How Ohio State researchers made COVID-19 testing on campus more efficient

Bringing students safely back to campus for the start of the 2020-21 school year — a year unlike nearly any in The Ohio State University’s 150-year history — could not have happened without a robust and accurate testing program to identify cases of COVID-19.

But efficiently testing thousands of people each week seemed, at the beginning, like an almost-impossible task. How to move that many people through a testing facility, while keeping them at least 6 feet apart, without causing long lines and backups?

Ohio State got it done by relying on the same researchers who solve similar logistics problems for industry.

“We had a great flow going, but we did realize, that, as we started to scale, there could be additional bottlenecks and issues,” said Christy Bertolo, who is part of Ohio State’s COVID-19 implementation team and who manages the university’s relationship with an external company for testing.

About a week after testing began, Bertolo said, she was awake in the middle of the night, trying to figure out how to make the tests run more smoothly. Bertolo’s non-pandemic job at Ohio State involves working with centers and researchers around campus and connecting them with businesses and industries that might benefit from their help.

Then, she thought of the SIM Center — Ohio State’s Simulation Innovation and Modeling Center. The center creates simulations and models of different logistics issues. Its researchers have built models and simulations for production lines around the world, with the goal of reducing or eliminating logjams and creating or amplifying efficiencies.

In the beginning, Ohio State was testing around 1,000 people per day. But in order to scale up and allow the university to reopen with the best possible testing information, that number needed to increase to around 3,500 to 4,500 people per day. To get that many people through safely and quickly would take serious logistics.

Bertolo sent an email to the center’s director, Shawn Midlam-Mohler, asking if he could help.

Midlam-Mohler thought about it, and realized the process to test 12,000 or more students, faculty and staff each week was not that dissimilar from the factories and production lines he modeled to make them more efficient.

“You’re trying to process a large number of people in a small amount of time in a very controlled way,” Midlam-Mohler said. “It really is the same type of way that you would look at an assembly-line process.”

To simulate the testing site, Midlam-Mohler first visited the site in person. Asymptomatic testing is done at the Jesse Owens North Recreation Center on Neil Avenue. He noted a number of points where things could back up: Students enter the building. They check in. They collect a testing kit. Get directions on how to take the test. Walk to a table, where a worker observes them as they spit into a small tube. Seal the tube. Walk the sample to a drop box. Check out. Leave the building. All while staying 6 feet apart.

Midlam-Mohler said the system was already operating fairly smoothly the day he observed. But he did note a few points where there was the potential for a backup, especially in the check-out area, where students had to stop to drop off a sample and scan their BuckIDs.

He took the numbers Bertolo and the testing team supplied — the number of people coming in per day, the number of testing stations, the average amount of time it took from the time a person entered the building to the time they left — and ran some calculations.

The math behind it is, in the simplest forms, the same types of math that middle- and high-school students study: algebra, statistics and calculus.

“The math behind the scenes gets a little more complex,” Midlam-Mohler said. “Think of it like a conveyor belt: You have a conveyor belt that has a maximum speed, and that’s the person walking. And you’ve got the distance between people on the conveyor belt, and that’s 6 feet, for social distancing. And then you assign values to the individual steps in the process and you see how long they take.”

The simulations showed that Ohio State could test around 20,000 people per day if the university had to. They showed that the system would operate more smoothly if students scanned their IDs when they first arrived at the center, rather than as they left. And it showed that the system Ohio State had in place could, if necessary, be scaled up.

Those tweaks made the testing system more efficient, Bertolo said. As of Monday, Ohio State had conducted more than 100,000 COVID-19 tests. And this week, the university started running diagnostics on those tests on-campus — a shift from the previous two months, when tests were sent to a lab in New Jersey for evaluation.

“I think the biggest thing that I was impressed by was that the system was already working really well — they were already able to process a very large amount of students,” Midlam-Mohler said. “And the biggest thing we were able to do is make sure that when the student testing ramped up, we were ahead of there being delays.”