It starts with a phone call. You may have been in contact with someone who has coronavirus, a kind-sounding voice tells you: You should self-isolate and get tested for Covid-19, if possible. The caller is careful not to reveal the identity of the infected individual you’ve crossed paths with; if they’re good at their job, they’re sensitive both to your fears and the others’ privacy.
The voice on the line is a contact tracer — a person trained to track down and notify all the people who might have been exposed to coronavirus by someone who’s known to have the infection.
Contact tracing is being pursued as a key part of the global effort to target Covid-19 hot spots, slow the spread of infection, and reopen local economies. Most experts agree that to truly halt the disease, it will take far more widespread testing and, ultimately, an effective vaccine. But by carefully following the movement patterns of the human disease vectors that carry it, jurisdictions hope to be able to target their lockdown measures at individuals who are most at risk, allowing a semblance of normal public life to resume.
To safely reopen, the U.S. might need to train and deploy at least 100,000 contact tracers. But some hope that this laborious process could be at least partially automated. Tech companies are jockeying to introduce infection alert systems, which use phone data to do some of the legwork of tracking recent contacts. Apple and Google have teamed up to develop a tool that uses Bluetooth to identify which other devices (and the people who own them) you come close to each day; if someone gets sick, anonymous signals go out to the people that had entered their orbit, urging them to quarantine or stay home.
Outside the U.S., such tracing technology is already in the field. South Korea’s much-talked-about coronavirus control regime leans heavily on the smartphones that 95% of residents carry; authorities can piece together and share anonymized routes of potential carriers using location data, credit card records, and security tapes. The U.K.’s National Health Service is introducing a Bluetooth tracking app akin to Apple and Google’s; since it became available to the 800,000 residents of the Isle of Wight last week, more than 55,000 people had downloaded it. Iceland’s app uses location data, not Bluetooth data. India is making it mandatory for residents to download its app, on penalty of fines or firings.
The promise that Big Tech will come through with a miraculous tool to fight coronavirus has been a recurring theme of the U.S. pandemic response: Remember President Donald Trump’s promise of an all-powerful nationwide Google testing site that turned out to be little more than an FAQ page? Or the more recent rush to develop and buy up thermal scanners to detect fevers in public spaces, which are reportedly not very good at detecting fevers?
When it comes to contact tracing, public health experts warn that technology cannot and should not replace the need for trained disease detectives, and not only because of the privacy concerns any location-based tracking service introduces. Manual contact tracing — and the human connections they create — will remain critical to any robust tracing response, they believe. And right now, the U.S. is only just beginning to assemble the infantry needed.
Mustering an army of disease detectives
Contact tracing is an old tool in public health, often used to alert cases of their potential exposure to tuberculosis, HIV, and other STDs. But the scale and speed of the coronavirus pandemic offers an unprecedented challenge. Given the rapidity of community spread — each person can infect two to three people on average — the National Association of County & City Health Officials has suggested jurisdictions deploy 30 contact tracers per 100,000 people, or at least 100,000 for the entire U.S. population. To assist local efforts, researchers estimate it’s going to take an additional $3.6 billion in federal funding to pay enough full-time contact tracers to work for a year.
Right now, only about 11,000 contact tracer jobs have been staffed in the whole country, and local public health departments are ill-equipped to fill these roles themselves, thanks to millions in cuts to federal funding for public health programs over the last 15 years. Recession-era downsizing also eliminated 50,000 public health roles that were never recouped.
The speed and scale of hiring has been spotty, varying by state and city: only one state (North Dakota) was equipped to handle its estimated need, NPR reported, but other jurisdictions have rapidly been expanding hiring goals. Massachusetts, for example, has worked with the organization Partners in Health to onboard around 1,600 people to train, lead and execute contact tracing efforts — but the state has only just started transitioning from a volunteer corps to a bank of full-time workers.
Hard-hit New York, meanwhile, plans to hire up to 17,000 contact tracers statewide. With the help of the city’s Fund for Public Health, New York City leaders set a goal of hiring 1,000 tracers as soon as possible. During the first week of May, FPHNYC told CityLab it was still in the resume-reviewing process, and more challenges may arise during onboarding. Originally, the city’s Department of Health was intended to lead efforts, drawing on experts who had decades of experience tracking other communicable diseases. But in a move criticized by public health professionals, Mayor Bill de Blasio put the city’s public hospital system in charge instead. This could accelerate hiring, its head told the New York Times, and get people directly connected with testing and isolation resources once they’re contacted.
