“There’s already discussions about a mask mandate, which is not based upon science.”
— House Minority Leader Kevin McCarthy (R-Calif.), remarks to reporters, July 29
“The ‘game changer’ data the CDC used for the mask mandate is from a single study from India. The study was rejected in peer review. But CDC used it anyway.”
— Rep. Dan Crenshaw (R-Tex.), in a tweet, July 28
The Centers for Disease Control and Prevention on July 27 revised its recommendations on mask-wearing to prevent the spread of the coronavirus. With the emergence of the more contagious delta variant, the CDC shifted course and recommended that even people who are fully vaccinated should wear masks indoors in areas with substantial transmission.
Two months earlier, the agency had said most vaccinated people did not need to wear masks indoors. The theory was that vaccinated people did not spread the disease. But now the CDC says it believes vaccinated individuals who have a breakthrough infection can spread the virus.
Republicans have cried foul, saying the public health agency is relying on flimsy data.
Let’s explore what’s going on here.
When we queried an aide for Stefanik, the third-ranking House Republican, about her claim that the CDC did not rely on science, the aide pointed to the CDC’s apparent reliance on a study in India that was a “preprint,” meaning it had not yet completed peer review and been formally published. He noted that the Science Council defined science methodology as including “verification and testing: critical exposure to scrutiny, peer review and assessment.”
That’s how it is supposed to work. But in a fast-moving pandemic, experts say, public health officials do not always have the luxury of waiting until formal peer review is completed. Lives are at stake and waiting even weeks might mean many more people will die.
“In a public health crisis, the preprint process has been extremely effective in pushing out information quickly that is needed to help form policy,” said Monica Gandhi, professor of medicine and associate division chief for HIV, infectious diseases and global medicine at the University of California at San Francisco. “Although we in the scientific community usually like to wait for peer-review, the covid-19 pandemic is a unique situation and basing policy from preprints and non-peer reviewed studies is likely fair to do as long as the study is reviewed independently by the CDC.”
But Gandhi, a specialist on masks, thought Stefanik was onto something when she questioned whether the information cited by the CDC actually supported urging vaccinated people to wear masks.
Gandhi dismissed the India study (more on that below) and has not been impressed with the follow-up information the CDC has cited as supporting the change in policy. “If the CDC is using this reason to put back masks, they don’t have enough evidence to do so (e.g. that vaccinated are as likely to spread as unvaccinated),” Gandhi wrote in an email. She pointed to other recent studies that cast doubt on the CDC’s analysis, though she is not opposed to the CDC decision.
The CDC’s rollout of the new policy arrived in fits and starts. On July 27, when it announced the new guidance on masks, a CDC document had this section.
Studies from India with vaccines not authorized for use in the United States have noted relatively high viral loads and larger cluster sizes associated with infections with Delta, regardless of vaccination status. These early data suggest that breakthrough Delta infections are transmissible. Unpublished data are consistent with this, and additional data collection and studies are underway to understand the level and duration of transmissibility from Delta vaccine breakthrough infections in the United States and other settings.
The India study found as much as a 50-percent breakthrough of covid-19 among health-care workers at hospitals who had been fully vaccinated but were exposed to the delta variant.
“Although the [CDC] document does cite numerous studies, those citations are meant to support different points than whether a vaccinated person who suffers a rare breakthrough infection can transmit the virus,” said Mark Bednar, an aide to McCarthy. “The only study cited by the CDC to support the hypothesis that a vaccinated individual can transmit the guidance is the India study.”
Justin Discigil, a spokesman for Crenshaw, said the lawmaker’s Twitter thread was developed after he realized that there was only one public study available to support the hypothesis that fully vaccinated individuals can transmit the virus.
“The way this was written would indicate that the printed data alone, before any additional support from the unpublished data, was sufficient to prove that vaccinated individuals are prone to transmitting the virus,” Discigil said. “Once you take out the unpublished data, there is only one study that discusses transmissibility in vaccinated individuals (using viral load, the CDC’s proxy), and that’s the India study.”
Three days after the policy change, the CDC revealed the mysterious unpublished data — that three-quarters of the people infected during a coronavirus outbreak in Provincetown, Mass., were fully vaccinated. But even though 469 people developed covid-19, only four fully vaccinated people were hospitalized. That would suggest the vaccines were helping to minimize the severity of the disease.
Complicating matters is that when people looked up the India study, a notation said that it had been rejected during peer review. That was quickly changed after Crenshaw highlighted that fact in his Twitter thread. Research Square, the preprint site, said in a Twitter thread there had been a technical glitch. “There was an issue relating to how our system communicates with the journal,” Phillip Bogdan, a spokesman for the company, told The Fact Checker. “And as a result, for a brief time, the review status of this paper was incorrectly displayed.”
But Ravindra Gupta, the director of the team at the Cambridge Institute of Therapeutic Immunology and Infectious Diseases that conducted the study, confirmed to The Fact Checker that the article had been initially rejected during peer review because a reviewer “was not happy with certain aspects.” The paper is now on its fourth revision — Research Square only shows the first version — and Gupta said that the “high viral loads” cited by the CDC essentially disappeared in the current version as more information was obtained from a third hospital. He said that revised paper is still under review for publication.
“The key message of the paper is the vaccines do not work as well on the delta variant,” Gupta said in a telephone interview. (Several recent studies, though not in the United States, also have found reduced protection of two-dose mRNA vaccines versus symptomatic delta infections.)
Gupta noted that the study had been posted six weeks before the CDC’s new guidance was issued, meaning it was relatively old. So he said it was likely that it did not have much impact within the CDC until agency officials received the data from Provincetown, involving U.S.-approved vaccines. “Then they thought, ‘This is for real,’ ” he said.
Interestingly, the CDC has sought to distance itself from the India study, even though it was cited in the CDC document and also in a CDC slide presentation obtained by The Washington Post. “The India data were not at all used to make this assessment,” CDC Director Rochelle Walensky told Fox News on July 30.
We sought comment from the CDC. Though we were told a response was being prepared, nothing was received before publication.
Gupta said that he understands that it’s a “horrible thing” to tell vaccinated people they need to mask up again. “Questioning the reversal was correct,” he said. “But the CDC’s guidance was spot on.”
The Pinocchio Test
Stefanik sets a high bar during a pandemic to suggest that all studies used for policymaking should be peer reviewed. McCarthy also cited the fact that the paper was not peer-reviewed. Crenshaw noted that the paper was initially rejected under peer review, which turns out to be correct.
To some extent, it’s a matter of opinion about whether scientific studies that have not yet been approved for publication should inform policymaking. The CDC presumably reviews the data and makes its own assessment, just like any peer reviewer. But the agency invites pointed questions when it unveils a new policy without disclosing rock-solid data to support it.
Given this issue involves a mix of opinion and legitimate disputes about scientific research, we will leave these statements unrated.
Send us facts to check by filling out this form
Sign up for The Fact Checker weekly newsletter
The Fact Checker is a verified signatory to the International Fact-Checking Network code of principles