Are large indoor events safe? It’s impossible to know, making “personal choice” a false one.

Are large indoor events safe? It’s impossible to know, making “personal choice” a false one.

Jeremy Faust — Read time: 6 minutes

Cover photo

Are large indoor events safe? It’s impossible to know, making “personal choice” a false one.

The math on events like the D.C. Gridiron dinner is variable and complicated, especially without testing or masks.

With the Omicron wave receding, I’m frequently asked about whether large gatherings are safe. Notwithstanding the possibility of a wave of the BA.2 or XE variant, the answer depends on local conditions and individual risks.

There’s a line of punditry that says that since vaccination and our growing armory of effective therapeutics are now widely available, large gatherings (without masks or tests) can resume and that whether to attend such an event is a matter of choice, reflecting any given person’s individual risk tolerance.

The problem I’m having is that “individual risk tolerance” is doing a lot of work here, work it can't do. In some sense, shrugging off all the complexities might be fine in theory. That is, if anyone had a reasonable idea of what their risks truly are. Unfortunately, they do not.

At both extremes of the cautious-to-eff-it spectrum, my sense is that nobody has a good handle on what their individual risks truly are. How could they? As I’ll describe, the math is remarkably complicated. We can’t expect people, even experts, to do this on the daily. That makes the “leave it all up to individuals” approach pretty unworkable at the moment.

The problem I’m having is that “individual risk tolerance” is doing a lot of work here, work it can't do.

Two recent events in Washington, D.C., illustrate how far most people are from grasping their risk: the State of the Union Address (March 1st, 2022) and the Gridiron Dinner (April 3rd, 2022). Both hosted around 650 people for a couple of hours with similar demographics. So they’re interesting to compare.

The day of the State of the Union, four members of Congress were found to have Covid-19 and didn’t attend. The next diagnosed case in Congress did not occur until March 11th. While 7 members of Congress got Covid-19 in the two weeks after the address, 53 attendees have reported infections in the week since the Gridiron dinner.

So why was the Gridiron event more likely to have been a superspreader event than the State of the Union? After all, masks were not required at either event and local case counts were actually higher in early March than April.

The State of the Union required a negative PCR test to attend. The Gridiron event didn’t. This key difference blows up the entire comparison. Comparing two similar events, one in which universal testing was required before and one not, is easy. The Gridiron event was a risk tsunami in comparison to the State of the Union, even given relatively low case counts in the area.

Depending on how you run the numbers, the range of credible risk that a contagious person attended either event is wide. Assuming that every infected person is contagious at all times (not true), and that public data only catches around 25% of new cases (probably true), you could calculate the chance of a contagious person being present was over 90% at either event. But a same-day negative PCR test would decrease that to under 1%. Testing requirements lower the odds dramatically. If 100 similar events were staged without testing, over 90 of them would host a contagious person. If 100 events were staged with testing, maybe 1 of them would include a contagious person.

Now let’s factor in the reality that not all cases are contagious, and that many infected people are symptomatic and isolating at home. Even knowing that we miss perhaps 75% of infections in our public reporting, the odds of a contagious person attending the Gridiron event were, according to my math, likely above 33%. That’s astoundingly high.

But we’re not done. Exposure and infection are not interchangeable, remember? Even in households, the rate of spread is nowhere near 100%, though the rates probably span from 25% to over 75%, depending on the situation. So while the risk of becoming infected after, say, an hour of exposure to a contagious person isn't trivial, it’s actually far lower than most realize. A colleague and I ran the back-of--the-envelop math recently, and we estimated that if you’re exposed to a person with Omicron, the risk of infection might “only” be 5% per hour. The problem is, we are constantly being exposed to this virus, especially as we abandon mitigation.

So why don’t superspreader events occur all the time? The reason is that the exercise above is not really about calculating the risk of a superspreader event. It’s about any spread. Only a small fraction of people are “superspreaders” and likely only for a short period of time. (It’s worth noting that we still do not know what makes a superspreader. Is it something about the person’s illness or is it a matter of timing? Maybe superspreader events happen when a pre-symptomatic person in the 99th percentile of viral shedding happens to enter a crowded space during the 2-hour period in which they are 10-times more contagious than at any other time of their 200-hour contagious window. We don’t know). It’s likely that in large gatherings, a few people do in fact pick up infections routinely these days. Sadly, infections are now so common that we simply no longer notice, unless it’s a high-profile situation and dozens of people are infected.

How can we expect people to behave in ways that reflect their risk thresholds when they don’t have a prayer at accurately staking what the risks are?

Where does this leave me? Right now, given the math, I would not attend a large indoor gathering without masks that did not have a testing requirement; I have a nearly 4-year-old child who can not yet be vaccinated. Knowing that she’ll likely be able to be vaccinated soon (fingers crossed), I’m willing to hold out. I would occasionally attend a large indoor event where I can mask (which rules out banquets, I guess). So I’m still avoiding indoor restaurants, based on the most vulnerable person in my household: my kid. But we have flown with her and will again soon (she’s great about masking). Though that travel poses risk, that’s risk we are willing to abide so she can see her grandparents.

We’ve actually run the numbers and our personal choices are driven by data. But not everyone can do this. This is what I do for a living, and honestly, the math is not easy! (In all transparency, I’m a little irked at my colleagues out there giving advice on this without having done the math, but I’m not here to shame anyone. I’m here to share ideas).

Was the State of the Union safe? Probably. Same-day universal testing actually makes masking a lot less essential (which is something the “cautious caucus” seems to not understand). Was the Gridiron event safe? Probably not; that is, if avoiding infection to keep a vulnerable person at home safe was a priority.

Nevertheless, large events are happening. But until everyone of all ages and risks has adequate protection, I don’t think we can responsibly just leave this to everyone’s perceived personal risk tolerance. How can we expect people to behave in ways that reflect their risk thresholds when they don’t have a prayer at accurately staking what the risks are?

So, yes, I sincerely doubt people can correctly grok the math for their individual scenarios. That’s why until everyone is authorized for vaccination (children ages 6 months to 4 years cannot yet be vaccinated), and until we have better access to prophylactic medications and therapeutics for the most at-risk immune-compromised members of our communities, I favor continued testing and masking for large gatherings whenever that’s feasible.