Editor’s Note: David Paltiel is professor of health policy and management at both the Yale School of Public Health and the Yale School of Management. Rochelle Walensky is chief of infectious diseases at Massachusetts General Hospital and professor of medicine at Harvard Medical School. The views expressed in this commentary belong to the authors. View more opinion at CNN.
US colleges and universities face an existential crisis. Caught in the grip of an immediate cash flow problem, most schools believe they must reopen or face financial ruin from dropping enrollments if classes are taught entirely online in the fall. But reopening risks calamity.
According to the Chronicle of Higher Education, 24% of surveyed colleges say at the moment that they will hold classes fully in person or mostly in person. Sixteen percent say that they intend to provide a hybrid model, which will mix in-person classes with ones held online. A number of institutions have announced that they will test all students frequently, but many others plan to rely on random testing or symptoms-based testing – which our research found is not enough to prevent outbreaks.
Our study, published in JAMA Network Open, shows that there is a way to manage the coronavirus and reopen residential campuses safely: All students should undergo a rapid, inexpensive Covid test every two or three days and follow safety precautions like wearing masks and maintaining social distancing. Isolating students who have tested positive is also critically important to preventing campus outbreaks.
Our study mathematically modeled the spread of the virus at a hypothetical college with about 5,000 students. We started with ten students who showed up to campus already infected, and then seeded the student population with small numbers of new cases (say, from an off-campus party) over the course of an 80-day semester. We ran several scenarios based on assumptions about how infectious SARS-CoV-2 was and we watched to see what would happen to infection numbers when we simulated regular testing for all students and isolation for those who tested positive.
We found that by regularly testing all students, not just the ones who had symptoms, universities could prevent outbreaks and keep the campus safe. When we modeled symptom-based testing alone, outbreaks occurred in every single one of the scenarios we investigated.
Our model suggests that this frequency of testing should cost somewhere between $120 and $910 per student for the semester, a high price but one that won’t be prohibitive for many schools. Preliminary studies and pilot programs suggest that saliva-based tests will soon be cheaper, faster, and more accurate than nasal swabs, which will make regular testing more feasible for universities.
It’s important to note that we found the frequency of testing to be the most important factor that a university can control, even more important than the accuracy of the tests. In fact, we found that a cheaper, lower-accuracy test worked better than a more expensive one if the cheaper test was administered more frequently and the turnaround time was fast.
Suppose a university has a hundred infected students. A 70% sensitive test will catch 70 students the first time it’s administered. The next round of tests, if conducted in a two to three-day window, will likely catch another 21 of the remaining 30 cases, identifying a total of 91% of the infections in just a few days.
We were surprised to find that it is possible to test too frequently. Over-testing, even with a test that only produces false positives 2% of the time, ends up filling isolation facilities with uninfected students incorrectly identified as infected which wastes money and resources, makes people anxious, and undermines the testing program’s credibility. To address the possibility of over-testing, schools might begin the semester by testing every two to three days, and then consider dialing back the frequency if the situation warrants.
We weren’t surprised to discover, however, that symptom-based testing alone in a residential college setting would not be sufficient to contain an outbreak. University officials cannot move swiftly enough if they wait until students start exhibiting symptoms – the virus is too readily transmitted by highly infectious, asymptomatic, ‘silent spreaders’ living in close proximity. This is especially a problem if there are sporadic parties or other gatherings that lead to outbreaks. Even now during the summer, with only a few people on campus, a number of colleges have suffered outbreaks among students who have gone to athletic practices or parties. You can’t play catch-up with this virus.
But if students are tested every two to three days, universities would potentially have enough time to isolate the confirmed cases and keep campuses safe. Students and administrators need not be at odds when it comes to staying safe –universities should be instituting a broader program that not only conducts tests, but also provides compassionate, realistic guidance on low-risk options for social connection and human intimacy.
We acknowledge that regularly testing all students sets a very high bar – logistically, financially and behaviorally – and that that might be beyond the capacity of many universities and the students in their care. But in the absence of regular testing, the consequences of an outbreak might be deadly, and members of staff, along with the older, more vulnerable members of the surrounding community, are likely to suffer disproportionately.
Our research suggests that many schools can reopen safely with rapid, less expensive tests administered on a regular basis. Colleges and universities can’t afford to reopen based on random testing or symptom-based monitoring. At the end of the day, given the dangerous risks involved, a school that cannot test all its students regularly or maintain control over good prevention practices must ask itself whether it has any business reopening at all.