Put to the test: detecting COVID-19 on campus

In a previous post, we described the early days of designing Dartmouth’s COVID-19 response. Here we further explore the testing approach we developed, which was critical to Dartmouth’s reopening plan.

“We can’t simply test our way out of this pandemic.”

                                 - Remark to our COVID-19 Task Force from a senior public health official

Before testing was widely available, we introduced an electronic screening portal in May 2020. This online questionnaire asked students and employees who were on campus about symptoms, contact and travel. Any positive responses directed students to contact college health services and employees to contact our occupational medicine service, a third-party vendor hired to conduct testing, and later vaccine administration, on campus.

This questionnaire also asked where on campus the individual was going to be, with the intention of using this information to assist in contact tracing should they test positive. However, like many apps that promised detailed information on contacts using smartphone Bluetooth technology, no contact tracing system proved better than interviewing an individual about where they had been and with whom they spent time.

Initial Testing Protocols

In June 2020, we announced a rigorous plan for routine PCR testing, along with reduced occupancy, for the upcoming fall term. Like many New England colleges and universities, we contracted with the Broad Institute in Cambridge, Massachusetts to perform our PCR testing. We used a courier to shuttle samples once daily, or twice daily on high volume arrival days; results were typically available the next day. Results were also sent to our college health and occupational medicine providers so they could immediately reach out to arrange medical care and isolation for those testing positive and begin to identify contacts.

Determining the correct frequency of testing was challenging. We eventually settled on once weekly. We learned that some schools were testing less frequently (e.g., every other week) and others were testing more (e.g., twice or even three times weekly). Once again, we found ourselves having to set policy based on incomplete information. We consulted weekly with our usual peers and joined other ad hoc check-ins with two newly formed consortia, one of northern New England colleges and another of mostly Boston-based universities. We took every opportunity to compare approaches. At least one school created a model to assess risks associated with re-opening and to inform testing cadence.

Our goal was simple: to identify cases quickly to prevent an outbreak. We knew we needed masks, reduced room capacity to enable distancing and to use only those spaces with optimal air exchange (which in some cases required renovations to achieve that threshold). But clearly, frequent universal testing for everyone on campus was going to be our first line of defense.

Next, we tackled how best to ensure compliance with testing requirements. Enforcement spanned from simple automated warnings whenever someone missed a single test to stricter consequences for more severe violations. We had to consider whether we could — or should — turn off card key access to buildings. What about students who lived on campus and needed to access their dorms? Setting each COVID-19 policy was the first challenge — determining how to best monitor them and set fair consequences for violations posed additional challenges and required constant revisiting.

Preparing for Arrival on Campus

Another hotly debated testing policy was prearrival testing of students returning to campus. We ultimately decided to utilize a saliva-based, video-observed PCR test. Test kits were sent out to students approximately 10 days prior to their arrival. This allowed us to identify infected students before they departed for campus. This also meant that infected students had to alter their plans to isolate at home before travelling.

Prearrival testing wasn’t available to everyone since it required a domestic mailing address. Thus, students living overseas or those in transit (driving cross country or hiking the Appalachian Trail, for example) could not be included. Fortunately, this number was small, less than 5% of returning undergraduate students. A total of 1,635 fall term prearrival tests were completed.

Arrival testing was then performed on days 0, 3, and 7 of students’ return to campus. After day 7, weekly PCR testing began. In January 2021, we increased the testing requirement to twice weekly for all, and then in April 2021, we adjusted the testing frequency again, returning to once weekly for those who were fully vaccinated if they provided documentation.

Throughout the year, we maintained a separate testing site for those who were symptomatic. This included a newly outfitted negative pressure room in the college health services clinic and a drive-up kiosk on campus for employees. More than 59,000 tests were performed during the 2020 fall term; with the increased testing cadence in the 2021 winter and spring terms, this number reached nearly 100,000 each term.

The last major revision to our testing protocol was incorporating rapid antigen tests. In March 2021, we piloted a rapid antigen assay as we prepared to receive students returning to start spring term. The highly specific rapid antigen tests were used in high-risk or symptomatic individuals and always performed alongside a confirmatory PCR test. Due to the delays caused by performing two tests at the testing site, we eventually discontinued the routine use of the antigen tests and further restricted their use to symptomatic testing only.

The final critical feature to our re-opening plan was ensuring adequate isolation and quarantine space for students who tested positive. With a plan to welcome back all graduate and professional students and half of the undergraduates, we secured roughly 500 dorm rooms for isolation and quarantine. We anticipated that this capacity would be sufficient to house students in the event of an outbreak in our student population, regardless of whether they were living on or off campus. Unfortunately, in February 2021, we had an opportunity to test this hypothesis. In a future post, we will discuss how we managed an outbreak on campus.

For further reading

First case on campus:
https://vtdigger.org/2020/03/02/nh-testing-dhmc-employee-for-possible-coronavirus/ 
https://www.vnews.com/Possible-coronavirus-patient-being-tested-at-DHMC-Lebanon-NH-after-trip-to-Italy-33037141

Dartmouth College COVID-19 Task Force members and subgroups:
https://covid.dartmouth.edu/taskforce

Dartmouth College resources on Health and Prevention:
https://covid.dartmouth.edu/health-screening-guidelines

Dartmouth College resources on Vaccines and Testing:
https://covid.dartmouth.edu/testing-health

Dartmouth College Community Conversations about the COVID-19 pandemic:
https://news.dartmouth.edu/community-conversations

NEJM Catalyst case study on developing a surveillance system for COVID on campus:
Mahraj, K., Chaiyachati, K. H., Asch, D. A., Fala, G., Do, D., Lam, D., … Volpp, K. G. (2021). Developing a large-scale covid-19 surveillance system to reopen campuses. NEJM Catalyst, 2(6). doi:10.1056/cat.21.0049