Sarah Cavey, a Denver real estate agent, was delighted last fall when Colorado rolled out an app to warn people of possible coronavirus exposures.
Based on software from Apple and Google, the government smartphone app uses Bluetooth signals to identify users who are in close contact. If a user later tests positive, that person can anonymously notify other app users that the person may have crossed over with on restaurants, on trains, or elsewhere.
Ms. Cavey immediately downloaded the app. After testing positive for the virus in February, she couldn't get the special verification code she needed from the state to warn others, she said, even after calling the Colorado health department three times.
"You're promoting this app to make people feel comfortable," said Ms. Cavey, adding that she has since deleted the CO Exposure Notifications app in frustration. "But it doesn't really matter."
The Colorado Health Department said they have improved their process and are now automatically issuing verification codes to anyone in the state who test positive.
When Apple and Google announced last year that they were working together to create a smartphone-based system to contain the virus, their collaboration seemed like a game changer. Human contact tracers have struggled to keep up with the rise in virus levels, and the trillion-dollar competing companies, whose systems power 99 percent of the world's smartphones, had the potential to quickly and automatically alert far more people.
Soon after, Austria, Switzerland, and other nations introduced virus apps based on Apple's Google software, as did around two dozen American states, including Alabama and Virginia. According to an analysis by Sensor Tower, an app research company, the apps have been downloaded more than 90 million times to date.
However, some researchers say that companies' product and policy decisions limited the usefulness of the system and raised questions about big tech's ability to set global standards for public health tools.
Computer scientists have reported accuracy issues with Bluetooth technology, which is used to detect proximity between smartphones. Some users have complained about failed notifications. So far, there has hardly been any rigorous research into whether the apps' potential to precisely alert people to virus exposures outweighs potential disadvantages – for example, incorrectly warning non-exposed people, over-testing or not recognizing users who are exposed to the virus.
"It's still an open question whether these apps help actual contact tracing, are just a distraction, or even cause problems," Stephen Farrell and Doug Leith, computer science researchers at Trinity College Dublin, wrote an April report on Ireland Virus alert app.
In the United States, some public health officials and researchers said the apps had shown modest but important benefits. In Colorado, more than 28,000 people have used the technology to alert contacts of potential virus exposure. In California, where a virus tracking app called CA Notify was launched in December, around 65,000 people have used the system to alert other app users.
"Exposure notification technology has shown success," said Dr. Christopher Longhurst, UC San Diego Health's chief information officer, who manages the California app. "Whether it's hundreds of lives saved or dozens or a handful, when we save lives it's a big deal."
In a joint statement, Apple and Google said, "We are proud to work with health officials to provide a resource that has enabled millions of people around the world and that has helped protect public health."
Let us help you protect your digital life
The Apple and Google system, based in part on ideas developed by the Singapore government and scientists, includes privacy measures that provide health officials with an alternative to more invasive apps. Unlike virus tracking apps that continuously track users' whereabouts, Apple and Google software use Bluetooth signals that can estimate the distance between smartphones without knowing where users are. It also uses rotating ID codes – not real names – to log app users who have been in close contact for 15 minutes or more.
Some health officials predicted last year that the technology could inform users of virus exposure faster than human contact tracers. Others hoped the apps could warn commuters who were sitting next to an infected stranger on a bus, train, or plane – those at risk who contacted Tracer would typically not be able to identify themselves.
"Everyone who uses the app helps to keep the virus under control," said Chancellor Angela Merkel in a video to advertise the country's warning system called Corona-Warn-App last year.
However, the apps never received the large-scale efficacy tests that were normally done before governments introduced public health interventions such as vaccines. The software's privacy features, which prevent government agencies from identifying app users, have made it difficult for researchers to determine if the notifications were hindering the transmission of viruses, said Michael T. Osterholm, director of the university's infectious disease research and policy center from Minnesota.
"The apps played virtually no role in investigating any outbreaks that occurred here," said Dr. Osterholm.
Some restrictions occurred before the apps were released. For one, some researchers find that exposure notification software inherently excludes certain vulnerable populations, e.g. B. older people who cannot afford smartphones. Second, the apps could trigger false positives as the system is not set up to take damage control factors into account, e.g. B. whether users are vaccinated, wearing masks or sitting outside.
Proximity detection in virus alert apps can also be inconsistent. Last year, a study of Google's system for Android phones carried out on a Dublin tram reported that the metal walls, floors and ceilings distorted Bluetooth signal strength enough that the likelihood of accurate proximity detection would be "similar" Trigger notifications by randomly selecting passengers.
Such glitches pissed off early adopters like Kimbley Craig, the Mayor of Salinas, California. Last December, when virus rates rose there, she downloaded the state exposure notification app on her Android phone and soon tested positive for Covid-19. After she entered the verification code, the system was unable to send a notification to her partner, with whom she lived and who had also downloaded the app.
"If it doesn't take a person in the same household, I don't know what to tell you," said Mayor Craig.
In a statement, Steph Hannon, senior director of product management for exposure notifications at Google, said there are "known challenges in using Bluetooth technology to approximate the exact distance between devices" and that the company is continuously working to improve accuracy.
Company policies have also influenced usage trends. For example, in certain US states, iPhone users can turn on exposure notifications with one click by simply enabling a feature in their settings. However, Android users need to download a separate app. As a result, around 9.6 million iPhone users in California had enabled notifications by May 10, far exceeding the 900,000 app downloads on Android phones.
Google has built its system in such a way that states work on a wide variety of devices and can be made available as quickly as possible.
Some public health experts admitted that the exposure warning system was an experiment where they and the tech giants learned and built improvements over time.
One problem they discovered early on: To prevent false positives, states review positive test results before a person can send exposure notifications. However, it can sometimes take days for local laboratories to send test results to health authorities, which limits the ability of app users to quickly notify others.
In Alabama, for example, the government's GuideSafe virus alert app has been downloaded around 250,000 times, according to Sensor Tower. However, state health authorities said they could confirm the positive test results from only 1,300 app users. That's a much lower number than health officials expected, as more than 10 percent of Alabamians tested positive for the coronavirus.
"The app would be much more efficient if these processes were less manual and automated," said Dr. Scott Harris, who oversees the Alabama Department of Health.
Colorado, which automatically issues verification codes to people who test positive, has reported higher usage rates. In California, UC San Diego Health has set up a dedicated hotline that app users can call if they haven't received their verification codes.
Dr. Longhurst, the medical center's chief information officer, said the California app proved useful as part of a larger statewide public health push that included wearing masks and virus testing.
"It's not a panacea," he said. But "it can be an effective part of a pandemic response."