We are stepping tentatively into summer. Even, I dare say, many of us with unvaccinated children. We, in the US, are enormously fortunate that the success of the vaccine rollout, in most places, has lowered COVID case rates. The result is we are finding way to dip our toe in the both literal and metaphorical water. My kids are going to camp, outside, largely unmasked. It feels odd, but also very good.
This is different from saying that the pandemic is “over”. It’s not over in much of the world, and even in the US we remain worried about later spikes, a fall recurrence, the variants. For fully vaccinated people, the messaging is still consistent: vaccines are protective, get out and do stuff. But many of us with unvaccinated kids are still struggling with what, precisely, is okay to do. There is a facile answer which is, “Well, it depends on your risk tolerance.” This is surely true! But that is paralyzing in its own way. How, precisely, do I know what my risk tolerance is? And when I hear about the variant — should I change what I’m doing right away?
This uncertainty may feel most frustrating when it comes to decisions we make a lot. Indoor playdates with other kids. Camp. Family dinner at a restaurant. And so on. I do not want to re-evaluate these choice every time I make them, but I do want to make them right, and be aware that I might need to alter them in response to the virus.
I think there is a way to do this better, and it draws on some ideas in The Family Firm (spoiler alert!) on how setting up guidelines and rules in your household can make day-to-day decision making faster. The basic idea is to think through — all at once, in a structured way — some “rules for engagement” for different situations, based on the virus situation. Once you’ve done that, you do not have to try to re-evaluate in every moment.
Basically, it’s work to do up front, but once you’ve done it, later decisions are made for you. To be more concrete….
Step 1: Define the Situations
We likely all have a common set of situations that are constantly running through our minds. I have a few listed out below — there are others which probably apply to you (sports events? concerts? family gatherings?).
The goal, then, is to make a decision about what COVID “triggers” (i.e. virus measures) you’ll look to, and at what level, in deciding whether these things are in or out.
Step 2: Choosing the Trigger Variable
There are different triggers people might choose. Most of us are likely to be drawn to triggers which capture, in some way, the rate of COVID in the population. If I’m going to bring my unvaccinated child to interact with 5 people, or with 100 people, how many of them do I expect to have COVID-19?
I think the most obvious number to base this on is local case rates, such as the New York Times reports in their daily tracker here. These rates capture the number of detected daily cases per 100,000 people per day in a given state or county.
The reason case rates are central is that they capture — in one number — a lot of different pieces.
For example (and this is important): they capture variants. There is a lot of discussion of the delta variant, which is more contagious. The greater contagiousness will drive case rates up. So the presence of the variant is captured in the cases; it’s not a separate thing. Case rates also capture how much interaction people are having that spreads COVID. They capture vaccination rates, since vaccination prevents COVID. All in all, case rates capture some version of how much COVID is out there.1
The case rate number is not a direct answer to the bolded question above, but it’s related. We just need a little math. The case rate is quoted in detected cases per 100,000 people per day; in the US right now, the rate on average is 3 in 100,000. I tend to assume that (a) people are infectious for a week and (b) about half of infectious cases are detected. If we apply those assumptions to the case rate of 3 we get:
3 (case rate) X 7 (days in a week) X 2 (half of cases detected) = 42
That’s an estimated 42 active cases per 100,000 people; a share of 0.042%. In other words: you’d expect 1 person with COVID for every 2380 people.
A higher case rate implies more people with COVID-19. Where I live at the moment, the case rate is 2, meaning (by these assumptions) I’d expect one person with COVID for every 3571 people. The state with the highest case rate the moment is Missouri (case rate of 11: estimated 1 case for every 649 people). [If you’re interested: Alter Assumptions in Calculator Here].
Regardless of whether you choose case rates or something else as a trigger, the goal is to map them into activities. In other words: When you think about indoor playdates, the rule should be something like “Yes if [insert trigger] below XX; No if [insert trigger] above XX.” Leading, then, to the obvious follow-up question: what is a good trigger level?
Step 3: Choosing Triggers
If I told you that 1 out of every 10 people you interact with was expected to have COVID-19, you’d probably be very cautious. If I told you it was 1 out of every 100,000 you probably be pretty relaxed. The question is: what number is “small enough” to make you comfortable with a particular activity?
There is no one answer (see above on “it depends on your risk tolerance!”). I think the key is to think about benchmarks based on other exposures, which are a way to better understand how you (implicitly) think about risk.
- In a typical flu season, 8 to 10% of the US population has the flu over a roughly 13 week flu season. This amounts to 0.7% of the population having the flu in a given week of flu season.
- An estimated 10% of children under 5 are seen for outpatient RSV infections each year. RSV is more limited in the summer (although perhaps not this summer), so over a 3 season period this would amount to about 0.25% of children seen per week. This understates infections, measuring only those which result in a doctor visit.
- An estimated 1% to 20% (I know, it’s a big range) of children under 18 have pinworms.
I think the first two are probably the most useful. We have flu vaccines, but they are imperfect, and there is no RSV vaccine. And yet for the most part many of us let our children do normal stuff — go to school or child care and playdates unmasked, go to restaurants, visit indoor bounce parks, etc.
Given this, if you’re thinking about “benchmark” triggers based on these kind of exposures you typically have, the flu number is equivalent to a COVID case rate of roughly 50 per 100,000. The RSV number is equivalent to a case rate of around 17 per 100,000.
These triggers you set may also reflect perceptions of benefits. Perhaps a higher trigger for in-person school than playdates. And they should reflect — if possible — some sense of the risk of the activity. You likely want a different threshold for indoor versus outdoor, and for masks versus no masks.
What’s the output? It’s a table like the above, but with numbers. Filled in with things like “Case Rates < 20 per 100,000” or whatever your triggers are. Just fill in the table and you’re done.
Step 4 through N: Caveats
Of course, not everything is resolved. You cannot necessarily just turn on a dime, and if your cutoff for camp is 30 in 100,000 and one day it is 31, you do not necessarily want to pull your kid that morning. You can think of these perhaps better as “re-evaluation cutoffs”.
And you’ll still need to pay attention to things like variant escape from vaccines (so far, all our vaccines seem very good against variants.)
Even putting the caveats aside, it shouldn’t be lose that actually doing this is hard. It’s a lot of work to make thoughtful choices about this, but the benefit is having made the choices thoughtfully once. Approaching these decisions in this way gives you free brain time back later, generates better thought out decisions, and may lower conflict. Disagreeing in the moment with a partner about today’s playdate is more frustrating than disagreeing (and discussing) in principle in advance.
There are other numbers you could use — the positivity rate (share of tests which come back positive), hospitalization rate, death rate. I talk a bit about positivity versus case rate in this post but, basically, I find positivity rates harder to interpret. Hospitalization and death rates are very lagged indicators
