Your university’s pandemic plan will only work if people follow it.
That simple truth – the need for widespread, voluntary adoption – may be the greatest challenge in any strategy to mitigate coronavirus spread.
The University of Illinois Urbana-Champaign reopening plan rests on two standard practices: Regular COVID-19 tests for every active individual on campus, and fast notification of potential coronavirus exposure through a mobile app developed by the University on its Rokwire open source platform. For reopening to work well, enough of the university’s more than 60,000 students, faculty and staff will have to choose to take regular tests and use the app. They’ll have to do these things as a habit, all semester long.
It’s not enough for the reopening plan to work on paper – it must work for human beings. And humans, we all know, are tricky.
“I keep throwing around the word ‘value’,” said Dr. Sanjay Patel, an Illinois professor of Electrical and Computer Engineering who led the development of the Safer Illinois mobile app’s system of exposure notifications system. Patel ran a tech startup for 10 years – he said prioritizing the user’s experience, like a startup company might, became a necessary guide in his team’s work.
“If you don’t give the user value, they don’t use it,” said Patel. “The value here is, Safer Illinois is a safety net.”
Here are 5 ways the need for voluntary user adoption changed development of Illinois’ reopening plan.
1. You get your status at a glance.
A system of widespread testing and exposure notifications emerged pretty quickly as the university’s reopening plan – but what crystallized the mobile app’s development, Patel said, was the decision to include a “digital status card” within the app.
“You could walk up to a building and show your entry status and be granted or denied entry based on that status,” Patel says.
Over time, that “digital entry card” became two separate statuses in the mobile app: There is a simple “admit/do not admit” status that displays as part of the in-app campus ID. But there’s also a color-coded health status, visible only to the user, that indicates at a glance whether it’s time to get a COVID-19 test or whether their testing is up-to-date and negative.
For this status, the specific colors and their meanings are set by local county health departments. But it means one look is all a user needs to know how certain their health status is and what they need to do to guard community health each day. The system offers users clarity – and clarity is value.
2. Testing is fast and painless.
The conventional COVID-19 test – a nasal swab on a long stick – is infamous for its discomfort. Patients generally don’t find it painful – but the test can activate tear ducts or a gag reflex. It feels strange, and that’s enough to make it something many patients aren’t excited to repeat.
Try asking 60,000 of them to take that test twice a week, all academic year.
Illinois doesn’t. Instead, the campus developed their own test: a saliva-based test that requires the patient to drool into a test tube. It takes less than a minute to administer, and it’s more sensitive than the nasal swab.
3. Testing locations are along your way.
The university placed 18 test locations throughout campus, most in easily identified white tents. And they’re located at key campus entry points: Near bus stops, parking lots and common points of pedestrian access. Tents are also found at campus focal points like the athletic center and the quad.
All that reduces barriers to voluntary testing – if it’s quick, it’s painless, and you’re there already, there’s a lot less reason not to do your part.
4. The app tells you when you’ve been exposed – but it won’t make mountains out of molehills.
The university’s mobile app, Safer Illinois, uses a system of anonymized phone ID numbers to notify its users of potential coronavirus exposure. When your phone is near another phone running the app, it receives an ID from the other phone with Bluetooth. Each phone stores IDs received form other nearby phones over the last 14 days.
Then, when an app user tests positive, a list of codes their phone has generated is uploaded to the cloud, where other users’ phones can cross-check that list of codes with the list of phones they’ve been in contact with.
The upshot: If you’ve been in physical proximity in the last 2 weeks with someone who went on to test positive for COVID-19, and you’re both using the app, your phone will know. (It won’t know who tested positive or how to identify them. It just knows your phone was near a phone on the “positive” ID list.)
What should your phone do with that information now?
The obvious answer is to warn the user every time – but that’s not practical. Not all exposures are significant. Passing someone on a sidewalk who tested positive a week later is one thing – sitting next to someone in class who tested positive that same day is another.
