There’s no crystal ball that can tell us what’s going to happen next with COVID-19, but there are computer models that provide some insight into the direction things are heading. Scientists at Boston Medical Center and Massachusetts General Hospital have developed a COVID-19 outbreak predictor computer model, and it’s showing that infection rates are increasing at an alarming rate in many Massachusetts counties. WGBH News Reporter Craig LeMoult spoke with Benjamin Linas, associate professor at the Boston University School of Medicine and an infectious disease physician at Boston Medical Center, who is one of the computer model’s developers. The transcript below has been edited for clarity.
Craig LeMoult: So what is the outbreak predictor and how does it work?
Benjamain Linas: So the outbreak predictor is trying to answer the question that I think it's a really relevant one right now in Massachusetts — are we in a surge currently? Ironically, it can be difficult to know. As dramatic as a surge is, and we all experienced that in April and May, when a surge is first beginning, you can't actually see it. And so that might be where we are now in the Boston area and even more widely in Massachusetts. We know certainly that cases have been ticking upward recently. What's not clear is if this is the beginning of an upward swing and we're going to be going back into, sort of, one of the peaks like we saw earlier in the spring, or if this actually represents some sort of a slight growth in cases, that's then going to level back out into some sort of a new normal as we're sort of going back to businesses and back to schools. So this outbreak detector really looks at the recent trends, as well as some of the policies in the area, to try to estimate how long will it take for the number of cases in your county to double. That's the doubling time. One thing to distinguish is that it's not trying to be a crystal ball to look a month from now or two months from now to anticipate where the next outbreak might be occurring. Some people have developed tools like that. This tool is really more for the situation we're in today, which is we see that the cases are going up. Are we in a surge? What is our doubling time?
LeMoult: So what is the tool telling you right now?
Linas: So this tool is showing some concerning signs right now. And I think it's agreeing with what a lot of people may have been experiencing or hearing about in the media, that our case counts are going up. And the tool says that our doubling time is accelerating. The rate of expansion is increasing. All around the Boston area, we're starting to see counties turn, on our map, an orange color, which means that the doubling time is somewhere around three to four weeks. Which, I will be clear, our doubling times at the peak of our surge in May were measured in days. And so I want to be clear that we're not in that kind of situation right now. But the evolution of it is such that the counties are starting to turn more and more red. In western Massachusetts, in Hampshire County, that county right now on our predictor is coming up as less than a one-week doubling time, which is very concerning. There's the beginnings of discussion of should we be moving forward in Massachusetts with our reopening plan? Is this perhaps a time to pause? What's the right approach now? I think there's reason to be concerned. There's certainly not reason to be panicking and walking backward. But it might not be the time to be continuing to move forward and to double down on our efforts on masking and distancing and see if we can stop this now before we end up with surges like we had in April and May.
LeMoult: Is this information being used right now to help our planning?
Linas: I hope so. I believe so. We are sharing these data with our colleagues at the Department of Public Health and the COVID planning forces for Massachusetts. And I know they're interested in it and they're definitely integrating into that decision process.
LeMoult: Now, you've aimed the outbreak predictor at seeing the impact of the NFL and NCAA football games, right? What are you looking for there?
Linas: So at the same time that we have these questions about where the epidemic is, it is notable that we are moving forward with and NCAA and NFL football season, which struck us as potentially irresponsible. And so with this kind of tool, it's possible to put into the predictor model things like, does this county have a NFL or NCAA football stadium? Have there been games played there and how many people attended? And if we can predict better when we take into consideration football, then that suggests that football is playing a role in driving the epidemic. We have not yet done that — there's not enough data. On the tool now, you can see, though, there are counties that are concerning. They're in a surge situation and at the same time planning to open up and go forward with their football season. I'm concerned that actually we're right and that this could be a problem. But we'll find out. That's why we do the research.
