Question: I was recently reading that the latest evidence on school reopenings shows that “schools aren’t superspreaders.” What do you think?

Answer: I read that Atlantic piece with interest too! For those of you who haven’t read it, the upshot is that based on preliminary survey data that Brown University economics Emily Oster has been collecting in partnership with Qualtrics online survey platform company, schools (grades K-12) are seeing infection rates generally slightly lower than local community infection rates. This is very heartening news. My major thought (described further herein) is that schools can be sites of superspreader events, but that risk is greatly mitigated based on the measures schools implement to keep their school community safe. When we put our effort and resources to it, we can protect our kids, teachers, and ourselves!

You can access the Qualtrics dashboard here and the most recently findings are further described in this recent Insider article (formerly Business Insider). The most recent data stem from 1,268 schools that have opted into the twice monthly survey. As I understand it, the survey is completed by the school principal, who answers questions in the baseline survey about student enrollment, reopening policy, mitigation measures used and answers questions about new infections and any changes to policy/mitigation measures in bimonthly follow-up surveys. When it comes to preliminary findings, I have the following thoughts:

1. We generally see super-spreader events when we have a COVID-positive individual (oftentimes before they are showing symptoms, but at the height of viral shedding) mixing with a crowd of people indoors with poor ventilation, oftentimes in loud settings, and with limited mask wearing. Schools may OR may not have environments that elevate super-spreading risks. It all depends on: a) the baseline condition of the school environment (ex: overcrowding); b) the mitigation measures schools have implemented (ex: universal mask wearing); and c) level of community spread (ex: the risk that someone in school will come to school infected). In the context of mask-wearing, good ventilation, small cohorts of students working together with their teacher the risk of super-spreading events would be low. And I think this is what the data are showing us. Yay!
2. That said, we do have examples of schools as the site of superspreader events. Here’s a database of superspreader events from around the world from researchers at London School of Hygiene and Tropical Medicine. And yes, schools K-12 do show up on the list. Schools can be sites of superspreader events, but that risk is greatly mitigated based on the measures schools implement to keep their school community (students, teachers, other school professionals) safe.
3. Finally, I’m having trouble finding the Qualtrics survey methods. Based on what I have read, their survey does not appear to be a representative sample survey but instead a convenience sample of schools that have opted in. According to the National Center for Education Statistics, as of 2010, there were >138,000 public and private K-12 schools in the US. With <0.1% of schools participating in the Qualtrics survey and with those schools being self-selected (rather than being randomly chosen based on robust statistical sampling design), I would not put much weight on extrapolating these findings to a set of schools beyond those included in this survey. For example, if the sample is biased (say, more schools that are implementing multiple mitigation measures are participating OR more schools in areas with low community transmission are participating, etc.), then the overarching finding — that schools are not the sites of superspreader events — may not hold true for the broader school population.
4. All that said, the fact that the schools participating have avoided superspreading events is something to cheer. It shows us — like the NBA bubble — that when we put our effort and resources to it, we can protect ourselves.