On Oct. 1, New York state released an app that can notify you if you’ve come into contact with a person who has tested positive for COVID-19.
Called “COVID Alert NY,” the app is one of 10 currently active in states around the U.S. that are based on Google and Apple’s decentralized contact tracing system, which was developed to maintain privacy while also giving health authorities a potentially powerful new tool to clamp down on outbreaks of the virus.
“Everybody’s wondering, ‘I was next to this person, I was next to that person,’ but this can actually give you some data,” said New York Governor Andrew Cuomo, announcing the app’s release. “I think it’s going to not only bring contact tracing to a new level, but it’s going to give people comfort.”
Contact tracing—either via an app or the old-fashioned method of interviewing people—is crucial to notify people who don’t know they’ve been exposed to the virus, so they can isolate. But it always involves a tradeoff between privacy and public health. In Asia, countries like South Korea used contact tracing with great effect to keep a lid on their epidemics, but at significant cost to individuals’ privacy. As well as deploying legions of interviewers, South Korean authorities also used phone logs, card transaction records and surveillance camera footage to monitor infected citizens’ locations and find their contacts. In more privacy-conscious Europe, many citizens and lawmakers were uncomfortable with such a big privacy tradeoff.
Instead, over the summer, many European countries adopted a voluntary app-based model. From the beginning, it was clear this would be less far-reaching because it requires large segments of the population to voluntarily opt-in. In May, researchers suggested that in order for a voluntary contact-tracing app to control an epidemic without any other severe measures (like lockdowns) being taken, around 60% of any given population would need to download it, use it, and follow its instructions.
But no country in Europe has seen anywhere near that level of uptake. In Ireland, some 34% of the eligible population now use the local app on a daily basis—probably the highest rate in Europe, going off the available statistics. But the app only helped identify around 2,000 people who may have been exposed to the virus between its release in July and late September — even though cases rose by more than 7,000 over the same time period. By contrast, South Korea’s more extensive system was able to identify more than 10 contacts per individual case of the virus between January and May.
On top of the 10 where apps are already active, at least seven more U.S. states, and Washington, D.C., are planning to release contact tracing apps in the near future based on the same privacy-preserving system used in Europe. The main lesson from Europe, where at least 20 countries have rolled out apps based on Google and Apple’s system, appears to be that apps with built-in privacy protections have their uses but are not a one-stop magic solution to stopping the spread of COVID-19.
Experts instead say the apps are useful as one weapon in a larger arsenal of measures to combat COVID-19. “This app can and will save lives, it will prevent infections, and it’s an additive thing that supplements other interventions like social distancing, hygiene, mask wearing, and much more testing,” says Christophe Fraser, an Oxford University professor, referring to England and Wales’ contact tracing app, which is based on Google and Apple’s system. “I think the lesson for the U.S. is that of an integrated approach,” he says.
Here is what can be learned from Europe’s rollout of apps.
Communicating the privacy protections is vital
Across Europe, download rates of contact tracing apps have fallen short of the numbers that researchers had hoped for in the early days of the pandemic, when the success of contact tracing in some Asian countries seemed to indicate that apps might provide the key for a return to normal life.
While prominent epidemiologists and privacy researchers have given the green light to decentralized apps based on Google and Apple’s system, the task of convincing large numbers of people to download the apps is still a work in progress.
Soon after the long-awaited contact tracing app for England and Wales launched on Sept. 24, David Bonsall, a virology researcher at Oxford University, decided to type up a Facebook post to answer a list of common misconceptions.
“Is it going to steal my data?” was the first on the list. “No,” he wrote. “It’s completely private. All your data stays on the phone, if you delete the app, you delete the data.”
Bonsall, also a scientific adviser to the U.K.’s Test and Trace program, quickly saw his post receive some 3,000 likes and thousands of comments, many of them raising further misconceptions. “There are lots of rumors that are not true about the app,” he says. “They were not substantiated, from what I saw going on.”
Concerns about user privacy are generally unfounded in the case of apps that rely on Google and Apple’s system. Typically, when you have an app based on that system installed and active, your phone transmits a Bluetooth signal that is picked up by other phones nearby with the app installed. The signal is a series of unique strings of random numbers and letters, and is not connected to your name or any other facet of your identity. If you subsequently test positive for COVID-19, you can tell the app, and it will send your unique, but de-identified, strings to a central database. Other people’s apps automatically check against that database for matches and then notify them if they’ve been in close proximity to you for a risky amount of time. But neither you, nor the person you’ve come into contact with, nor the government, nor Google or Apple can deduce any personal information from that data.
One measure among many to stop the virus
However effective an app is, contact tracing doesn’t work without a broader public health infrastructure, specifically an effective testing regime. People can only ever know they’ve come into contact with a potentially infected person if that person has been tested for the virus. That means there’s no point having a popular contact tracing app if your country or state doesn’t have widespread availability of testing, which many countries are still struggling with months into the pandemic.
Other factors are important too, like ensuring that people who’ve been told to quarantine by the app actually follow the instruction. In the U.K., only 18.2% of people who developed symptoms reported completely self-isolating after the symptoms began, according to a King’s College London survey. And just 10.9% of people who were told by contact tracers they may have been exposed to the virus (who may not have had symptoms) stayed at home for 14 days, as required. The study, which has not yet been peer-reviewed, found that non-adherence was highest among people with dependent children; members of socioeconomic groups hit hardest by the pandemic; and essential workers, suggesting that state support for a person who is instructed to change their lifestyle is essential for a contact-tracing system to work effectively. “If you ask people to isolate and quarantine, then somehow they need support,” says Fraser, the Oxford professor.
By contrast, in Austria, which has a generous system for paying a full wage to workers forced to quarantine and fines for those who don’t comply, only 1.5% of people told to self-quarantine were found to have broken the guidelines, according to the Washington Post. (Austria has a decentralized contact tracing app that has been downloaded by some 12% of the population.)
Consistent public health messaging is crucial
Contact tracing apps also can’t work effectively without consistent public health messaging, experts say. When messaging is contradictory or rapidly changing, people tend to become confused about what measures they should be following, and lose trust that their government is acting in good faith. In the case of contact tracing apps, that consistency is an essential precursor for citizens to feel prepared to agree to their government’s requests that they download and use a tracking app—a voluntary change in behavior that they likely would never have agreed to before a global pandemic.
“In public health, you have to have a clear message, and you have to repeat it consistently over a lengthy period of time,” Fraser says. “Further increasing confidence in the system is one of the things that’s required” for the apps to have an effect, he says.
In the U.S., where there is a political climate of deep polarization and skepticism toward scientists and medical professionals, people may have less of a willingness to adopt contact tracing apps. “It’s a similar question to what drives uptake of mask use,” says Fraser. “To me it seems clear that if you’ve got a lack of clear leadership, and a lack of a clear strategy, it really is not helpful.”
Experts predict that, in the U.S., this means apps will likely face problems given the fraught political climate, since using a contact tracing app requires a deeper level of trust—in tech companies and the government—than the simple act of wearing a mask does. “If you have a president telling people not to be afraid of the virus, how likely is it that they’re then going to go to the extra effort of downloading an app?” says Samuel Woodhams, a privacy researcher who has compiled a database of different contact tracing apps in use around the world. “A lot of people still have privacy concerns about these apps. A lot of the time, they may be misguided, but they still exist. And without clear public messaging, their effectiveness is only going to be limited further.”
In his new unofficial job as a debunker of misconceptions about the England and Wales contact tracing app, Bonsall has found one way of getting the message across. “When people have said to me, ‘look, I don’t trust the government on this app,’ one of my first responses is: you don’t have to,” he says. “The system that is built into this app uses the code from Google and Apple. You can trust them to look after their reputations.”
“Apple and Google have put their brand behind this piece of code and sent it to all governments,” he says. “They are very concerned about the potential for their technology to be used for nefarious purposes. So they built a system that was totally private.”
Manage expectations for success
Although researchers stress that contact tracing apps can still be helpful even with low download rates, overall, the available evidence from around the world suggests that contact tracing apps have so far had a “marginal” effect at best, says Woodhams, the privacy researcher. “Even when we see cases in which apps have been downloaded a lot of times, and alerted a relatively high number of people, in terms of what that does for overall numbers and the resurgence of the virus, I think it’s important that we understand the limitations of these apps,” he says. “Contact tracing apps are not going to be a silver bullet.”
But contact tracing apps don’t just benefit society at large—they keep individuals personally more safe too, Bonsall says. “People are thinking like chief medical officers, asking what the population needs to do,” he says. “But it’s okay for people to think about themselves in this.”
When he was advising the U.K. government on the contact tracing app for England and Wales, Bonsall and his colleagues ran simulations and found that uptake was likely going to be patchy across the population. But they also found something surprising: “If you’ve downloaded the app, then you’re more likely to be in a group among other people who are also downloading the app,” he says. “That means you’re actually quite likely to be getting a benefit, even if uptake in the population is quite poor. People think about this as saving other people’s lives. But actually, if you’re in a social network of people using this app, you are [more] protected from COVID-19 too.”