By interviewing students and tracking their data, like their PUID swipes and connections to Purdue Wi-Fi, University health officials say they are confident their contact-tracing methods accurately portray that no classroom transmission of COVID-19 occurred this year.
Eric Barker, dean of the College of Pharmacy, plays an administrative role in health monitoring and surveillance, which focuses on COVID-19 contact tracing and case management. He described the contact tracing the team performs as a two-layered process.
The first layer consists of manual contact tracing. Protect Purdue staff interview students who test positive and attempt to identify close contacts and others who may have been exposed to the virus.
This first stage of contact tracing isn’t always reliable, however.
Dr. Libby Richards, a professor in the College of Nursing, said the process is only as accurate as the presence of symptoms and available COVID-19 testing will allow.
Human error is a crucial factor, she said. People may struggle to recall past exposures or be dishonest about certain activities.
“If they were at a party and they don’t say that they were and they don’t share that information,” Richards said, “then the contact tracing won’t be as accurate.”
The second layer of contact tracing tries to mitigate this inherent error by diving deeper into the existing datasets that track students as they go about their day.
Barker said he and the University’s data analytics team set out last April to build tools that draw on existing information about student activity, like their residence hall and room number. Information like class schedules, Purdue Wi-Fi access and PUID swipes could also help contact tracers to identify frequent acquaintances of students who tested positive.
The team built data integration systems that view all the student data and pair students based on certain criteria, like how many times they swiped in at the same dining hall in a week, Barker said. The results illustrate to the team the identities of other potentially positive students.
“When we looked at the classroom data, we didn’t see any correlations or linkages there,” Barker said.
One caveat in processing this data was that the attendance numbers for in-person classes were down across the board during the 2020-21 academic year, as students were provided with alternative ways to attend class.
To control for this, Barker said he asked the data team to analyze instructional laboratory courses for any transmission linkages. Labs tended to be in person and had higher attendance than traditional classes.
The team again didn’t find any transmission linkages. The lack is partially due to precautionary Protect Purdue lab protocols that were in place.
Eric Palmer, a professor in the School of Nursing who teaches a course with a lab component this semester, said he lectures asynchronously but added that in-person clinicals his students must take happen requiring them to be close to each other.
“We follow all the Protect Purdue guidelines,” Palmer said. “The times the nursing students are in a lab setting, when they are close to each other, we make them wear the face shield and the mask and we preach a lot about hand hygiene.”
Palmer said measures like the mask mandate and physical distancing are vital in mitigating the long-term risk of exposure to the virus. He likened the threat of virus exposure to sun damage: the more time you spend out in the sun, the higher the chance you’ll get a sunburn.
Barker’s team traced roughly 89% of cases to places in which people typically take off their mask for prolonged periods, like in domestic setting and restaurants.
Even after the threat of COVID-19 diminishes, Palmer said mask-wearing should stay in vogue.
“When cold and flu season come around, I hope people consider wearing masks,” Palmer said. “It’s the exact same principle (of exposure) that we need to be avoiding.”