Editor’s note: This story is available to Tribune subscribers only. Thank you for supporting important local journalism.
I am very happy to report: coronavirus cases are down, and in a big way. Here’s Utah’s cases by day:
We are on the other side of the mountain! Things are significantly less awful than they used to be! It is far too soon to declare victory, but it is honestly very pleasing to see us take positive steps. And we know the decline is true, because we see it from a variety of sources: cases, deaths, hospitalizations, even the amount of coronavirus detected in our sewer systems, it’s all down. Coronavirus in our state hasn’t been at these levels since September.
As you’d expect, the decrease in cases also means a decrease in our transmission rate, or Rt. Remember, Rt is the average number of people infected by each person with the coronavirus, and this is one of the most important measures of the pandemic. If Rt is above 1, even by just a little, cases will rise. If Rt falls below 1, cases will fall.
For months, Utah’s been playing this game with its contagion rate, dancing above and now below the Rt = 1 line. You can compare the last two graphs: When Rt was above 1 for months, cases rose and rose. As it’s fallen below 1, cases have fallen.
So what’s working? Well, the decrease in Rt can only be due to three reasons:
1. People are getting vaccinated — making them immune to the virus. Vaccinations significantly reduce transmission.
2. People have already caught COVID-19 — making them quite immune to the virus (slightly less than if they were vaccinated). The antibodies also significantly reduce transmission.
3. People have modified their behaviors — we’re doing things differently than in the months of rapid COVID-19 acceleration. There could be a bunch of factors at play here: fewer holidays in January and February, less travel, government orders, and so on.
Figuring out the relative role of these three reasons is really important. If the biggest factor is modified behavior, then we’ll have to continue doing whichever behaviors are necessary to keep our Rt below 1. If this decrease in cases is because most people are immune now, then we can be more aggressive in removing coronavirus restrictions.
So which of those factors is playing the biggest role? Let’s find out.
This is the easy one to figure out. At this writing, there have been about 420,000 vaccinated with at least one dose of the vaccine, about 13% of Utah’s population.
That being said, the vaccine doesn’t work immediately. It does seem to work relatively well after one dose: recent estimates put it as high as 90%, but that’s after 21 days. In that first week or two, it’s not very effective yet.
Only about 250,000 Utahns had received at least one dose 21 days ago, which is about 8% of Utah’s population. So at this point, vaccinations have probably reduced the number of people who can catch the virus by about 8%.
Now, that doesn’t necessarily mean an 8% decrease in the average transmission rate. We vaccinated a whole lot of elderly folks, who probably weren’t seeing as many people in their everyday lives as many Utahns. On the other hand, we also vaccinated a whole lot of health care workers, teachers, police officers, and so on, who see more people in their everyday lives than the average.
So far, about half of our vaccinations have gone to people over 60, and half below, according to Utah state data, so maybe those two groups even out. That 8% is just an estimate, but it’s what we have to work with.
People who already had COVID-19
People who have been infected are significantly less likely to be infected again. One study followed people for over seven months and found that the protection that previous infection gives is over 90% — hey, just like the vaccines!
We’re less sure about how long some degree of immunity lasts, but the evidence here is largely positive. Even if the antibody levels don’t last very long, the responses in other parts of the immune system can last for years and years — people with the original SARS-CoV-1 saw an immune response that lasted for over a decade.
Overall, 368,601 Utahns have had a positive test for the coronavirus, or about 11% of the population.
But testing was very sparse early in the pandemic. And at the beginning of the school year, there were all of those remarkably foolish people who refused to get tested. And of course, there are legitimately people who got it but were asymptomatic — and yes, even people who don’t show symptoms largely get immune system protection anyway. Lucky ducks.
So how can we estimate how many people had the virus without getting a positive test? Well, we have a bunch of methods of tracking the pandemic that aren’t as reliant on testing: hospitalizations, deaths, the number of times people Google Search various symptoms, wastewater surveillance, and so forth.
