As a follow-up to my recent COVID post, this weekend I came across a survey database of infection rates at schools in the US. It strongly suggests that schools are not impacting the spread of COVID and that in-person schools are not an elevated risk for teachers.
These data were collected by a company called Qualtrics and the survey responses account for over 200,000 students in 47 states. The survey responses were voluntarily submitted by school districts and therefore do not represent a true random sampling of schools across the country. But even with that limitation, the results are enlightening.
The database looks at the “confirmed infection rate” – which is the number of confirmed COVID infections for students and teachers, respectively, over a two-week period divided by the estimated student or teacher attendance. It also looks at a “confirmed and suspected infection rate” – which adds suspected cases to the numerator.
The database allows one to parse the data depending on the learning mode of in-person at full capacity, a hybrid partial in-person and remote approach, or completely remote. Here are the numbers:
| In-person | Hybrid | Remote | |
| Student Confirmed | 0.17% | 0.14% | No data |
| Student Suspected + Confirmed | 0.47% | 0.86% | No data |
| Teacher Confirmed | 0.24% | 0.25% | 0.32% |
| Teacher Suspected + Confirmed | 0.65% | 0.76% | 0.85% |
There appear to be no material differences between learning models. These data suggest that school has little, if any, effect on COVID rates. If teachers (the population most at risk of people at school) were contracting COVID at schools, the in-person and hybrid teachers should have higher infection rates than those teach remotely. The data do not show that.
The database allows one to select for other variables including the State, whether the school is urban, suburban, or rural, and whether the school is public or private. Changing those variables does not significantly change the infection rates, which further suggests that schools do NOT matter much when it comes to spreading COVID. There is no evidence in the dataset suggesting that remote learning schools are safer in terms of COVID.
Can we rely on this?
Before the stats people jump on me, I concede that the table above is not a substitute for a rigorous statistical analysis. But these data are best I have seen to date on the issue. I am unaware of any data showing the opposite – that schools actually are a nexus for contagion. Hopefully, similar, data driven studies are underway regarding school transmission (or lack thereof) of COVID, including additional detail on the effects of social distancing practices, if any, at schools.
We are a long way from March when schools shut down. The time for making decisions based on fear, politics, or relying on worst-case assumptions of what might happen is over. Administrators, teachers, and parents need to be provided with data and facts to make decisions. If the data in this study continue to be confirmed, I can’t see any real argument for doing anything other than opening schools fully back up.
Of course, my mind is open and if anyone knows of data that show schools-based COVID transmission above background community transmission is real, please let me know.
Vamonos Baby! 