Contact tracing for COVID-19 will be the most complex health investigation ever

On August 3, 1900, a woman who sold fish along the River Clyde in Glasgow, Scotland, started vomiting and couldn't stop. Within days, she and a granddaughter who shared her tenement were dead. Soon, health officials would confirm their worst fears: “Mrs. B” had contracted Bubonic plague—the first outbreak in the United Kingdom in centuries.

Authorities moved quickly. They reached out to anyone who’d had contact with the deceased. They tracked down a sick child most likely infected by a babysitter who’d attended the grandmother’s wake. They traced a factory worker’s illness to colleagues who’d visited the victims’ family. Health officials isolated at least a hundred people. They disinfected sheets and fumigated apartments. By the time they contained the outbreak, 35 people had died. It could have been far worse.

Thanks to modern medicine, doctors today can perform surgery using robots and lasers. But our best shot at tracking the path of a contagion such as the novel coronavirus SARS-CoV-2 is still the same labor-intensive process used during the reign of Queen Victoria.

The practice, called contact tracing, aims to identify those the virus might strike next. It’s an indispensable part of breaking the chain of a pathogen’s transmission and getting society back on track, according to most health officials including those at the Centers for Disease Control and Prevention. But in the United States, the idiosyncrasies and scale of this pandemic—and the intense polarization of this peculiar moment in American history—promise to make contact tracing the most complex and difficult health investigation in history.

‘It’s going to be hard.’

Around the globe, health workers in many countries already have spent months painstakingly calling people to let them know they may have come in contact with someone carrying the virus. This army of tracers urges those who’ve been exposed to isolate themselves rather than risk sickening others. From South Korea to Singapore contact tracing has slowed, though not stopped, the spread of COVID-19, the disease caused by the virus.

The U.S., despite having more cases than any other nation, is only just getting started. And for tracing to work in freedom-loving America, a lot needs to change—and change now. At the end of April, a bipartisan group of health officials suggested the country would need 180,000 tracers to detect and trail the virus’s movement from now until we find a vaccine. In early May, a National Public Radio survey of the states found that there were about 11,000 tracers.

Yet even as this workforce is ramping up, interviews with nearly two dozen experts suggests political tribalism, misinformation, the demonization of at-risk communities, and the lack of a cohesive federal response threaten to complicate the task. Tracing will only be successful if people trust public health officials, tell the truth, and respond by changing their behavior.

“It’s going to be hard,” said Janet Baseman, an epidemiologist and associate dean at the University of Washington School of Public Health. “The virus is new, and it’s poorly understood, which increases people’s fear. It’s a divisive time in this country, and when we’re in a climate like that, how do you rally people around being part of our response?”

We’re also perilously far behind. The number of new cases appearing each day is swinging wildly, from 13,000 on May 19 to 24,000 the next day, even though much of the country remains locked down. And those are just the cases we know about. Until the U.S. does a far better job of quickly figuring out who is infected, it simply won’t be possible to keep up with tracing.

Right now, the country tests roughly 300,000 people daily. That’s a fraction of what health experts have said we’ll ultimately need; estimates include 900,000, 3 million, and 20 million. Many of those tests require days-long waits for results—delays that let the virus extend its reach. And the Trump Administration has insisted it is unlikely to get beyond more than 500,000 tests.

Marc Lipsitch, director of the Center for Communicable Disease Dynamics at Harvard University's School of Public Health, is pessimistic. Even if 80 percent of the country managed to trace cases effectively—“which would be miraculous,” he said—the remaining 20 percent would be a constant source of new infections. New research, not yet peer reviewed, suggests tracing would have to catch at least half of all new cases to reduce transmission even 10 percent. “My view is it’s probably not going to work,” Lipsitch said. “But given the lack of alternatives, we have to try—and hopefully demonstrate that I’m wrong.”

His colleague, Ashish Jha with Harvard’s Global Health Institute, however, believes these problems are surmountable. To stand a chance, though, this campaign will have to move fast.

Tracing techniques have been honed over decades

Contact tracing is often characterized as sleuthing, but the job is more akin to social work. To trace contacts, investigators ask ill people or their next of kin about their whereabouts back to a few days before they showed symptoms, seeking anyone the victim had come within six feet of for at least 15 minutes. Those contacts are then passed to a tracer who reaches out by phone.

This second round of conversations follows a vague script. Callers explain to people that they may have been exposed. They answer questions and ask about symptoms and access to care. They ask the exposed people if they have food and a bathroom. They urge them to isolate at home and call their doctors. Callers ask these people who they, in turn, might unwittingly have infected.

Krysta Cass has made hundreds of such calls in the last two months as a tracer in Boston. “There’s a point in time where you want to pause and let them really digest what you said,” she explained. “You have to wait and listen, and then be a friend.” Her first task is to put people at ease.

“The human component can’t be underestimated,” said Nahid Bhadelia, medical director at the special pathogen unit at Boston Medical Center and an associate professor of infectious diseases at Boston University’s School of Medicine. “You’re basically asking people to reveal parts of their lives to you.”

This process can seem intrusive, but public health agencies have honed it over decades. They use tracing routinely for measles and food-poisoning outbreaks or to track HIV/AIDS and avian flu. It was central to stamping out Ebola in West Africa after the 2014 outbreak. “We get people to tell us who their public contacts are when they have syphilis,” said Kristen Pogreba-Brown, an assistant professor of epidemiology at the University of Arizona. “We know how to do this.”