The Centers for Disease Control and Prevention did this through December and estimated that only 1 out of every 4.6 coronavirus infections was actually reported to a health department. That ratio, if it stayed stagnant, would mean 52% of Utahns have been infected with the virus. That’s probably a little high, as testing has continued to improve throughout the pandemic.
A research team at Columbia University built a second model that also found about a 1 in 4 rate nationally. But even better, they provided state-by-state estimates. At the beginning of February, the Columbia model estimated about 1.3 million Utahns have been infected at some point, or about 41% of Utah’s population.
A third model, at COVIDestim.org, uses a different approach to estimate that 29% of Utahns have been infected.
Regardless of the actual number, whether it is closer to 29% or 52%, it well exceeds the number of people who have been vaccinated, and explains the drop in Rt in the past few months to a huge degree — as more and more people caught the virus especially in November and December, there would naturally be downward pressure on the number of people available to catch it again.
How do people spread the coronavirus? By hanging around other people, of course.
So we want to track how much time Utahns spend around other people and how those measures have changed. To do so, we can ask people what they’ve been doing in self-reported surveys, and then verify those answers with real data from their cellphones.
Carnegie Mellon’s Delphi Group have been compiling data from Facebook surveys that do just that, asking a whole bunch of people every day about their activities. People then self-report if they spent time with someone outside their household, went to a store, went to a bar or restaurant, or attended an event that had more than 10 people in the last 24 hours. They’ve been doing this since Sept. 8.
Here are Utah’s results. Click the drop down to change the question people were asked.
You can see the biggest shift in behavior wasn’t actually at the holidays, but came in early November. That’s when the state of Utah issued its emergency order prohibiting people from meeting socially with people outside of their household. This was the one with the scary late night emergency warning on our phones.
The number of people visiting each other outside of their household still hasn’t recovered fully. Even once that order ended, I suspect the seriousness of it taught many about how much spread occurs through out-of-household social contact.
You can see this pattern repeated in the other parts of the survey data. Fewer people are attending large events than in September and October, fewer people are leaving home for work or school, and fewer people are going to stores and restaurants. The upturn really had an impact on our social behaviors, an impact that’s verified by a look at the cellphone tracking data.
But how much did it change? Well, about 46% of Utahns spent time with someone outside of their household in the week before the order, and about 33% did after. This week, we’re at about 40% of Utahns.
Of course, it’s not just whether or not you spend time with someone new, but how many people you spend time with. In the week before the state’s order, about 15% of people said they went to an event with at least 10 people in the last 24 hours; in the week after, that reduced to about 7%. Now, we’re at about 11%. A sick person spending time around dozens of other people means a potential super-spreading event.
Overall, I’d estimate the role of changing behaviors to be the second largest factor of the three: somewhat less than the role of those who already were infected, but more than the impact of the vaccine so far.
You can see this reflected in the Rt graph, too. If immunity was the only thing playing a role in Utah’s transmission rate, you’d expect to see a constant negative slope. But instead, Rt has stayed pretty flat since the middle of January. Well, this data shows us one reason why: people are spending more time with more other people in the last month, after they saw cases were declining.
We’ll need to keep a close eye on it, though. With coronavirus variants starting to spread in Utah — the variants are somewhat more transmissible — our Rt might be a tad more unpredictable moving forward.
The extremely good news is that the cumulative impact of Utah’s behavior changes and the spreading variants to this point hasn’t exceeded the rate of increasing immunity. So far in 2021, we’ve done a really good job of finding that balance. Both residents and the state Health Department will have to continue to work together on that, but so far, so good.
We can’t open everything up right away. But as immunity increases — mostly through vaccinations this time, thank goodness — we can implement even more relaxed coronavirus standards.
So yes, relatively soon, we’ll be able to be able to gather in groups, and yes, we’ll be able to shed the mask. We’re approaching the light at the end of the tunnel.
Andy Larsen is a data columnist. He is also one of The Salt Lake Tribune’s Utah Jazz beat writers. You can reach him at firstname.lastname@example.org.