For California, a state of 39 million people, Governor Gavin Newsom aims to hire 20,000 contact tracers. As of now, CalMatters reports that around 3,000 have started working, and that the state will be aiming to train 3,000 more per week. San Francisco currently has a corps of about 150 contact tracers from UCSF working with the health department, said Susie Welty, who usually works in global public health at the University of San Francisco, but has recently started helping the San Francisco Department of Public Health with contact tracing efforts. The goal is to scale up to 250 by the time the city’s shelter in place is lifted at the end of the month, while passing along most of the duties to existing civil servants who are out of work, like librarians and court staffers. “There aren’t enough contacts now to keep that many busy, but there will be as things open up,” she said.
Some third-party efforts have been able to act faster to find eligible candidates, though they can’t deploy them directly: CONTRACE, an organization that helps health departments find and train contact tracers that launched on April 21, has already recruited and screened more than 20,000 people. Company founder Steve Waters says they they’ve since adjusted its target to 25,000. They pass on info about qualified candidates to health departments, who take on hiring from there. “The NY/NJ/CT effort looks to be very promising, and MA was very much ahead of the game,” Waters wrote.
With projections for national unemployment expected to crest 20%, finding available workers to fill this need shouldn’t be difficult. Both the job and the training can be done remotely: The Johns Hopkins Bloomberg School of Public Health just released a six-hour online Coursera course, which anyone can take for free.
The art of the trace
But not everyone is cut out to be a contact tracer: The ideal candidate must be smart, empathetic, and discreet. Public health training isn’t a must, but it’s a bonus; knowledge of local languages is preferred. “A lot of our contact tracers at the health department have trained in things like motivational interviewing skills and cultural competency and cultural sensitivity,” said Sarah Braunstein, the director of the NYC Department of Health’s HIV Epidemiology and Field Services Program, who was a co-lead of the department’s contact tracing task force. “So they really have experience in knowing not only how to understand themselves but also translate for the public.”
Though the work is done remotely, hirers look for locals who live in the area they’re canvassing, or at least understand its culture, customs, transportation network, and resource landscape. That, too, helps build an early connection, says CONTRACE’s Waters. “For example, if you have someone with a deep New York accent calling someone in Louisiana, and talking to them with the demeanor of a New Yorker, it is less likely they will gain the trust of that person than if they sound like and had the demeanor of someone from Louisiana,” he wrote.
Contact tracers don’t divine coronavirus cases by putting their ears to the pavement, nor are they supposed to follow up on self-reported presumptive cases, like stories shared on social media. Instead, the process begins at the moment a local health department is notified of a new case after receiving a positive lab test result, says Braunstein. Other departments might also follow up on cases that have not tested positive but are deemed a likely case by a public health professional. The coronavirus subjects are then called by a contact tracer, who asks about their illness history, living situation, demographic profile, and a list of all their recent “close contacts.”
Armed with that contact sheet, the tracers get on the phone and start dialing. They call anyone the person can remember spending more than 10 to 15 minutes at a time with in the two days before symptoms presented themselves. (They won’t call the Trader Joe’s you stopped at, but if you work at Trader Joe’s, they will call your shiftmates.) Welty says the average number of contacts has been 3.5 in San Francisco, mostly people they live with; in the next 2 months, when more people return to work, she expects it to rise to 7 to 10. In Massachusetts, the average is closer to two contacts, WBUR reported.
Because of privacy rules, contact tracers can’t reveal the identity of the coronavirus-positive acquaintance, but Welty says she tries to carefully discern whether they already know who it might be. That helps the tracer understand when the potential contact happened, which influences how long a case should isolate.
What comes next varies slightly by jurisdiction. In San Francisco, for example, every person who’s identified as a close contact is instructed to get tested. If they test positive, they’re instructed to self-isolate for the next 10 days, and continue to until it’s been at least 72 hours after they’ve had a fever; if negative, they’re still asked to self-quarantine for an additional 14 days. In other places where testing is less available, tracers might tell people to immediately start quarantining until it’s been 14 days since their contact with a case, and get tested if symptoms begin to develop.
After these instructions are given, the work doesn’t stop. Via daily text alerts or, more infrequently, phone calls, contact tracers check in on their subjects. They ask what symptoms have appeared, how severe they are, and whether they need medical attention.
Emma Ryan works as a volunteer contact tracer in the Boston area, where she spends five to six hours a week calling positive coronavirus cases. (Disclosure: She’s also a former college classmate of mine.) Though Partners in Health told the Boston Herald in April that only about half of the tracers’ subjects pick up the phone, Ryan told me she’s surprised at how receptive people have been to her cold-calls.
