Of the colleges and universities that have chosen to hold classes in person this fall, most are not conducting widespread testing of their students for the coronavirus, an NPR analysis has found. With only weeks remaining before many of those schools plan to send students home for the end of the semester, the findings raise concerns that communities around the U.S. could be exposed to new outbreaks.
The data from more than 1,400 colleges were compiled by the College Crisis Initiative at Davidson College and analyzed by NPR. They show that more than 2 out of 3 colleges with in-person classes either have no clear testing plan or are testing only students who are at risk — mostly when they feel sick or have had contact with someone who has tested positive for the coronavirus.
Of colleges with in-person classes and more than 5,000 undergraduates, only 25% are conducting mass screening or random "surveillance" testing of students. Only 6% are routinely testing all of their students. Most, instead, are relying on only diagnostic testing of symptomatic students, which many experts say comes too late to control outbreaks and understates the true number of cases.
The data set is the largest national look to date at how colleges have approached testing for the virus, which has varied greatly from campus to campus and from state to state. In lieu of a federal effort to gather data on college reopenings, Chris Marsicano, a professor at Davidson College, stood up a group of "scrappy undergraduates" to create one. "This testing data is clearly data we need. We think of it as our way to help during this crisis," he says.
Experts say that because up to 40% of people with the coronavirus do not show symptoms, frequent and rapid testing of large numbers of students is the best way to fend off major outbreaks on campuses. Research from this summer shows that most cases are spread by people without symptoms, often called silent spreaders.
"You can't play catch-up with this virus," says David Paltiel, a public health expert at Yale University who co-authored a study on the importance of frequent testing. "Any school that thinks it can get away with nothing more than symptomatic monitoring is a fire department responding only to calls once houses have already burned down," he added. "You need to do more."
But the high cost of acquiring thousands or tens of thousands of coronavirus tests puts such comprehensive testing out of reach for many colleges.
The first six weeks of the semester has taught colleges an important lesson: "It's not simply testing — it's testing, testing, testing," says Terry Hartle of the American Council on Education, a national group of college presidents, "but it's an expensive undertaking."
In a letter to congressional leaders late last month, the American Council on Education and other higher education organizations requested at least $120 billion from Congress to help colleges and universities with added coronavirus costs, including testing.
"The amount of money colleges and universities will spend on testing is likely to dwarf every projection we would have made a few months ago," says Hartle. "Schools are just spending substantially more money on testing than they anticipated."
Tests can cost more than $100 each, though some schools have found cheaper options. The Broad Institute, in partnership with the Massachusetts Institute of Technology and Harvard University, is working with more than 100 colleges, including many small private schools in New England, to provide regular coronavirus testing. Through that partnership, tests are $25 each. At the University of Illinois at Urbana-Champaign, where officials are using a saliva test they developed themselves to regularly test students and staff, individual tests are about $10, but given the frequency of testing, officials estimate they're spending about $1 million a week.
"The fact that a majority of schools in the data set are not testing students weekly shows how much institutions are financially struggling," says Samuel Owusu, a senior who helped compile the data under the direction of faculty at Davidson's College Crisis Initiative. He says the major risk is high numbers of undetected cases, and without significant change in the price per test, "this could spell bad consequences for cash-strapped universities anticipating reopening in person in the spring."
The lack of testing on campus is consistent even for colleges in COVID-19 hot spots, NPR's analysis found. About two-thirds of full-time undergraduates who attend a college in a hot spot county are on campuses that do not require routine or surveillance tests.
One big reason for the wide disparities in testing approaches is the lack of clear information from the federal government on how to open colleges during the pandemic. Last week, the Centers for Disease Control and Prevention issued new testing guidance for colleges that laid out the many ways they might test — a change from previous guidance that included testing only students who showed symptoms or had come in contact with someone who tested positive.
The updated guidance tells colleges which people on campus should be tested first if there is an outbreak and gives examples of the different approaches colleges are taking, including widespread testing, but stops short of clearly recommending any specific strategy.
For the fall semester, colleges appear to have leaned heavily on earlier guidance from the CDC and some state governments to justify their more limited testing programs. At Furman University, a small private school in South Carolina, officials consulted with the state health department about testing students weekly but ultimately determined that testing only symptomatic students was best.
When NPR visited the campus at the start of the semester, leaders there said they worried that regular testing would give students a false sense of security. When you test negative, "we think psychologically, you feel safer about your own health and well-being," Ken Peterson, the provost at Furman, told NPR in late August. "So we actually think you're less likely to mask up, you're less likely to distance."
By early September, the school had changed course, announcing it would test all students on campus and those who were returning to school after an outbreak off campus. By late September, a mandatory surveillance program was in place, testing at least 20% of students weekly.
Peterson still believes students take more risks when they are assured by negative test results, though he says he hasn't actually seen evidence of that on campus. "In my view, any such effect appears to be small and offset by our enhanced ability to stay ahead of [the] virus due to surveillance testing," he wrote to me last week.
