May 18, 2020

Institutional Error, the Tech Bro, and Fall Quarter: Q&A with a COVID-19 Expert

Sarah Cobey, an associate professor in the Department of Ecology and Evolution, runs a computational lab at the University of Chicago. Cobey is a leading expert in her field, and she and her research have recently been featured in Scientific American, Nature, and press releases by the Office of the Governor of Illinois (links here). Cobey spoke with The Maroon about her firsthand perspective on the academic and public responses to COVID-19. Her recent COVID-19 population modeling research is publicly available on GitHub.

The Chicago Maroon: I have had the opportunity to look into your background, but I was wondering if you could quickly introduce yourself to our readers?

Sarah Cobey: Everyone calls me an epidemiologist now, and it's fine to label me like that, but that’s not the only work I do. As an associate professor in the Department of Ecology and Evolution, my research focuses on modeling population dynamics, which is much more fundamental than epidemiology. My group’s primary research areas are flu transmission, flu evolution, the development of immunity to flu, and the impact of vaccination upon these processes.

CM: On April 24, the [Chicago] Tribune reported that your COVID-19 growth model predicted a peak in Chicago cases on May 6; in our online correspondence, you indicated that this was not exactly the case.

SC: Ah—common misconception. Models don’t predict points—they predict statistical ranges.

CM: Could you explain the primary implications of your model to a layperson?

SC: Actually, I hesitate to see these models as predictive. We can never be sure of what the future holds, since it is determined by how people respond to policy by their perceptions of risk, executive orders, et cetera. Instead, we use historic data to simulate potential scenarios.

We just released new numbers that forecast mortality rates for the state of Illinois through June 6. Fitted on data through April 30, our models forecast four different scenarios: continued sheltering in place and 20 to 40 to 60 percent increases in transmission rates.

CM: Researchers often arrive at differing predictions, such as the team behind the University of Illinois at Urbana–Champaign (UIUC) model, which had projected a substantially more optimistic COVID-19 model. How does academic collaboration and disagreement play out during these times?

SC: Our biggest collaboration is in trying to get data, which we need to model accurately. For example, although Northwestern [University] researchers had an early data-use agreement [DUA] with the state, UIUC and University of Chicago researchers have waited over a month to get access to the same data. We’re still waiting, although apparently the DUA was just signed. We also collaborate by working through various inconsistencies we find in the data.

As for the disagreement you were referring to—the UIUC forecast has since shifted a lot. It had originally included an optimistic seasonal effect on transmission rates. Without the effect, their current forecast is now closer to our original forecast. But our forecast has since shifted to reflect continued transmission under shelter-in-place, including some surprising spread in rural areas.

CM: Could you comment on how the University of Chicago has been responding to the crisis?

SC: The University response has been slow. Everyone could have been given better and earlier notice about classes not being held, teleworking, and etc. There was no question by the middle-to-end of February that these things would have to happen, yet many people were acting as if things were normal until mid-March. But all this is definitely not unique to the University of Chicago.

CM: Could you shed light onto how you collaborate with public health officials when they reach out to you for your expertise?

SC: Generally, we have approached them, they haven’t approached us. We communicate regularly now with the Illinois Department of Public Health and the governor’s office. We have had almost no interactions with the Chicago Department of Public Health, although I have tried emailing them a couple of times.

CM: Wow. Could you elaborate on your perspective of the public response?

SC: People do ask questions and use our research, but we are not being asked in the way we had expected them to ask. I could characterize our modeling as emergency technical support that is helping to guide the response. But a lot of what we’re showing quantitatively is—I don’t want to call it common sense, but it is something very close to that, for the field. We can do much more nuanced research than this, but without full access to data and closer collaboration, the best we can do to help is to work on these basic models, keep making suggestions here and there about the next steps, and eventually just publish recommendations on our own.

There is definitely a lot of expertise that is not being used. Researchers are asking to be involved in the policy response but feel that decisions are being made without taking advantage of all the information that is available. For example, we have said for many weeks that our quantitative methods can be used to guide strategies for reopening, but we are only asked for rough inputs at the last minute.

CM: How do you respond to people who claim that the crisis can be managed with sufficient data and testing alone?

SC: We unquestionably need good data to manage the response going forward. Mostly, we need to improve surveillance, which is not the same thing as testing a lot of people. Surveillance will tell us where we are and if transmission is increasing or decreasing. Models show that widespread testing alone will not be enough. We will also need some kinds of distancing, mask-wearing, and probably contact tracing. But I don’t think people realize that a major problem is a lack of established surveillance, which is the foundation for figuring out which interventions are working.

I am a little cynical about some data sources that are being pushed way too heavily, at the expense of solid but maybe boring public health surveillance. This is related to larger problems of who gets attention. The disproportionately male and white Tech Bro problem is serious. They often have expertise in some area, yet they assume that it translates to other disciplines with significant bodies of knowledge that other people have been working on for decades. They think that massive amounts of data are enough to solve things if the data can be used to make pretty figures or dashboards or match a curve. But you need to understand the underlying biological processes and the behavior that the system is capable of to identify what data and approaches are needed. They have good intentions, but their efforts often push questionable policy that crowds experts out of the discussion.

CM: In your opinion, how can—or rather, is it possible for—academics to meaningfully engage with individuals and groups like the Wisconsin lockdown protesters, who exhibit a high degree of skepticism?

SC: I agonize about that, but not in a sufficiently scholarly way to have good ideas about what to do. People need to be able to trust the data streams that are being reported and have enough of an understanding about epidemic risk. Outside of large cities especially, people often perceive that their risk is low and petition to reopen. Because they haven’t had as many cases and deaths yet, they assume that they’re at intrinsically lower risk. But it’s only a matter of time before their communities are hit too, if they haven’t been already. Surveillance rears its head here too.

CM: So this is a question forefront on the minds of most, if not all, students—in your opinion, what is the likelihood that the University of Chicago will be returning to on-campus instruction for the fall quarter? What steps, if any, could the University take to safely reopen campus?

SC: Well, phased reopening is already starting at a national level. The governor of Illinois announced a reopening plan yesterday, although there are no clear indicators of reopening before the end of May. Some labs are already open, such as researchers in the Biological Sciences Division working on COVID-19. We’ll learn a lot over the next few months.

I would be surprised if we held classes in the usual way—and we definitely can’t hold large lectures. It really depends what happens over the summer, although I do expect continued risk over the fall. Traditional living arrangements, like the dorms, would have to be modified for it to be safe for people to come back. We need testing. And we need to brainstorm policies to avoid superspreading events that occur when people are interacting and talking a lot in small spaces. Considering its resources, the University of Chicago should be paving the way to see how universities and businesses can reopen.

CM: What are your thoughts on what people can do to fight the pandemic beyond traditional social distancing?

SC: People should help others socially distance and isolate, when necessary. In our society, it’s too hard to stay home when we’re sick. We should help get food and essentials in place so that people can take care of themselves when they’re feeling unwell, and just generally be more accommodating. This is something I’ve thought about for many years with the flu, but I think coronavirus might finally prove the point. If we go out with a mild cough, we should not assume that what we transmit to others will be a mild cough. It could be much worse. It should go without saying that we need paid sick leave, better childcare, and other policies in place to support public health.