“Everyone has been super willing to talk and answer questions about like, whether they had diarrhea or not,” she said. “People also seemed to be pretty willing to help out, because [they know that] part of the reason we’re collecting info on symptoms is to help better understand how it presents in different people.”
There’s no app for this
The interpersonal skills required to build this relationship is one reason why public health experts are skeptical about using technology to replace manual tracers.
“The human element of contacting someone on the phone, gaining their trust, having them view you as an ally, and accepting your recommendation to self-quarantine and share who they’ve had contact with is very important and somewhat of an art,” said Waters. “I’m not sure a tech-based solution could replicate that.”
Early in the coronavirus outbreak, Singapore’s contract tracing efforts were singled out for praise by the World Health Organization. The city-state of nearly 6 million acted aggressively to contain the virus, supplementing its vast network of human tracers with a tracing app, TraceTogether. But since its March launch, only a fifth of its residents has downloaded it, and Singapore has more recently experienced a surge in cases. New outbreaks have been sourced to dormitories where migrant workers lived, where Welty says it’s unlikely such an app would have helped; Singapore’s human army stepped in instead.
Oxford University researchers suggest that a minimum 60% adoption rate is necessary for a tracing system to be effective. Of all the countries using technology as a tracing supplement, MIT Techology Review found that Iceland’s app has the highest adoption rate, at 38%. For the country of 364,000 people, which has largely managed to flatten its curve, the app “wasn’t a game changer,” an official overseeing tracing efforts told MIT.
Jason Bey, the product lead for Singapore’s TraceTogether app, put it even more simply in a Medium post: “If you ask me whether any Bluetooth contact tracing system deployed or under development, anywhere in the world, is ready to replace manual contact tracing, I will say without qualification that the answer is, No.”
The prospects of a tech fix may be even dimmer in the U.S., where privacy concerns are more widespread. Based on polling by the Washington Post and the University of Maryland, many Americans don’t trust tech companies with their public health information, despite assurances from the companies that they won’t save location data or store it in a centralized location. Three in five respondents reported being “unable or unwilling” to use such a system, and most expressed a greater trust in public health departments than Big Tech. Of people who owned smartphones, only 41% said they’d be willing to download the app.
That hasn’t stopped the emerging patchwork of coronavirus-tracing apps that are appearing at the local and state level, BuzzFeed News reported, which could introduce new incompatibilities between health departments and further scuttle any sort of coordinated national response.
Adding to the challenge, many communities especially vulnerable to the virus in the U.S. — migrant workers, people experiencing homelessness, and undocumented immigrants — do not have access to an internet-connected device, says Welty. They also might have more reason to hesitate before sharing movement patterns with an entity that they fear could pass them along to the government.
Even if an app was capable of locating potential contacts and asking the same series of automated questions, certain subtleties can’t be replicated, as when Welty carefully probes her subjects to try and determine when the contact happened. And a human tracer can also provide additional public health information, she says. For example, if a subject is told to stay inside but responds that they have to keep working to provide for their family, Welty submits their name and address to the city, which sends them food, cleaning supplies, or even diapers.
“We really need to be able to provide the services they need to isolate,” she says. Acts of support like this can help build trust among whole neighborhoods, like the communities of Latino San Franciscans Welty works with, who are testing positive at far higher rates than white residents.
“People also need guidance about where to get resources, how to communicate with their employers; some people fear loss of income or job,” said Dr. Emily S. Gurley, an infectious disease epidemiologist at the Center for Global Health at the Bloomberg School, in a press conference. “At the end of the day, this endeavor is about helping people in a difficult and scary time.”
But technology can be an effective ally in the Covid-19 battle. For those willing and digitally able to opt in, apps can be used by patients to record the results of daily temperature checks, or alert public health departments of their need for services, Braunstein said. And tracers like Ryan could use the assist when hunting down cold leads. Most of the people she’d contacted during recent shifts had been tested three weeks prior, which made remembering and finding the people they’d seen while developing symptoms much harder. One woman’s office used internal CCTV footage to help retrace her steps as soon as they knew she tested positive, Ryan said. By the time Ryan called, the brunt of the tracing work had already been done.
A lot has changed since the Ebola outbreak of 2014, when contact tracers in west Africa sat in centralized call centers or went door-to-door to find subjects. Today, technology can continue to push the craft forward, even without tying your health data to your GPS coordinates. “There are many ways that tech can support the public health function: It can help them be faster, it can help them organize data, it can help cases and contacts respond in a quick way about their signs and symptoms,” said Gurley. “Technology can be a support all along the way. But it never supplants the public health function.”