China’s coronavirus app could have unintended consequences

The Chinese government app that lets people check whether they’ve been in proximity to somebody with coronavirus could backfire by creating a false sense of security or further stigmatizing sufferers, according to experts. 

The app was launched this week, and is aimed at limiting the spread of a disease that has so far infected nearly 50,000 people and killed more than 1,300. But while most criticism has focused on the fact that China’s surveillance state makes the technology possible, some infectious disease specialists say that it’s another example of technology being used without enough attention to social concerns. 

Security and stigma

Users sign up for the “close contact detector” app by registering using their phone number, name, and ID, and then scanning a QR code on their smartphones. The app will tell them whether they’ve been in proximity to someone who has been infected—which can mean family members or coworkers, as well as strangers on public transit. If a user is found to have been in close contact, and therefore could be sick or contagious themselves, the app recommends self-quarantine and an alert to local health officials. (Each person can also check the status of three other people.) It’s unclear exactly how the app determines close contact, although state-run newspaper Xinhua suggests that it uses data from national transport and health authorities. 

But because tests are never 100% accurate, especially for new diseases, false negatives and false positives are always an issue, says Monica Schoch-Spana, a medical anthropologist at the Johns Hopkins Center for Health Security. 

Especially in the case of coronavirus (now called COVID-19), many factors—from the fact that Chinese officials are not counting mild cases, which can still be contagious, to the many people turned away from hospitals—suggest that infection may be more widespread than official numbers suggest. There have already been many reports of false negatives. 

As a result, “not everyone who could be capable of exposing you to the disease is actually captured in the database,” Schoch-Spana says. This, she claims, could lead to a false appearance of safety that might end up with people letting down their guard. Plus, she says, it’s not just contact that matters, but how vulnerable someone is. The app doesn’t seem to account for that. 

Being marked as potentially contagious (whether accurately or not) could have a big effect, too, because of the social reaction to disease in China, says Christos Lynteris, an anthropologist and epidemic expert who has done research in the country. Though all infectious disease creates an element of a moral panic, the stigma in China is also connected with the concept of “face,” which is similar to honor and moral standing. The stigma of associating with a contagious person is intense, and stigma is contagious in China in a way that’s not quite the case in the West. That means the app is likely to create a situation where people preemptively avoid others to avoid being implicated in a “loss of face chain-reaction," Lynteris says. While this isolation could actually help prevent coronavirus from spreading, it’s a socially punishing way to do so. Lynteris has even considered whether this is part of the app’s design: “It’s a strange thing which makes me think, is stigmatization simply a by-product—or is it what the application relies upon in the first place?” 

Data—for good and bad

The close-contact app is fundamentally a data-gathering enterprise. If someone is marked as possibly contagious, the government has that person’s information down to their ID number. Whether that’s a good thing depends, in part, on the trust someone has in Chinese institutions. 

Residents of Wuhan, the central Chinese city at the epicenter of the outbreak, are already treated as pariahs in China itself. The authorities have hunted people from the city, encouraging citizens to inform on each other and offering bounties, with prejudice becoming so acute that even government officials called for people to show more understanding. An app that holds such detailed information, combined with what Lynteris calls the long history of public shaming in China—such as “struggle sessions” that publicly humiliate a target, or meetings in which citizens accuse each other of wrongdoing—could turn negative quickly. With social media now acting as China’s town square, there is no more effective place to do this, he explains. Though the app is not public, he says, it reasonable to expect that the information may be gleaned to identify or expose people who are “hiding” contact histories that they may be unaware of.  

However, Schoch-Spana points out that this feature also has the potential to be very helpful. “Health departments all across the globe are under-resourced, and particularly during crisis situations,” she says. If the government is able to use a communications tool to reach those who are possibly infected, it would help them reach those who need help and take some burden off the hospitals. There are also steps the government could take to avoid downsides, she says. This could include making sure the anonymity of the people in the database and using the app is protected; refusing to single out specific neighborhoods or spaces as places of contagion; and developing a companion service that could teach about risk of exposure or help people organize volunteer efforts. 

Ultimately, although this app may not have been possible in other countries, it is evidence of the widespread desire to use information to “get a big-data picture of what an epidemic is doing,” says Darryl Stellmach, an anthropologist and epidemic expert who works with Doctors Without Borders. For example, researchers from Google tried to forecast the flu using people’s searches. (The result, Google Flu Trends, overestimated the prevalence of flu in two seasons by over 50%.) China’s app seems like a dream for epidemic responders—something that is comforting for people who want to think that experts using technology can manage the situation. “But epidemics at their heart are a social phenomenon, and we can’t purely use technocratic means to address them,” he says. 

Tools like surveillance and epidemic maps need to be combined with a view of how people react under pressure, Stellmach says: “I’m very hesitant about technological solutions that are rapidly deployed, in settings where there’s a great pressure to act decisively and to be seen to act decisively.”