Sunday, August 23, 2020

Covid-19 Testing in Higher Ed

Abstract

No residential college or university operating in-person in the US is doing as much testing as they should and only about 20 are likely to be able to manage outbreaks this fall. Commuter schools may be fine. Testing everyone every 2 days is best but no school is doing that because tests are too scarce and expensive, and it takes too long to get test results. The two types of interesting schools to watch are military academies (some testing, lots of behavior modification and policing) and a handful of schools with in-house capability or special deals (mainly with the Broad Institute). Sufficiently cheap tests may be abundant in a few months. In the meantime, given the scarcity of tests, higher ed should only be using tests as part of efforts to make testing cheaper and faster for everyone.

Higher ed is not doing enough testing to safely reopen

Whether higher ed is doing enough testing is not an interesting scientific question: scientists know that few, if any, higher ed institutions are doing enough testing to prevent outbreaks. Part of the problem is the CDC, which discourages surveillance testing in its guidelines for reopening higher ed institutions, presumably for political reasons.

Covid-19 is hard to manage because of 2 A’s: asymptomatic and aerosol. Many people with Covid-19 are asymptomatic—they don’t show symptoms of disease—but can spread Covid-19, anyway; thus, you need a test to know if someone is contagious. Covid-19 spreads as an aerosol. When you exhale, your lungs release small particles of water and the small particles, aerosols, can hang in the air and spread around. Aerosol spread means that Covid-19 can be transmitted even if people stay 6 ft apart and wear masks. Distancing and masks help enough that we should maintain these practices but they are not so powerful that it’s easy to figure out everyone who a patient could have infected. Therefore, to manage Covid-19 outbreaks, surveillance testing is needed: frequent testing to find people who are infected and don’t know it.

This paper is the big one people point to saying that colleges need to test everyone every 2 days. It’s a modeling study. The authors really say that anything from weekly to daily might be right, depending on how well everyone follows the rules. In principle, if you could keep everyone locked up outside of class and everyone wore masks outside of their dorm rooms, a campus might be fine without any testing; that would require extreme modification of college student behavior. Realistically, testing every 2 days is needed unless an institution can figure out how to get college students this year to act quite differently from any previous class.

Just based on how Covid incubation works, testing every 2 days makes sense because it should catch new infections after they’re detectable but before they’re contagious.

No college or university is testing every 2 days.

You can’t test everyone in a college every 2 days. The normal tests you’ve heard of are PCR tests on NP (nasopharyngeal) swabs. Turnaround time is too long (>2 days) and it is very expensive to buy faster turnaround time. The machines and supplies needed for these tests are scarce. Even assuming normal tests did have good turnaround time, at about $80 each, 1 semester of testing for 1 person is $4,000.

Fast and cheap tests are in development and if we’re very lucky, they’ll be here by the spring semester. Antigen tests are the main contender in this category; some exist but are not yet sufficiently abundant for frequent surveillance testing. This article describes a lot of testing advances in development. We’re not there yet.

About 20 schools might be able to make an in-person fall semester safe

Two interesting sets of institutions to watch are the ones testing twice a week and the military academies.

The military academies are not actually doing a ton of testing but at least they’re doing some surveillance testing, i.e., testing everyone regularly, not just people with symptoms or people who were in contact with positives or presumed positives. The military academies are making up for not testing everyone frequently by being military academies: they know how to teach people to follow orders!

Because testing is so constrained and expensive, colleges need their own labs or special deals to be able to do frequent surveillance tests, which you need if you aren’t dealing with students who are not cadets or middies. A list of institutions doing testing at least twice weekly is below. It’s not a long list. Most schools on the list are getting testing through the Broad Institute, which has a special relationship with Harvard and MIT.

One interesting example is Colby College, a small, liberal-arts college that bet big on testing early on this summer. Colby’s getting their tests through Broad for about $30 each, or $1,500 for a semester. They’re spending $10M this year on health and safety for 2,000 students.

Another interesting example is the University of Illinois Urbana-Champaign, which, last week, did 1% of the total amount of Covid testing that has been done in the United States. They’re using an in-house developed test. The test uses PCR but it uses saliva, rather than NP swabs. Fewer supplies and special chemicals are needed for saliva based tests, compared with NP tests. The materials for these tests cost about $10. Even so, they still require PCR machines, which are scarce.

The other schools doing twice-weekly testing are mostly using Broad or in-house labs.

Schools testing twice-weekly:

  • Harvard University
  • Amherst College
  • Bates College
  • Boston University
  • Bowdoin College
  • Brandeis University
  • Clark University
  • Cleveland State University
  • Colby College
  • Franklin W. Olin College of Engineering
  • Massachusetts Institute of Technology
  • Mount Holyoke
  • Roger Williams University
  • Simmons University
  • University of Illinois at Urbana-Champaign
  • University of New Hampshire
  • University of South Florida
  • University of Wisconsin at Madison
  • Wesleyan University
  • Yale University

This list was pulled from this Google Sheet

Wednesday, August 19, 2020

It won’t be the students' fault.

It won’t be the students' fault.

Students trust college and universities to take care of them. They tend to believe that if college is open, it’s safe. Nerds like me flip through epidemiology modeling papers that say that regular (every 2 days) surveillance testing, with short turnaround times are needed for safe reopening; well-adjusted college students don’t.

I’m pretty bad at empathizing with students. I’m realizing that I have a weird set of values, i.e., I forewent hundreds of thousands of dollars of engineer salary to get a PhD and do a postdoc so that I could get the job I have that I’d rather do than stuff I could get paid more for. My students are in college, not because they love learning, but because they want to get jobs, which is totally fair, because how else would they pay off the debt that pays my salary? While they’re here, the parts of college that they find meaningful are inevitably in the authentic experiences they have with each other rather than in the interactions with me that yield 3 college credits for 3 hours of class for 14 weeks, plus homework. In contrast, the meaningful part of college for me is in the classroom. The difference in our understanding of what college is about often gets us frustrated with each other. For that reason, I’m trying to approach my work not by thinking about what the classroom is like for me but by remembering what college was like for my classmates.

I was a great undergrad and, even then, campus ministry, the dining hall, having friends, having crushes, staying up until 4 am trying to share the Gospel with my roommate—those were the parts of college that mattered most to me. Why should I expect students to be any more studious than me?

Students will be blamed for having parties and drinking. A lot of Covid-19 will, indeed, be spread at parties. When they get back to campus, students are going to most want to see the friends they have missed since March. When a lot of students get together, we call that a party. At parties, alcohol appears but I don’t think alcohol is the point of parties. Most years, faculty get together at the start of the school year but we call that a meeting or happy hour; we're not so different from students. We aren’t going to happy hour this year because our priorities—being nerds, many with kids and spouses and houses—are different from those of people who work in retail or food service, still play sports, and who don’t know where their lives will take them. I’m afraid of getting Covid, mostly because if my spouse got sick, too, we would have a hard time taking care of our toddler: what’s at stake is tangible. Students are less risk-averse than faculty, partly due to how brain development works and partly because the arcs of their lives are not yet clear so the effects of risk for them on their futures are abstract.

It’s unfair to blame students for having the maturity level and priorities of any previous group of students, including my classmates and me.

I think I’m getting better at empathizing with students. I’ll love them as much as I can this fall.