Kids and the Delta Variant: Should you Act Differently?

My most asked question these days, by far, is about kids and the delta variant.

We thought it was the summer of fun! Camp! Popsicles! Followed by the fall of normalcy: back to school shopping, full school days, maybe even no masks. And now…delta.

There is a lot to unpack with the delta variant, including how it interacts with unvaccinated adults, its contribution to global cases, its intersection with the question of whether we should donate vaccine doses, and on and on. I want to mostly focus today on one specific question: How much does this change how we think about kids (if at all)?

Before we get into that, though, I wanted to briefly talk about vaccine efficacy. The vaccines we have remain highly effective against the delta variant. Estimates of efficacy after full vaccination (two doses) with Pfizer in the UK and Canada were 88% and 87%, respectively. Data from Israel has seemed to suggest lower efficacy (64%), but many experts have questioned this given the UK and Canada data. It’s difficult to calculate these numbers in a real-world setting, and the Israel number may simply be too low.

In general, though, the efficacy numbers against any infection are slightly below the wild type. But they are still very effective and extremely, extremely effective against serious illness and death. The presence of the variants is an extra reason to get vaccinated, not a reason to not be vaccinated. Virtually all COVID deaths in the US at this point are among unvaccinated individuals. Bottom line: vaccines work — they still work — and if you are fully vaccinated, you still do not need to worry very much about infection and, especially, serious illness from COVID.

However: kids under 12 are not eligible for vaccines yet. So, what about them?

Kids & Delta

When we turn to the question of kids, we need the right frame. Which is, to me:

What, if anything, does this change?

We have been living with COVID-19 for over a year now. We grappled with questions about how much to engage our kids in the world, how we wanted to mask, what case rates we wanted to respond to. At this point, for better or worse, you have probably made those choices.

When we introduce the delta variant into the equation, then, it would be a huge mistake to undo all this work. I think we are sometimes falling victim to the idea that this should cause us to completely re-evaluate everything. But you did so much work to make these decisions before! Do not throw it away!

Instead, we want to simply ask whether this should change what we do. The value of this frame is that it really narrows what we need to know. There are two reasons that our behavior might change in response. One is if this makes kids more likely to get COVID-19. The second is if it is more likely to lead to serious illness in kids. Let’s deal with these in turn.

Is Infection More Likely?

The Delta variant is more contagious; exactly how much is unclear, but it seems in the range of 60% to 100% more. (100% more contagious = twice as contagious). This means that if an interaction with an infected person had a 10% chance of leading to infection with the original COVID-19, that same interaction has a 16% to 20% chance now.

This means that everyone — kids, adults, etc — are more likely to be infected for a given interaction with an infected person. However: the data does not suggest a relatively greater degree of infectiousness for kids. That is: it doesn’t look like children are relatively more susceptible to this variant.

The best data we have on this is from the UK, where frequent sequencing and the dominance of the delta variant for months has made it possible to look at the age patterns of infection. The graph below shows infections by age group over the last months. Just as in earlier parts of the pandemic, rates in younger children remain extremely low. The most affected group in this time frame are people 16 to 24, a group which in the US has been eligible for vaccination for several months.

The group aged 2 to 11 is perhaps the most relevant here, and the rates are low and flat even though there has been unmasked in person school during this period. This should be reassuring.

It is important to note that over the next months we will continue, in the US, to see children be a larger and larger share of cases. (Not number, share). This is because older people are vaccinated. (For example: UK vaccination rates for older people are very high, and you can see that in their low rates in the graph above). In the limit, if all eligible people were vaccinated, we’d expect to see kids be a really large case share. But that’s not because the variant is more infectious in children.

Bottom line here: infection is more likely for every unvaccinated person, including children, but not relatively more so in kids.

Is infection more serious?

Serious infection or death from COVID-19 in kids is extremely rare. We know this, and it continues to be reinforced with data. Just this week, several studies out of the UK showed extremely low child death rates. Of almost half a million infections, there were 25 deaths, 15 of which were in children with serious underlying illness. Any death is tragic, and death is not the only thing we are worried about, but this reinforces the conclusion that children are extremely low risk.

