And now, a follow-up to a story we brought you in May about professors at Bard College in Dutchess County, and others, who won a National Science Foundation grant to help develop forecasting models to better capture certain aspects of the COVID-19 pandemic. They’ve now completed modelling COVID-19 spread in small colleges and how safe reopening is. WAMC’s Hudson Valley Bureau Chief Allison Dunne spoke with the primary investigator on the study.
The new model suggests that careful student behavior may be just as important as strong policies to safely reopen small colleges. This comes after a tumultuous week or two of larger universities across the U.S. pivoting to remote instruction due to early infection spikes, plus smaller flare-ups in the Northeast. Matthew Junge is a math professor at Baruch College, formerly with Bard College, and the study’s primary investigator.
“What amount does administrative policy contribute versus student adherence to that policy? And we show that even if you’re in a very strong policy, strictly intervened campus, without the student variable also putting in a good faith and changing their behavior, it’s a hopeless situation,” says Junge. “So student compliance is an essential ingredient to opening a campus is what our results suggest.”
“Certainly in our model with no interventions, we saw, after a semester, over 80 percent of everyone there was infected but, of course, this is with no interventions in place and not completely realistic towards the end that that would be happening. But, of course, that’s not tenable and not recommended to do nothing and no one is doing that,” says Junge. “But then there’s a lot of variation, how much intervention colleges are bringing, and we’re offering fairly concrete guidelines and, I want a disclaimer, this is for small colleges, our modelling was specifically for colleges with 2,000-3,000 students, but we’re recommending that they only bring back half as many students typical, that they test everyone at least weekly, and that — this is from the policy side of things — and that these tests come back fairly quickly, like after one or two days, that there’s not a lot of delay with turnaround on test results.”
“There’s two reasons why we focused on smaller campuses. So, first off, they hadn’t been explicitly studied in much detail before this, and we believe that smaller colleges have a different level of intimacy, if you will. I think you have different social groups, like closer social groups that are more tightly knit, and you tend to have central locations for, like one library, one dining hall that are, again, smaller spaces,” Junge says. “So we think that they are actually a different environment than these larger schools where everything is more spread out but there’s more people in that space. And another big advantage of small colleges is because there are less people and it’s sort of a simpler, smaller space, we can model it with a lot more detail and so that helps us make better predictions.”
“What are the limitations of this kind of model?” asks Dunne.
“The biggest limitation, and this is facing almost every model that’s coming out right now, is we don’t have good data to check our findings against, or even basic parameters for COVID spread are not especially well known right now,” says Junge. “So an example is we still don’t know the efficacy of face masks, for instance, or how infectious COVID is with no intervention present, like how many people, on average, you infect.”
The study was conducted by a team from Baruch College, Grinnell College, Bard College and Cornell University.
“Something surprising that we found in our study is that building closures are not necessarily a good thing, and the advantage of that is you can actually control that space. What is even more concerning to us is what students do when they’re not in one of these buildings that the administrators can control, so in their unstructured time where they can socialize, and obviously they’re not putting two chairs of social distance between each person when they’re hanging out with friends outside of classes,” Junge says. “So what we found is that keeping this buildings open actually may help keep infections down because administrators can control how people move and use those spaces so much better.”
The team released the study as a so-called pre-print, which means it has not yet been peer-reviewed. Because of the time-sensitive nature of the results, Junge says it was important to get out the study now as the first semester will be over by the time the study is peer reviewed.
“Two questions that we thought are a little bit grey with reopening colleges regarding face masks and testing — first off, how frequently tests should be administered and with what latency. This is really just a cost versus benefit analysis for schools because the more frequently and the faster they come back, it costs more money, and they have to decide what is the balance point for this. So we make a recommendation regarding testing and a lot of other papers have done the same thing,” says Junge. “And, in the face mask vein, a place where it’s always a bit ambiguous whether you should wear a face mask or, at least, adherence tends to be quite low is when you’re socializing in private with friends. And we know that a lot of students are going to be inclined not to do that, I think a lot of people just in general are not doing that, and we built into our model that feature where we can have people when they’re with friends socializing, are they wearing a face mask and we can compare it to students not wearing a face mask in that setting It truly does make a very big difference, those two, if you turn that off or on. So a concrete behavioral guideline that we’re offering to students is to wear a face mask when you’re socializing.”
“And another, this is a bit more vague, but we recommend socializing less, so choosing to spend time alone when you might be inclined to socialize,” Junge says. “We realize these are somewhat painful, big asks from students, but a big part of our message is urging them to realize that they’re returning to unusual campuses with all these extreme measures in place, and their behavior is one of these extreme measures, and we strongly suggest that they take that into consideration and enjoy college life a bit differently than before, but at least they get to enjoy college life in some form.”
Junge says researchers plan to build on modelling COVID spread at colleges for the remainder of the grant period.