Behind the numbers in the breakthrough infection spike

In the first 18 days of this month, the state documented 2,822 people who were fully vaccinated against COVID-19 and still got the illness.

For all of this month, the number of so-called breakthrough cases will at least triple the July total of 1,612, according to data the state compiled. And July was two-and-a-half times the next highest month, April, which had 654 breakthrough cases.

So we’re seeing a spike in breakthrough cases in Connecticut. Why? Blame the delta variant. Blame complacency by vaccinated people who thought they had some kind of kryptonite shields coursing through their veins.

On the bright side, blame the fact that Connecticut is pushing toward 2.5 million people vaccinated, which should put the few thousand cases in perspective. Even with far fewer eligible people unvaccinated — about 900,000 and dropping — we’re still seeing twice as many cases in unvaccinated people as we see in people who have courageously sat for the painless jabs.

“In the end, the message is still that if you’re unvaccinated, you’re still a lot more at risk than if you’re vaccinated,” said Dr. Lynn Sosa, deputy

state epidemiologist at the state Department of Public Health.

That’s true, with a warning and a caveat to the warning. The warning: Numbers show the vaccine just isn’t as protective as it was in the spring, when we were begging the state for the shot, rather than the state begging us to take it.

Sure enough, on Tuesday the U.S. Centers for Disease Control and Prevention published a report showing the vaccines — mostly the mRNA ones from Pfizer and Moderna — dropped from 91 percent effective before the delta variant became dominant, to 66 percent effective in the grip of delta.

That follows studies in Israel and elsewhere that show the same thing. And it’s roughly what the numbers in Connecticut suggest: In July and August, you were four to five times more likely to catch a case of COVID if you were unvaccinated than if you were.

The caveat: We don’t have the data to compare what’s happening from month to month as all this unfolds. Different coronavirus bugs, different behavior as we returned to parties and crowded bars, different people getting the illness in different places.

“It’s important to see the numbers in context,” Sosa said.

Deaths are even harder to compare over time because as of mid-August, the state documented just 49 among people fully vaccinated, including 18 in August.

And we have different methods of collecting data. Remember, friends, we’ve never done this before, although I studied the monthly charts from the 1918 flu epidemic and they did a nice job of collecting numbers all things considered.

For example, waaaaay back in the first week of August, the state reported a grand total 1,171 so-called breakthrough COVID-19 infections since the start of vaccinations, with just 38 in the six days ending Aug. 3.

That can’t be right, state Sen. Mathew Lesser, DMiddletown, tweeted, after I reported those numbers. He figured he personally knew about eight people with COVID that very week who were fully vaccinated, making the 38 number obviously too small.

Turned out Lesser was right. The state DPH, to its credit, went back and matched up all COVID reports with actual vaccination data and found 6,137 breakthrough cases from February through the middle of last week. That includes the 2,822 cases for the first half of this month.

Lesser, a strong proponent of expanded, statesponsored health coverage, thinks a big part of the breakthrough problem is parents with kids under 12 — who aren’t yet eligible for the vaccinations. Those kids bring home the illness along with harmless scrapes on their knees.

And of course, as always, we don’t know how many people actually have COVID-19 because folks aren’t seeking out tests. So we’ll never know how many breakthrough infections are in our midst, even if Connecticut is far ahead of just about all the other states in testing.

“We just need a much more robust testing regime,” Lesser said.

What are the numbers? As of Friday, data the state

DPH showed me reveal that 149 out of every 100,000 Connecticut residents fully vaccinated contracted a case of COVID in the first half of August — well under the 465 out of 100,000 who were not vaccinated and came down with the illness in that same time.

For July, the state’s numbers show COVID cases in 316 of every 100,000 unvaccinated (if you’re using the old Dewey Decimal system, that’s three-tenths of 1 percent) and 85 of every 100,000 who were fully vaccinated.

So, a great deal for the low, low price you pay for the vaccination. There is no national or state-by-state measure of breakthroughs and if you see one, don’t trust it. Every state does things differently.

Hospital numbers show that in recent weeks 25 percent to 30 percent of people sick enough to be admitted were vaccinated. That suggests that although infections in inoculated people may be less severe, they’re not way less severe.

As for cases, it’s actually a bit better than the state reports in favor of vaccinations. By my calculations, using the actual, reported number of people fully vaccinated and the reported number of cases, the first half of August showed 617 cases for every 100,000 unvaccinated and 128 per 100,000 vaccinated, not 149 per 100,000.

In my method, each month shows a similar difference with the state’s figures. But of course, the message is the same: You’re three or four or five or six times safer with the inoculations. Not ten or 20 times safer, like we thought, but safer nonetheless.

There are lots of ways to look at the numbers and lots we still don’t know. In the coming days and weeks, we hope the public health team of Sosa, Geballe, Gifford & Lamont reveal the following: A breakdown of who’s getting the breakthrough cases by age within each month; and by location within the state. Are there many clusters of breakthroughs, like we saw in Provincetown, Mass.?

How about the number of weeks people were vaccinated before they tested positive? That matters a lot as we look at boosters. And, is there a pattern that shows improperly administered doses?

We build this machine as we fly it, nowhere more than in measuring breakthrough infections.