Nate Silver’s been beating the drum about this study all morning, wanting to know why American media isn’t giving it the same breathless attention that it gave to the questionable Provincetown study on which the CDC’s new mask guidance is based.

The chief virtue of the UK analysis over the U.S. studies, he notes, is that it’s based on a random sample of the population. The P-town and Wisconsin studies relied on by the CDC used self-selecting samples, relying on data taken from infected vaccinated people who sought out testing on their own initiative — probably because they had a bad enough case that they were experiencing symptoms. But not all who’ve been vaxxed and then end up with a breakthrough infection will have symptoms. In fact, it may be that most don’t. We have no idea because, again, the Provincetown sample wasn’t random.

It may be true, in other words, that symptomatic breakthrough infections produce viral loads in vaccinated people that are on par with the viral loads in the unvaccinated, possibly making them equally infectious to others. But that wouldn’t tell us anything about how infectious the average breakthrough infection is, which matters a lot to the CDC’s new universal masking recommendation. If it turns out that most vaxxed Americans aren’t actually as infectious as those in the P-town sample after being infected then why do we all need to mask up?

Why can’t we just ask people with symptoms to stay home?

The key bit from the study:

“Cycle threshold” refers to how many cycles a PCR machine needs to perform on a given viral sample before it detects viral RNA. The fewer cycles needed, presumably the more virus there is in that particular sample. That means lower Ct values should indicate higher viral load, which in turn could indicate greater infectiousness. In the Provincetown study, the Ct values in the vaxxed and unvaxxed were similar. In this study, the values are lower in the unvaxxed than in the vaxxed. Which means that the two groups aren’t equally infectious, assuming that Ct values are a reliable proxy for infectiousness. (A questionable assumption!) The unvaccinated should be more infectious.

The hard numbers:

That’s a big, big sample, Silver points out, vastly bigger than the one in Provincetown that the CDC relied on, which means the margin of error in the data here is much smaller. As you can see, people who received zero doses of the vaccine had consistently lower Ct values than people who’d received one or two, which means they probably had higher viral loads on average. In graph form:

If the UK study is accurate then it’s not necessarily true that the vaccinated are as infectious as the unvaccinated, which was the key conclusion justifying the CDC’s new mask recommendations. At best the data is divided, and due to its methodological superiority the UK study should be seen as more reliable than the Provincetown or Wisconsin ones.

But even so, I’d like two questions answered:

1. Although the vaccinated might have lower viral loads on average than the unvaccinated in the UK data, are those loads still high enough that it’s fair to assume that they’re infectious? The numbers above show that both groups are consistently under 30 in Ct value. In the Wisconsin study published last week, the authors noted that “Previous studies have suggested that Ct values of ~30 or lower are consistent with the recovery of infectious virus in biological specimens, an indication of potential contagiousness.” If that’s true then many of the vaccinated even in the UK study might have been contagious to others, if not quite as contagious as the unvaccinated. Which would mean the CDC’s conclusion that it’s worth having the vaxxed mask up would have some merit to it after all.

2. The table above shows no meaningful difference in Ct values between people who’ve had one dose of the vaccine and people who’ve had two, a confounding result given all the hype about how one dose of the vaccine doesn’t protect you much from Delta. Only after two doses do you have strong immunity. Why are the one-dosers seeing similar Ct values to the two-dosers? The best I can do by way of a theory is behavioral. Because the one-dosers knew they weren’t protected yet from Delta, they hunkered down after their first shot; the two-dosers, knowing they had top-notch immunity, went out and partied and ended up being exposed to more virus on average.

While we’re on the subject of the UK, you’ll be pleased to know that their decline in Delta cases has continued since we last wrote about it, although it’s leveled off lately. Hospitalizations are now in decline too:

Scientists continue to debate whether that’s a durable remission due to growing herd immunity or some quirk due to a conflation of factors, like kids suddenly being out of school, the Euro 2020 tournament being over, and hundreds of thousands of people isolating voluntarily due to exposure to someone who’s infected. (“However, Paget cautions that there is some indication that infections might now be creeping up again.”) The one unfortunate wrinkle is that deaths here will probably rise a bit more steeply than they have there because the UK has a much higher percentage of senior citizens fully vaccinated than we do. Britain didn’t come close to setting a new pandemic-era record for hospitalizations in its latest wave, but some places here have. I’ll leave you with this:

🔥12,408🔥 now hospitalized w/COVID in Florida, setting new record, per updated HHS data.

>1 out of every 1,750 residents in a bed right now, >90% because they didn't get the free, effective vaccines.

— Steven Dennis (@StevenTDennis) August 4, 2021