Besides preventing symptoms, do vaccines combat COVID-19’s spread? A UCSC expert thinks so

UC Santa Cruz epidemiologist Marm Kilpatrick, who has gained fame within his field for analyzing the latest COVID-19 research on Twitter, has completed a preliminary calculation which he believes shows significantly reduced risk in virus transmission from people who have been vaccinated.

It’s one of the first scientific efforts to quantify this, and the public health implications at stake are enormous.

Most of the approved COVID-19 vaccines are incredibly effective in preventing severe and even mild symptoms of the infection. But because the trials for Moderna and Pfizer did not include regular testing of the participants, it’s an open question whether they prevent infection. If they don’t, vaccine recipients could become infected and spread the disease.

But last week, Kilpatrick shared (on Twitter, of course) that he’d investigated a clue buried in the Moderna data: Before trial participants received their second dose of vaccine, they were tested for COVID-19. Of the 28,000-plus trial participants, only a tiny fraction (14 in the vaccine group and 38 in the placebo) tested positive without displaying symptoms. Kilpatrick used the results of these tests, along with information he got from the scientists leading the vaccine trial, and other previously published COVID-19 research, to estimate how much the vaccine prevented infection.

The reduction in transmission Kilpatrick calculated for the Moderna vaccine was around 90%. “But the uncertainty around that is (such that) if you ended up getting the gold standard data, and it showed it was 75, or 80%, would I be shocked? No, not at all,” Kilpatrick said. “But if you got that data, and it showed it was like, 30%, then I would say, ‘wow, I got something really wrong.’”

Rebecca Goldin is a professor of mathematics at George Mason University and the director of STATS, a group from Sense about Science that helps journalists and the public correctly communicate statistics. Goldin pointed to several assumptions and sources of uncertainty in Kilpatrick’s calculations: among other things, he had to apply information about the Pfizer vaccine to the Moderna data, estimate the percentage of COVID-19 infections that are truly asymptomatic, and account for uncertainty in each of the various measurements he used to arrive at his answer.

Even if the overall conclusion is true — that vaccinations reduce transmission to some extent — that doesn’t mean the exact numbers in Kilpatrick’s calculations are exactly right, as Kilpatrick himself pointed out. “Even when we’re confident there’s a difference in the two groups, we may have less certainty about how big the difference actually is,” Goldin said.

Joanna Masel, a biologist at the University of Arizona, says that during this pandemic “a lot of things have been held back because people want perfect data,” which isn’t always realistic. When perfect data isn’t available, ”we just have to use everything else we know, and common sense and so on.” Masel agrees with Goldin that no one should set too much store on Kilpatrick’s specific numbers on how much the vaccines reduce transmission.

“This is a snapshot in time,” she said, and pointed out that anyone can still get COVID after getting a vaccine, and that as the Brazilian and South African variants proliferate, “your probability of getting COVID after the vaccine will be much higher.”

The news that the vaccines could prevent transmission doesn’t mean vaccinated people should throw out their masks. “I think vaccinated people can definitely do more,” Masel said.

But no one knows exactly how much more risk they can take on. This is why public health officials are erring on the side of caution, and advising vaccinated people to continue to follow physical distancing guidelines.

Masel is excited about Kilpatrick’s estimates because they will be helpful for projecting the trajectory of the pandemic, and potentially for helping design vaccine distribution.

Understanding how well inoculation prevents transmission is of critical importance not only to how vaccinated individuals should behave, but also to public policy — should vaccine doses be prioritized for those most likely to suffer severe illness, or those most likely to infect lots of people?

Masel called Kilpatrick’s calculations a “scaffold,” that other scientists can now critique, revise, and build upon, which is just what he intended.

The unfortunate truth is, we may never have a precise answer to this question — or at least not any time soon — because the infectiousness data was not collected during most of the vaccine trials.

Kilpatrick’s transmission “tweetstorm” has already been cited by prominent economist and media personality Emily Oster, among others, but the significance of this work isn’t only its content, but how he did it.

“Science has had to operate a little differently during the pandemic. We’ve had to do science at pandemic speed,” Masel said. “We need the data now, which doesn’t mean that we can skip peer review. It means that peer review needs to change during the pandemic — and Marm is an absolute model of how peer review can still work but completely differently.”

Masel said that Kilpatrick “basically conducts peer review on Twitter,” by conducting detailed analyses of preprints (scientific papers that have been finished and shared, but haven’t yet undergone peer review and publication). He even acts as an editor sometimes, contacting other experts like Masel to “review” papers on Twitter where their expertise is relevant.

“He has offered up so many valuable insights, through his Twitter and through private conversations,” University of Florida biostatistician Natale Dean said, in an email to Lookout. “I am grateful for his expertise and the care and attention he brings to his analyses.”

Kilpatrick continues to make waves with his analyses online — already in the past 24 hours he’s tweeted about the relative risks of indoor versus outdoor dining and the potential under-counting of COVID-19 deaths in Africa. He’s now up to nearly 9,700 followers and counting at his handle, @DiseaseEcology.

Here’s the start of the original tweet thread.

Have a question about the pandemic? Ask Lookout: