Between a sweep of mountains and an expanse of dark waters, a 14-story building looms over Prince William Sound. Most of Whittier, Alaska’s 280 residents live in the peach-colored confines of Begich Tower, which was built in 1956 as a U.S. Army barracks. The building has its own post office and grocery store. An underground tunnel leads to the town’s small school. “We are our own petri dish—we share the same ventilation system,” says Jim Hunt, the city’s manager.
When COVID-19 reached the state, controlling visitors was the town’s best hope of keeping the disease at bay—and for months, they did. There are only two ways to reach Whittier: By boat, or driving the 60 miles from Anchorage, passing through a single-lane tunnel. Then in June, 11 seasonal seafood processors tested positive for COVID-19 and departed to isolate in Anchorage. A month later, two more cases appeared among workers at businesses along the harbor. Finally, in August, the virus penetrated Begich Tower. An employee who worked in maintenance—which includes COVID-19 disinfection—tested positive, along with five members of his family.
The employee chose to get tested in Anchorage, and there’s no obligation between the two city governments to discuss cases. But the busybody nature common to small towns eventually delivered the information to the city manager. Without the rumor mill, Hunt might not have known.
“If you test positive in another community, we wouldn’t know,” says Hunt, adding bluntly that “we have no contact tracing—none—outside of anecdotal evidence. You need human resources for contact tracing.”
Whittier’s dilemma may sound extreme, but it’s become an alarmingly common problem: Contact tracing, the process of identifying who may have come into contact with a known case of COVID-19, requires people, training, funding in at-risk places, and time—all resources the United States has been short of devoting during the pandemic.
It’s hard to overstate how big of a problem the lack of American contact tracing poses. The Johns Hopkins Center for Health Security issued a national plan back in April, reporting that until there’s a vaccine, “management of the COVID-19 pandemic will rely heavily on traditional public health methods for case identification and contact tracing.” The director of the Centers for Disease Control and Prevention warned Congress in June that the U.S. would need at least 100,000 contact tracers—and billions of dollars to fund these programs. Other estimates put the number much higher. But by the end of July, a joint NPR/Johns Hopkins Center for Health Security survey found that the U.S. only had 41,122 contact tracers—even though 11 states said their underfunded programs included staffing by unpaid volunteers.
Data discrepancies and understaffing have made contact tracing harder in some states than others. But the basic concept—find the people who may be sick and keep them from infecting anyone else—is the same, and everyone from mayors to universities to nonprofits are improvising ways to manage even when resources are almost impossibly tight.
“Personally, it’s hard not to be very concerned,” says Candice Chen, a physician and associate professor of health policy and management at the Milken Institute School of Public Health at George Washington University who has worked on contact tracing estimates for COVID-19. “I’ve noticed a definite sense of hopelessness. But we can’t afford to give up.”
Money and speed
Contact tracing capacity ultimately comes down to money, says Tom Hennessy, an infectious disease epidemiologist who spent 25 years working with the CDC and who has been advising Anchorage on its COVID-19 measures.
“Budget cuts have really had a negative impact on public health departments,” he says. “This is a really good example of what happens when you have an organization whose job is to respond to emergencies but is staffed only for a base-level of activity.”
Persistent lags in test results, which is its own ongoing problem in the U.S., can also complicate contact tracing. Tracing efforts start with a positive test result, but many people in the U.S. are waiting 10 days or longer for results. In a paper recently published in the journal Lancet, researchers found that just as for testing, timing in contact tracing is important.
“If you find fewer contacts, but you find them very fast, you can have more impact,” says Mirjam Kretzschmar, the paper’s lead author and a professor of the dynamics of infectious disease at the University of Utrecht in the Netherlands.
Up to 50 percent of transmission can take place before a person feels any symptoms, and 40 percent of people may never develop symptoms but can still be infectious. That means people who are potentially infected need to isolate as soon as possible, before they infect anyone else. “The earlier you can catch them, the more effective contact tracing is,” Kretzschmar says. She found that reducing delays in testing and tracing could prevent as much as 80 percent of onward transmission of a diagnosed case.
Switch to mobile
But all contact tracing isn’t created equal. In her study, Kretzschmar considered both traditional human-run contact tracing and the use of mobile apps. Mobile apps can tell investigators where people were while they were potentially infectious more accurately than people themselves, who might not remember everywhere they went. Those apps can also instantly notify users if they may have potentially been exposed to the virus as opposed to waiting for a health department to call.
Surprisingly, even if only 20 percent of a population uses this software, Kretzschmar’s model found that app-based tracing remained more effective than conventional tracing, due to its inherent speed. Apps are still limited by the turnaround of positive test results, but conventional tracing run by health departments is also restricted by the capacity of the human tracers and their success at finding the necessary contact information. She notes that the speed of contact tracing may be especially important as schools reopen, since children generally have a greater number of close contacts than adults.
Critics of mobile-based tracing cite concerns with data privacy. The Netherlands, where Kretzschmar is based, has much stricter data privacy laws than the U.S., so the health ministry itself is developing an app that works via Bluetooth, so it never centrally stores information, reducing the potential to leak private information to either governments or businesses.
By contrast in the U.S., tracing platforms designed by private companies or local governments have sparked serious privacy concerns over how sensitive health data will be stored and used. For example, an undocumented person might not want their geographic location to be tracked at all times; trolls or foreign operatives could falsely report COVID-19 infections in a strategic location. These kinds of misuses might be more likely to happen when data is collected by third-party companies, especially as American data laws allow more uses of private information. On September 1, Apple and Google announced the expansion of its Bluetooth-based system that simplifies the tracing process for state health authorities, allowing officials to gather and share this information without launching a standalone app.
Others worry that mobile apps fail to encompass the full gamut of tracing’s activities, which include connecting ill people with resources and information, such as advice on how to isolate safely. Human contact tracing at its best builds a relationship based on trust, and tracers often call multiple times over the course of someone’s quarantine to make sure people have what they need. Massachusetts is collaborating with Partners In Health, a nonprofit healthcare organization that typically works in impoverished settings, to develop a contact tracing program that delivers resources such as food stamps or alternate housing to people who need to isolate.
“To get people to quarantine safely, we have to make sure people have material resources to do that,” says Joia Mukherjee, the chief medical officer of Partners In Health. No one is going to quarantine if it means they’ll lose their job or housing.
People hate scam calls, so to ensure essential information reaches potential patients, Partners In Health helped the state government work with telecom companies to make sure caller ID showed that tracing calls came from public health departments. Partners In Health created and staffed a hotline, and tracers were also allowed to leave voicemails and answer questions via text messages.
These measures are far from standard. In Louisiana, at least some contact tracing calls appear with an 800 number. The Federal Trade Commission warns that scammers are already impersonating tracers, and advises against clicking on links in any texts from tracers. A real contact tracer won’t ask for your Social Security number or financial information.
The trick is figuring out how to make communication easy, but not too easy. The Massachusetts program required an enormous investment—at least $94 million according to the Boston Globe—from the state and private grants. Partners In Health hired and trained 1,900 people to do case investigations, collaborating with 400 local health departments. Mukherjee says the payoff is worth it, since the information that tracers learn often holds valuable clues as to how transmission is occurring.
“Nothing replaces in-person work,” Mukherjee says, noting that the teams also focused on finding tracers who spoke languages used in particularly vulnerable communities and translating information to provide those communities. “We are always trying to solve for inequities.”
When money is tight
Alaska has some of the lowest coronavirus rates in the country, which makes it even more remarkable that the state’s ability to contact trace was overwhelmed this summer. This shortfall illustrates the monumental task states such as Texas and Florida currently face. (Texas actually reduced its tracing workforce as its cases surged.)
"In Alaska, we've never before attempted contact tracing on this scale and for this long. Past efforts relied upon a small group of people who could effectively track cases using simple tools,” says Hennessy. “Now we're trying to expand that to hundreds of persons around the state using a new data system they are not familiar with."
In July, when new daily statewide cases were as high as 119, the state’s largest city—Anchorage—announced it’d hit its capacity to investigate new cases. Contact tracers were finding that people who tested positive had hundreds of contacts during their potentially infectious period. As a result, the Anchorage Health Department decided to publicly name bars and restaurants where there were possible COVID-19 exposures, much to the establishments’ ire.
“When you don’t have resources, you have to improvise,” explains Mayor Ethan Berkowitz, who added that the city would have preferred to contact people privately.
Since then, Berkowitz says Anchorage has hired more nurses and developed a training program with the University of Alaska to offer a contact tracing certification. The state health department has also switched from Excel spreadsheets to an online database to track cases in real-time.
“I don't think any city in the country was set up to respond to this pandemic,” Berkowitz says. “When you contrast us with other countries, you can see the deficiencies.”
Wuhan, China, for example, had 9,000 contact tracers communicating with tens of thousands of contacts every day back in February. New York City, in contrast, was very slow to develop an adequate contact tracing program, only officially launching June 1. Since then, workers have complained of confusing training and malfunctioning equipment.
Though best practices can be shared, and expertise from experienced field workers like Partners In Health recruited, we don’t have much time to learn. Back in Whittier, the end of the infected family’s quarantine is near, but there’s been another test-confirmed case in the tower. Without contact tracing, everyone’s just holding their breath, Hunt says.
“We keep waiting for the other shoe to drop."