Turning to delta: there is disagreement about whether the delta variant leads to more severe disease in general. Some people have suggested it does, based on one study out of Scotland. However, this is generally a challenging question because in many places with good monitoring, the most vulnerable people are vaccinated already. The broad expert consensus seems to be that serious illness risks rates are similar.

When it comes to kids, though, the data doesn’t point to anything that would look like alarming increases in hospitalization rates in recent weeks. The graph below shows estimated COVID cases and hospitalizations in the US for two age groups based on CDC data.1

Broadly, both cases and hospitalization rates have been declining in children over 2021 (note the hospitalization rates are on the right axis — they max out at around 1 per 100,000; very low even at their peak). Delta has not been dominant in the US in this period, but the UK data also does not show significant hospitalization spikes over the last few weeks.

The bottom line is there is nothing in the data that we have so far which suggests the variant is more serious for kids. The situation is murky enough that it is hard to rule out the idea that it might be *slightly* more serious (partly, the risks are so low that you’d need a huge amount of data to figure this out). It also might be slightly less serious. But: the fact remains that the risks of serious illness for kids remain really, really low.

(Yes, many people are worried about long COVID. But, again, this is a risk you thought about before. It hasn’t changed, which doesn’t mean you shouldn’t think about it, but does mean that you do not need to re-evaluate based on it).

What Does This Mean?

Should you do anything differently?

I revisited my calculator as a window to what changed. First: the presence of the variants may mean that case rates go up, especially if you are in an area with limited vaccinations. So you want to watch those — independent of the variants, you want to watch them. The NYT Tracker can show you rates in your state or county.

Second: the more contagious virus warrants an update of the transmission risk, which I did in the calculator.

More generally, let’s say you thought through some case rate cutoff for various activities before, under the wild type virus. Like, maybe you thought: I’m comfortable with my child in day care as long as daily case rates are below 60 in 100,000. The presence of the variant should reduce that threshold proportionality, since it’s based on an assumption about transmission risk. If the variant is twice as contagious, your new threshold should be 30 in 100,000.

And if an activity is extremely low risk — like being outside, even unmasked — this shouldn’t change what you do. Twice a close-to-zero risk is still close to zero.

In this way, you can use the decisions you made before. Don’t force yourself to remake them all.

A Final Thought

As I reflect on on our continued discussion around kids, I am recognizing a distinction in some of our thinking. It boils down to this: Do you want to think of COVID-19 for kids the way you think about other illnesses, or does the fact that it is new mean you want to treat it differently? I think at the core this distinction is what colors a lot of our continued disagreements or differences in approach.

On the observables — serious illness, symptomatic illness, mortality risk — COVID-19 for kids (variants or not) is a risk comparable or lower than diseases like RSV or the flu. Again, I know the long COVID fears are real, but the fact is that RSV, the flu and other diseases have these long tails also.

We’ve kept kids isolated to protect high risk adults, but as those groups get access to vaccines and we focus on kids, there is an argument for treating this like other diseases. Which means, in turn, accepting the risk that your kid might get COVID-19.

For some people, the residual newness of COVID-19 means they do not want to think of it this way. They may worry about the unknowns — could there be unexpected impacts of COVID-19 infection a decade from now? — and no amount of current data will help with this. If this is your take, you may want to continue to take precautions which are greater than those you would take in the face of existing diseases.

Both of these may be reasonable, but they do reflect a very different approach. And by recognizing this difference, we may be able to better understand our choices and — more importantly — the choices of others.

Weigh in or Share!

Keep the thoughts coming.

Leave a comment

Share ParentData

Write to Me!


This scales up the CDC hospitalization numbers to reflect that they cover about 10% of the population and it scales them down to reflect findings that at least 40% of children hospitalized with COVID tests are incidental infections. Case rates are converted to counts based on US population by age.