Thread by @JenRoesch

This article is widely circulating, what people want to hear and, unfortunately, it is dangerously misleading. A (very long) thread:
1/

https://twitter.com/ProfEmilyOster/status/1314572214621143040

She cites a .13% infection rate in her own study and one of less than 1/2 a percent "even in high-risk areas". However, there is no large-scale, randomized testing of students happening in schools. These #'s are students who voluntarily tested & were reported in schools. 2/

These rates use the total student population as the denominator, not the # of students tested. This is a different measurement than the positivity rate (% of those tested). Without widespread testing, we have no idea how many students are infected. 3/

Because kids are less likely to get sick, we won't detect their cases or if/when they transmitted until/unless someone for whom they were the source of transmission becomes sick. The lag time here is significant. Without backwards contact tracing, we may never know. 4/

However, we do know that kids are testing positive at much higher numbers than she cites. For ex, in NYC, labs reported that 2.3% of children aged 5-17 tested positive 10/2-10/9. The # of new cases in kids per day doubled between Sep & the beginning of Oct. 5/

1/2 these kids are attending school in person, but schools report a positivity rate less than 1/10th of a percent & only 14% of the cases reported by labs. This means that infections in schools are both undercounted & undetected. 6/

She cites FL as an ex of school opening not leading to spikes. However, the article she links to tells a more complete story. Cases among kids were declining rapidly b4 schools reopened. Post-opening, that decline shrunk-except where schools went remote & cases kept declining. 7/

Meanwhile, new cases amongst young adults are rising rapidly again in FL after schools reopened. They had been declining in late July and August. 8/

Evidence on kids & transmission remains limited, but we certainly know that kids *can* transmit and that they can be super spreaders. The NYT just ran a story on a 13-year old girl who infected 11 relatives. 9/

The risks of opening schools cannot be measured solely by positive tests within schools (esp absent mass, randomized testing). There are multiple ways schools increase risk:

-Direct transmission risk in schools (evidence still limited, but we know it's a risk) 10/

-Secondary transmission as a result of becoming infected at school (by kids and adults) - primarily to family, neighbors, friends.

-Tertiary transmission of those people when they go to work or otherwise have social contact. 11/

-Increased transmission risk through increased social contact & behaviors triggered by schools opening: more people back to offices; on transit; more frequent grocery/errand trips & socializing once out to drop off/pickup kids. 12/

Because kids are less likely to get sick, undetected transmission will take time to spread through healthier groups of people before reaching those who are more vulnerable. It's when these people start to get sick & die that we will see the impact of opening schools. 13/

I've focused on how this piece overconfidently minimizes the risks involved in opening schools by focusing on a narrow, misleading & incomplete set of data metrics. However, there's a whole other debate to be had about how best to address inequality & meet kids' needs. 14/

Parents don't need false assurances that schools are "safe enough" or to be told that "no risk" was never realistic. Esp Black & Brown parents in communities that have been hit hardest, whose kids have gotten seriously ill at the highest rates & who are most demanding remote. 15/

Instead, they need paid parental leave, economic protections, devices & internet for their kids, more investment in & support for remote learning & more support & grace generally. 16/

Insistence on opening schools has come at the direct expense of these kinds of investments. Families who fear sending their kids into schools have been abandoned & dismissed. These kinds of articles do no favors for the vulnerable families they purport to speak on behalf of. 17/