And the software developers quickly realized that the app would need to know the difference.
“We could have the app say, ‘quarantine or test’ every time,” Patel said. “But that would create false positives so often that you’d turn the app off altogether.”
Instead, Patel’s team worked with scientists on the reopening effort to find methods of qualifying possible exposures and deciding which ones were worth notifying users about. The intent is to be cautious enough to catch possible infections early – but not so cautious that the app becomes a nuisance instead of a guardian.
“What we’re trying to do, and it’s still a work in progress, is tune these notifications,” Patel says. His team is teaching the software to use several methodologies as it determines which exposures are a risk, then weigh those methodologies to make a final call about whether a given exposure deserves a “quarantine and test” notification.
Over time, Safer Illinois will even learn from past experiences and weigh the methodologies differently. If one method is found to result in more false positives, it can be weighed less. Notifications should get more reliable as the semester goes on.
5. Using Safer Illinois and getting tested can cut quarantine time in half.
Limiting notifications isn’t the only challenge with a system that tells users about possible exposures. The amount of time they’ll be affected is a factor, too.
The CDC recommends 14 days of self-quarantine after a known contact with someone who’s tested positive for COVID-19. But when university scientists began thinking about test schedules, they found an opportunity: App users could reduce that time.
“It took us a lot of time talking with our colleagues,” said Dr. Rebecca Smith, an assistant professor of pathobiology at Illinois. As an infectious disease epidemiologist, Smith is part of the team working to decide when users should be tested for the best results.
The virus multiplies inside infected people for an incubation period of several days, during which time it’s not yet detectable. So testing too soon after an exposure won’t reveal the virus, even if it’s there.
But 14 days is awfully long to wait.
“The deputy director of the local department of public health is a member of our team,” Smith said. “At first he was saying, ‘you can’t test out of quarantine.’ But we had a group realization: Nobody’s going to use our app if it puts them in quarantine for 14 days every time they hung out on the quad to close to someone who later tested positive.” And that may not be necessary.
The group set out to see if “testing out of quarantine” could happen safely. If that test was negative, the person will be released from self-quarantine.
“We looked at the data,” Smith said. “What do we know about viral dynamics? What can our test detect? How long until we can most likely detect any infection that occurred?”
The group settled on a guideline: After an exposure, users should have two tests at least 24 hours apart. The first test should happen at least 3 days after the exposure, with the second test happening at least 5 days after the exposure.
That means, given two negative tests, a user could confidently end self-quarantine 5 days after exposure instead of 14. That makes the “Safer in Illinois” app more helpful and less hindering.
“We want the app to be protective rather than punitive,” Smith said. “You have to give guidelines people can live with.”
6. Soon, you might track your COVID-19 test the way you track your food deliveries.
Dr. Melvin Fenner is Senior Assistant Director of Medical and Mental Health Systems and Compliance for student affairs technology at Illinois. Part of his job on the reopening project is to integrate existing health technology with new software. To make sure COVID-19 testing and reporting were efficient enough to handle thousands every day, his team tweaked existing systems and developed new software to manage the journey from patient sample to test lab and electronically delivered test result.
Now, he wants to let users track that process for themselves.
“I call it the Grubhub experience,” Fenner said. “You know when your order’s been placed. You know when it’s on its way. You know when it’s been delivered.”
In the next weeks, Fenner wants to see a similar capability within the Safer Illinois app. A patient could visit one of the 18 test tents on campus and leave a saliva sample; then the app would show them when their sample had been picked up by a courier, delivered to the testing lab, and when results were on the way.
Real-time updates like this, Fenner said, would make the user experience better. And that’s likely to increase adoption and compliance in the long term.
“The reality is, this process isn’t going anywhere this academic year,” Fenner said. “If we’ve got to live with it, we want to make it as livable as possible.”
The Safer Illinois app, powered by the Rokwire open source platform, is helping create a smart, healthy community at the University of Illinois Urbana-Champaign.