, including health-care and essential workers and residents of long-term care facilities, to see how well vaccines are working. Some people are tested weekly, whether they show symptoms or not, Walensky said at a recent White House briefing.
“I would rather see a smaller population studied very, very well than a large population where you arrive at the wrong answers,” said Natalie Dean, a biostatistician at Emory University’s Rollins School of Public Health, adding that officials need to collect information about participants’ occupations, previous covid infections, vaccination dates and underlying health conditions. That’s much harder to do if all breakthrough infections are tracked because “it’s too many people,” she said.
But even those who support the agency’s approach note that until now, it has issued only a handful of reports from those studies – and none based on data more recent than April, well before the delta variant began making inroads.
“What is very concerning is that we’re not seeing the data come out,” said Tom Frieden, a former CDC director who served under President Barack Obama. “It needs to come out. What you can criticize CDC for, validly, is why aren’t you talking about the studies you’re doing of breakthroughs? Even like, what’s the methodology. Where are they being done? What are the results so far?”
There has been such a long lag time, Frieden added, that some wonder if the CDC is hiding results. “And these are the people who are potentially friendly to CDC, so you know you’re in trouble when even your friends are suspicious of your motives,” he said, adding he was not talking about himself.
The CDC has held data so closely that the public learned of key agency assessments of the delta variant only from an internal agency slide presentation published in The Washington Post on July 29. As part of that presentation, officials said it was time to “acknowledge the war has changed.” One slide described a higher risk among older age groups for hospitalization and death relative to younger people, regardless of vaccination status. Another estimated 35,000 symptomatic infections per week among 162 million vaccinated Americans. The document included references to CDC studies, as well as to international data the agency was examining.
“That exemplifies the problem,” Topol said of the leaked slides. “The moment those data were assembled, that should be shared, and that basically is full transparency and truth telling. People can handle the truth. When you find out there’s a leaked slide set from the CDC appearing from an article in The Washington Post, what do you think that does for the conspiracy theories and the mistrust?”
The day after The Post published the slide deck, the CDC did publish a much-anticipated study about several hundred cases in Provincetown, Mass., stemming from July Fourth celebrations, that found that three-fourths of the people infected with the coronavirus had been vaccinated. Almost all of those cases were mild or moderate. But the study also found that vaccinated people could spread the virus to others – a finding which the agency said was key to its decision earlier that same week to reinstate indoor-masking recommendations for the vaccinated.
The CDC “will tell us just how bad this delta wave was with beautiful science in four months, published in JAMA,” said Scott Gottlieb, former FDA commissioner under President Donald Trump and a Pfizer board member, referring to the prestigious medical journal. “It’s not an indictment of them. It’s the wrong agency. Their mind-set is we should polish it, vet it, peer review it.”
– – –
Critics say the agency’s delay in publishing its findings is particularly troubling in this moment, when the United States is battling a fast-moving variant that is causing nearly 140,000 newly reported cases a day – still well below the nearly 250,000 daily averages during the worst days of the pandemic last January, but more than 10 times higher than the U.S. caseload just a month ago, according to a Post analysis of data reported by states.
“In times like this, it’s important to move data quickly so policymakers can make quick decisions,” according to the senior CDC official. “There isn’t time to wait for this exquisite presentation.”
Experts say the CDC’s insistence on holding onto data until it is thoroughly vetted also means the public is caught off guard about the newest phase of the pandemic.
Both current and former CDC officials say bureaucratic realities hamper the agency’s ability to turn out real-time updates. The agency has nearly 200 separate budget line items, appropriated by Congress. That makes it difficult for even the director to move people and resources around to address urgent concerns. And despite its $8 billion budget, there is scant money for what former CDC director Frieden referred to as crosscutting activities, “someone to think big picture about what is it that we need in terms of covid data.”
The agency has flu experts, respiratory disease experts, virology lab experts, Frieden said, “but you really don’t have, structurally in your organization, groups that think broadly about where we are, what we need to be doing.”
The collection of such data is also easier in many other countries with nationalized health and data systems: Singapore, Israel and the United Kingdom, for instance, routinely collect and analyze data on a population level.
“Our for-profit health-care system is really a patchwork of different systems that don’t talk to each other,” said Edward Belongia, an infectious-disease epidemiologist at Marshfield Clinic Research Institute in Wisconsin. “To do these studies, we have to create these networks where people are sharing data, and that puts us at a relative disadvantage.”
Public health experts say global experience should inform U.S. policy, without dictating it. They fault CDC officials for failing to communicate to the public how data from other countries might be applicable to the United States, or to let Americans understand what they are observing from other countries’ experiences.
“You definitely want to be learning from other countries,” said Richard Besser, former acting director of the CDC and president and chief executive of the Robert Wood Johnson Foundation, the nation’s largest health philanthropy. “The way a virus acts in Singapore will help inform how it will likely act here, and the question then is at what point do you change the recommendation you have for people.”
The rate of breakthrough infections is one of several key questions that must be answered to better respond to the delta surge, said David Fleming, a former CDC official and former global health executive at the Bill and Melinda Gates Foundation. Others include risk factors for severe disease among vaccinated people and how the virus spreads in school settings, especially in younger children. None of those questions, however, is readily answerable through the routine surveillance in place in the United States, he noted.
This kind of insight can’t wait for the “perfect, well-designed large-scale” studies to be published in a top medical journal such as the New England Journal of Medicine, Fleming wrote in an email. The CDC, he added, needs to “step up to this challenge.”
– – –
The Washington Post’s Jacqueline Dupree and Tyler Pager contributed to this report.