How is the college tracking in-class transmission?

Date: September 2021


Dear Deans,

We have repeatedly seen stated that there is no evidence for in-class COVID transmission, and yet we know that students who are in class with a COVID-positive person do not count as contacts. How exactly is the college tracking in-class transmission (or lack thereof) in this case?


Concerned about contact tracing


Dear Concerned about Contact Tracing,

While it’s true that Cornell has no evidence of classroom COVID-19 transmission, the lack of evidence does not provide 100% certainty that no one has been ever exposed to COVID-19 or contracted it in a classroom. So far, despite rigorous contact tracing, we have been unable to link any confirmed cases back to the classroom setting.

Contact tracing is done by the Tompkins County Health Department, so if a student tests positive, the health department will attempt to identify people who were in close contact. That is defined as someone who was within 6’ for >10 minutes. The goal of contact tracing is both to attempt to identify where/when the patient was infected and to determine who they have been in close contact with since contracting the virus. Close contacts can include those in classroom settings and contact tracers would attempt to identify those people.

Perhaps your assumption that contact tracing excludes classroom settings is based on the notion that a student who tests positive might not be able to identify all her or his classmates who were sitting within 6’ during the preceding days. That can sometimes be a limitation, just as it might be impossible to identify all contacts in a bar, restaurant, bus, or party. However, there are several tools that Cornell has that are not available in those other settings. First, classrooms are a more controlled environment where nearly all occupants are both vaccinated and wearing masks. The data suggest this reduces transmission risks by about 99.5%. Second, Cornell conducts “adaptive testing” in situations where unexplained positive cases might be linked as a cluster.  Adaptive testing attempts to identify targeted populations for additional testing if there is cause to believe there may be an event where transmission occurred. So far, such transmission appears to be overwhelmingly associated with large unmasked social gatherings.   

Let me close by addressing a final concern that may be on your mind – that Cornell is simply not looking for classroom transmission so we can justify continued in-person instruction. As we’ve seen in some other locations, ignoring COVID-19 does not make it go away. In fact, it makes it much worse. The staff of Cornell Health, the county health department, the modeling teams, and the university leadership have been working nearly around the clock to keep COVID-19 transmission low and understand where the greatest risks occur. I’ve personally been impressed with their commitment and tireless efforts to keep us as safe as possible. It seems to be working as we move past the spike in numbers we saw in the first days of September and are now seeing positivity rates of about 0.45%.


Jason Kahabka
Associate Dean for Administration