Technology may help some this time around. Apple and Google and a half dozen other companies are developing cellphone applications that use GPS or Bluetooth technology. These apps can alert users that they were recently near an infectious person or allow them to retrace their steps to see if they crossed paths with a carrier. But most experts insist technology will merely support, not supplant, human tracing. And it may never be as effective in the U.S. as it has been in other countries.

“The things people are trying to do passively, with phone apps, aren’t the same as active contact tracing,” says Eric Perakslis, a data science professor at Duke University who worked in Sierra Leone during the Ebola outbreak. He has advised several states on COVID-19 tracing. “The Bluetooth stuff is proximity monitoring. It’s contact-tracing-like. It’s not the same as interviewing people … making sure they have food or medicine, asking if they need help. You’re not comparing apples to oranges. You’re not even comparing apples to pork chops.”

For starters, tech solutions are valuable only if widely adopted, and polls show Americans are skeptical and worry that apps could infringe on data privacy. The U.S. also has so many cases that it can’t rely on weak crowdsourcing with limited penetration to find them all. Meanwhile, in much of Asia, previous viral outbreaks bred enough comfort and respect for tracing that there is less need for the handholding provided by human tracers. In a few countries, notably China, authoritarian governments also leave citizens little choice but to participate.

Americans, on the other hand, are less sensitized to the need for tracing, more independent, and more suspicious of government monitoring. That brings its own complications.

Skeptics have sowed doubt about tracing

The whole point of tracing is to find people quickly, ideally within 48 hours, and get them to act to halt the pathogen’s spread. In the best of times, the job isn’t easy. Not everyone recalls every contact. Some people don’t answer their phones. Some can’t afford to stay home.

And modern America is almost perfectly conditioned to accentuate these difficulties. The virus has hit minority communities hard. But after years of stepped-up immigration raids and racial demagoguing, many are reluctant to speak up.

“There’s distrust, misinformation, a stigma, no united front,” says Rupa Narra, a pediatrician at New York University’s South Brooklyn campus who worked overseas for Doctors Without Borders and once worked as an epidemiologist for the CDC. She’s seen parents of sick kids in her hospital downplay their own symptoms even as she sees them coughing. “I think we’ve not done a great job reaching different parts of our population.”

Talking heads on cable news and lawmakers in several states, including Louisiana and Minnesota, have demonized contact tracing. Washington state Rep. Jim Walsh, a Republican from Aberdeen, a city near the Pacific Coast, attacked tracing at a recent rally. “I’m very concerned about the rhetoric about ‘armies’ who are going to compel people to give up names of who they’ve been around,” he said later in an interview.

Whether lawmakers fret about cost, consider tracing a barrier to states reopening, or view it as a cudgel to attack political adversaries, their criticisms have muddied efforts to develop public support. In mid-May, after complaints, Washington Gov. Jay Inslee, a Democrat, rescinded plans to require restaurants upon reopening to have customers leave names and contact information. He wanted owners to be able to forward that data to tracers if there were an outbreak in their establishment. The state’s health department also issued a rare statement to bat down rumors that tracing could lead to forced quarantines. Tracing is voluntary.

At the same time, a feature of public health—that most control is local—means the amount and priorities of tracing efforts will vary substantially from state to state. And privacy rules around health information generally prevent health districts from directly contacting individuals outside their state. That will likely prove cumbersome as states reopen and contact increases, especially as health authorities who are tracing cases race against the clock.

“If I call Joe Schmoe, and he says, I had these five contacts, and this person was visiting from Florida, technically I can’t call the person in Florida,” Pogreba-Brown said. “I have to work through Florida’s health system.”

Tracing and testing ramp up

So what do we do about that? The CDC should pull states together to improve data sharing and make sure contact tracing reaches the places that need it most. “This is clearly a national-level disaster that should be coordinated at the national level,” said Ben Brunjes, a University of Washington expert in emergency management and contracting.

In lieu of a rigorous federal response, Partners in Health, a Boston-based nonprofit, is embedding expert teams in health districts around the country to build up tracing programs. It is providing advisers to offer technical assistance. In April, the group worked with Massachusetts Gov. Charlie Baker to rapidly hire 1,000 tracers. By mid-May these workers had reached 32,000 people. The nonprofit now plans to add 600 more tracers in Massachusetts.

Meanwhile, Harvard’s Jha said testing is finally on the upswing. He has projected that the country needs 900,000 tests a day. He said logjams for ingredients like reagents to detect virus in samples can be loosened with a little cajoling. “I’ve talked to the companies,” he said.

Testing also could rise by orders of magnitude if samples are pooled. Since the vast majority are negative, authorities could combine a dozen specimens into one batch and then only retest the individual specimens if the batch tests positive. “If you have 20 test kits, you could still test a hundred people,” Jha said. If testing let people find out more quickly if they’re carriers, it would make them more responsive to tracing and less likely to break self-imposed quarantines.

To combat public wariness, experts say the U.S. needs a sustained campaign of accurate, apolitical information making clear that tracing is voluntary, confidential, and secure. Experts point to the Boston Public Health Commission, the nation’s oldest, which has spent decades making inroads with minority communities. Using a diverse staff that speaks eight languages, they’ve developed contacts that help navigate political barriers. “It’s about relationships,” said Thomas Lane, associate director of the commission’s Infectious Disease Bureau.

Health districts should also try to develop contacts among protest movements and tracing skeptics, Jha said. The idea is that hearing from a tracer should be no more controversial than getting a concerned call from your doctor.

In the meantime, Pogreba-Brown said, those who get sick should at least alert friends that they may have been exposed. “Getting a ping on your phone or a call from a stranger is very different,” she said, “from having your friend say, I feel like death, and you’ve been exposed, please take care of yourself.”