Thanks for the nice messages about the last newsletter. I’m looking forward to the non-COVID Mondays, COVID Thursday pattern. I’m starting up on Expecting Better revisions so look forward to some pregnancy related posts soon (Skincare? Side sleeping?)
If you are carefully following the news on COVID and child care and schools, you probably have come across the fairly scary statements coming out of Texas about cases in child care centers. CNN reported this week the latest — 441 children and 894 child care workers known to be infected, across 883 centers.
Obviously this is somewhere between jarring and terrifying, and a number of you have asked what I make of it. One part of this question is what does this tell us about what should happen in Texas. I am of the personal view that at this point Texas needs a more significant lockdown on all activities, but that view is driven by many things outside these numbers.
The second thing people are asking, though, is whether this changes how we think about kids and COVID and schools in general. The fact that we see these cases: does this suggest kids are at higher risk than we thought, or child care centers are more significant source of infections? This would have implications far beyond Texas.
To answer this, we need to put the numbers in some context. We need to understand what they really tell us about risk, not just about case counts. Part of this, naturally, is asking what data is missing to paint a complete picture.
By far the biggest missing piece here is the denominator. Any cases are worrisome. But to understand the risks, we need to think about the number of cases relative to the number of kids or staff at risk. Unfortunately, all that is being reported is the number of cases and not the number of kids and staff at child care settings. 441 cases in children means something very different if the total number of children is 1000 versus 10,000 or 100,000 or 1 million.
We can do a little back of the envelope calculation to get some sense. According to the Texas Tribune, in normal times there are about 17,000 child care centers in Texas serving 1.1 million children. At the moment, about 12,000 centers are open (this is about 70% of centers). These are undoubtedly operating at reduced capacity, so let’s imagine that the population is 35% of the typical child population. This is about 385,000 children. The 441 COVID cases represent, then 0.11% or about 1 in 900.
The staff calculation is harder. The required child care staffing ratios in Texas vary from 1:4 for infants up to 1:22 for five year olds. If we take the average of 1:10, this would say that for 385,000 children you’d have 38,500 staff. 894 cases gives a rate of 2.3%.
Again, these calculations are based on some assumptions. It would be great to have actual numbers here, and if your assumptions differ from mine, you can adjust accordingly. But remember that the relevant thing is rates, not counts.
A second piece I think is important to think about is the relative risk of child care centers compared to the overall world. This is especially relevant for staff. Very unfortunately, Texas is in the middle of a major COVID outbreak. According to Johns Hopkins, of the tests being done over the last 7 days, about 14% are positive. Given that testing is limited, this probably overstates the general population risk, but it gives a sense of where things are.
What this means is that if you tested people in Texas — even randomly chosen people — some of them would have COVID-19. When we think about child care settings, we want to think (at least in part) about how rates compare to the wider world. This applies in particular to staff. Some staff in these centers have COVID-19. Is it a larger share than you’d expect in the general population? Is this an especially risky job? If yes, that would be especially concerning.
It is difficult to know this from the Texas data. For context, very helpful reader Ashley Kubiszyn put together the table below, which at least puts the data in context with the total cases in Texas (which are all going up). But I’d love to see more information by profession and demographics.
Clustering and Sources of Infection
Are child care centers sources of infection or just places where some infected people are? This is important to understand. From a parental standpoint, you may not care — if there are some cases in child care centers, maybe you do not want to send your kid even if the centers are not themselves the original source. But from a policy standpoint if we want to, for example, figure out what this says about schools, we really need to know if the child care is fomenting infection.
How could we do that? Well, one way is with good contact tracing. When we see cases in these centers, do they seem to come from family members outside or other sources, or do they seem to be coming from others within the center. Unfortunately, contact tracing isn’t great, so this may be hard.
A less direct way to observe this is to look for clustering. Think about an illness like the stomach flu which every parent knows is spread in child care settings. If you took a typical week and you looked at stomach flu in child care locations, you’d see some locations with no stomach flu, and some where there were a lot of kids with it. In other words, you wouldn’t expect to see a lot of places with one or two kids only having it. Yes, this happens from time to time. But most of the time if one kid pukes, you know it’s going around to the rest.
We could look for similar patterns in COVID. If the disease is spreading in child care settings, we’d expect the cases to be at least somewhat concentrated. What is currently reported suggests cases are fairly widespread (that is, they are across a lot of centers — you can see that in Ashley’s table above) but it would be simple for them to report more detail about the distribution.
In point of fact, I was more worried about this one outbreak in Oregon, where we do see some clustering, or the small number of clusters in North Carolina. But, again, it would be great to have more information about these incidents.
A final, big thing. Texas child care centers were open with very limited rules about precautions. This doesn’t mean that places didn’t take some precautions and I’m sure many of them did. But we do not know the details, or who did what. It would be extremely helpful to understand whether cases are arising more in places with fewer precautions.
The Bottom Line
I am afraid sometimes that I’m coming across as a COVID denier or trying to minimize risks. I’m not. COVID-19 is a serious illness, and not to be taken lightly. Questions about kids, schools, child care — there are no easy answers here and I’m certainly not going to back the President and say all schools should just open normally and try not to think about it.
And in the particular case of Texas the outbreak at this point is so significant I can see a public health case for a more aggressive lockdown, completely independent of child care in particular.
But: I do want to make sure we are making rational choices with a full understanding of the evidence and not exclusively out of fear. The media coverage of facts like these isn’t sufficient to help parents and policy-makers make decisions. The title of this newsletter is ParentData and I feel like the refrain is always “We need more data.” This is boring, but often true. As it is here.
I get a lot of questions, and some of them I’ve already answered. As my husband puts it, though, I’m running a “content farm” these days and it may not always be so clear where to find things.
So a few links…
COVID-Explained: redesigned, now with child care highlighted, updates on treatment and testing.
Keep the thoughts coming. I cannot write back to everyone but I do read all of your emails, I promise.
Where to Find Me