Furman wasn't alone in its initial thinking. Research from the summer found that this "false sense of security" sentiment, which echoed health guidance from the Trump administration, was reflected on other college websites. According to our database, about a third of colleges with in-person classes are still testing only students who are at risk — mostly when they feel sick or have had contact with someone who has tested positive for the coronavirus. About 40% have no clear testing plan.
At the University of Houston, officials rely on students and staff to self-report if they've tested positive. Its website reads: "UH does not require mandatory testing" and cites CDC guidance to back that up.
"Schools are complying with CDC guidelines," says Yale's Paltiel. "If the CDC has issued schools a free pass, a get-out-of-jail pass that says, 'Go ahead and do nothing,' then what do we expect?" He argues that the CDC guidelines on testing have been "dangerous, disingenuous and not evidence based."
This past summer, CDC guidance also did not recommend testing students as they returned to campus at the beginning of the semester, but the new data show that about 250 colleges opening in person did so anyway, administering pre-arrival or on-arrival testing as the school year began. For about 70 colleges opening in person, that was the only testing they planned for the semester.
Though flawed in some ways on its own — a student could be negative on the day of the test and become positive just a day later — this "entry testing" does provide a baseline for a campus. "I think you really do need to know what your incoming student prevalence is," says Dr. Robert Schooley, a virologist at the University of California, San Diego. "When you understand that, you at least start with some knowledge and can then try to keep it at a low level, if that's what you see when you come in."
The new updated guidance from the CDC now says that entry testing combined with regular testing "might prevent or reduce" coronavirus transmission.
Reopening campuses likely contributed to a large increase in coronavirus cases across the country. Research from the CDC shows a sharp uptick of COVID-19 cases among young adults — up by 55% nationally between early August and early September, when many colleges opened for the semester. Another recent study linked an increase of about 3,000 more cases a day to areas in the U.S. that had colleges with in-person classes.
As colleges have seen outbreaks, many have reconfigured their approaches to testing. At the University of Wisconsin-Madison, after rising cases forced the campus into a two-week lockdown, students living in on-campus housing will now be tested weekly. After a surge in cases at James Madison University in Harrisonburg, Va. — a school that was testing only students who believed they had been exposed or were feeling sick — the university halted in-person classes, sent some students home and hired an outside firm to start a surveillance testing program.
"Testing of asymptomatic individuals can further decrease risk through isolation and contact tracing," the program announcement says. Starting this week, as JMU resumes in-person classes, the university will aim to conduct 300 tests a week. The school typically has about 20,000 undergraduates.
Surveillance testing is helpful because it can indicate how prevalent the virus is in a population of mostly young people.
"It's kind of like going into a dark room: When you suddenly turn the lights on, you can sometimes be surprised by what you see," says Schooley of UC San Diego. He says testing works best when you do a lot of it and when it's mandatory; but it provides fewer answers when it is voluntary or if you're testing only a small number of students.
"It's better than nothing," says Schooley, "but you're still going to have a lot of blind spots and it won't give you the kind of vision that you're going to need to have to really see outbreaks coming."
The University of Mississippi implemented an asymptomatic testing program called Sentinel. The university invited more than 18,000 students to participate, but only 394 students followed through with a test, illustrating the limits of a voluntary program. Other schools, like the University of Georgia, have seen a decline in voluntary asymptomatic testing. On its COVID-19 dashboard, it encourages the campus community to continue getting tested: "In order to keep our community safe we must continue to support the surveillance project."
Of course, testing alone won't prevent the coronavirus from spreading, as demonstrated earlier this semester by spikes in cases at the University of Illinois at Urbana-Champaign, where all students and staff are tested multiple times a week. That program has underscored that wearing masks, social distancing and strictly following quarantine and isolation policies are also essential. After a two-week lockdown to curb the spread, that large state university has seen its daily cases drop significantly.
Experts there say that without widespread frequent testing, campuses are operating mostly in the dark.
"I've looked at what other schools are doing, and there are a lot that haven't instituted the testing we have," says Becky Smith, the lead epidemiologist at the University of Illinois at Urbana-Champaign. "I am worried about it. Colleges are simply seeing the tip of the iceberg. Here, we're seeing the whole iceberg."
Largely because of fears of a second wave in the late fall, many colleges are planning to send students home before Thanksgiving and finish the semester online.
Without widespread testing and a clear understanding of the spread of the virus on campus, public health officials fear this could result in transmitting the virus to families in communities far from campus.
"I am deeply concerned about the fact that Thanksgiving may roll around and we may be sending all sorts of ticking time bombs home," says Paltiel.
In places that do not have aggressive testing programs, Smith recommends putting the campus in a two-week lockdown before students travel home. During the lockdown, students would continue with in-person classes but refrain from any unnecessary outings or gatherings, Smith suggests, so universities can encourage students to actually stay on or near campus, instead of simply leaving campus early because their schoolwork is online.
Researchers and college health officials are also looking toward the spring semester, when many colleges will have to confront these decisions all over again.