A notable example from Friday of a mainstream non-Fox news anchor expressing some polite skepticism about vaccine efficacy. I understand that the vaccines work to prevent hospitalization and death and that’s amazing, says Sara Eisen, but the impression given by public-health experts is that they largely prevent infections too, and, well, do they? Quote: “Are you too casual about the limitations of the vaccine? Because it does feel to me that these breakthroughs are happening, they’re happening regularly, and we haven’t really seen the government pay that much attention to them or warn about them too much.”

Five people in her own family, three of them vaccinated, caught COVID recently. Even larger circles of vaccinated people have experienced outbreaks. She’s not speaking hypothetically.

The most Fauci can do for her is assure her that infections are less common in the vaccinated than the unvaccinated, that boosters further reduce the risk, and that the CDC is doing studies now to try to better ascertain the risk of being infected after vaccination. Okay, says Eisen — but in that case, if the CDC doesn’t know how many breakthrough infections there are and isn’t even tracking “mild” cases, why does their own website casually claim that breakthroughs happen in only a small proportion of the vaccinated population?

How can they know that? Watch, then read on.

"If you look at the people who have died from COVID-19, overwhelmingly 90% of them are unvaccinated," says Dr. Fauci. "If you get vaccinated and get a break through infection, you are much less likely of having a severe outcome and much more likely you would be without symptoms." pic.twitter.com/gEFsIjTNWj

— CNBC's Closing Bell (@CNBCClosingBell) October 1, 2021

I’ve written about this question more than once, most recently a month ago when NYT columnist David Leonhardt guesstimated that the vaccinated have a one in 5,000 chance on any given day of being infected. If we assume that the vaccinated typically experience genuinely mild symptoms when infected; that they’re less likely to confirm their infection with a test because they’re not worried about getting seriously ill; and that they’re more likely than the unvaccinated to wear masks and socially distance in order to avoid exposure in the first place, then it may be that the CDC is missing many of breakthrough infections and/or attributing lower infection rates among the vaxxed entirely to vaccine efficacy when really some of it is due to more risk-averse behavior.

Are breakthrough infections happening more frequently than we were initially led to believe would be the case because Delta is so much more contagious than earlier strains of the virus, generating a thousand times the viral load in the infected? Or are they happening more frequently because waning immunity is more common than we’d hoped? One new study points to the latter:

Six months after receiving the second dose of the two-shot vaccine from Pfizer Inc and BioNTech SE, many recipients no longer have vaccine-induced antibodies that can immediately neutralize worrisome variants of the coronavirus, a new study suggests. Researchers analyzed blood samples from 46 healthy, mostly young or middle-aged adults after receipt of the two doses and again six months after the second dose. “Our study shows vaccination with the Pfizer-BioNtech vaccine induces high levels of neutralizing antibodies against the original vaccine strain, but these levels drop by nearly 10-fold by seven months” after the initial dose, Bali Pulendran of Stanford University and Mehul Suthar of Emory University said by email. In roughly half of all subjects, neutralizing antibodies that can block infection against coronavirus variants such as Delta, Beta, and Mu were undetectable at six months after the second dose, their team reported on Thursday on bioRxiv … ahead of peer review. Neutralizing antibodies are not the immune system’s only defense against the virus. Still, they “are critically important in protecting against SARS-CoV-2 infection,” said Pulendran and Suthar. “These findings suggest that administering a booster dose at around 6 to 7 months following the initial immunization will likely enhance protection against SARS-CoV-2 and its variants.”

Israel began to see its latest ferocious wave of COVID recede only after boosters began rolling out. One health official there who spoke to the Financial Times was blunt in defending that decision. “It’s good that the US and UK are giving boosters to their elderly,” he said. “But it’s a political decision not to follow Israel’s lead, not a medical one. That the vaccine is not good after six-eight months is no longer a secret.”

I don’t know if I would have phrased it that way. “Not good” implies that they’re not effective at all when actually they’re still holding up against severe illness months later. In fact, NIH is out with a model today estimating the number of lives saved by vaccination in the first four months of the national immunization effort. It’s in the six figures:

Based on the model, COVID-19 vaccines saved nearly 140,000 lives in the U.S. through May 2021. About 570,000 people died of COVID-19 in the U.S. through that time. The model estimated that there would have been about 709,000 deaths without the vaccines.

Some states had more effective vaccine rollouts. They were able to vaccinate more people faster. This led to greater protection of their population.

The study found that New York had the greatest reduction in COVID-19 deaths. Researchers estimated that vaccines led to nearly 12 fewer deaths per 10,000 people in the state.

Eisen’s likely right that the CDC is lowballing the number of breakthrough infections but the larger benefits of two doses are clear. Whether a third dose will lead to durable heightened protection against infection or whether we’re in for an endless series of boosters to try to mitigate people’s odds of getting sick is the next data puzzle to be solved.

I’ll leave you with this short clip from another interview Fauci did yesterday, when he was asked about Merck’s impressive new medication to treat COVID. Once that’s available, he’s asked, will people no longer need to be vaccinated? Of course they will, he replies, because Merck’s pill reduces the risk of death by 50 percent. What we want to do is reduce the risk by 100 percent by keeping people from getting infected in the first place. But … that’s Eisen’s point. The vaccines aren’t 100 percent effective at preventing infection and probably aren’t even close, so why casually imply here that they are? The reason to get your shots is because they give you a close to 100 percent chance of not dying from COVID if you are infected. He should focus on that.

Dr. Fauci lauds new Merck antiviral pill that reportedly cuts Covid mortality risk by half, but adds, "you know the way to decrease the risk by 100 percent? Don't get infected in the first place." pic.twitter.com/J1lTZKuv6x

— Aaron Rupar (@atrupar) October 3, 2021