Contact-tracing apps have consistently been hailed by governments as a key tool in fighting the spread of COVID-19, but we've yet to see many of them released to the public.
On Thursday morning I awoke to discover a COVID-19 contact-tracing app was finally available for me to download. Except the app I ended up downloading wasn't the one I'd initially anticipated.
Back in March, as the world was waking up to the realities of the coronavirus and was starting to go into lockdown, the government in the UK, where I live, boasted that it had already started work on a contact-tracing app that would be ready by May. Spoiler alert: It wasn't. And it's still not ready, and I've ended up using a different solution altogether.
Contact tracing is a long-established form of finding out who might've been exposed to a disease by asking someone with a diagnosis who else they've recently been in contact with. The digital version is effectively the same as the traditional approach, but instead of health professionals asking questions, your phone keeps an anonymized record of the people you've crossed paths with, using Bluetooth and a dedicated app, and it lets you know if you need to get a test.
From the start, the UK government put its app right at the center of its track and trace strategy. The app would be essential for tackling the spread of COVID-19 and reopening the economy when the time came, officials said.
Initially the UK went with a centralized model for its app, which meant that all data collected through the app would be uploaded to and processed by a central database. This would allow for the gathering of a certain amount of public health data, as opposed to a decentralized model, which involves data being stored on people's personal devices and processed fully anonymously only when necessary.
Privacy experts warned the government, and other governments working on similar projects, that the centralized model was a data privacy nightmare and that people might not trust the app, resulting in them not downloading and using it. Meanwhile, technical experts pointed out an even bigger flaw in the centralized approach: Apple's rules for apps meant that data collected in the background via Bluetooth (an essential part of how digital contact tracing works) couldn't be uploaded to a centralized database.
Effectively, they were saying that the app the UK government was building wouldn't work properly on iPhones.
But even when Apple and Google together released a protocol to help people build decentralized contact-tracing apps that would work properly on phones, the UK didn't change its approach. It said Apple and Google would slow it down, and it pressed ahead with trials and with promising that its software would be available to members of the public within weeks.
It took until mid-June for the government to recognize publicly that its attempt had failed and that the app's arrival wasn't imminent. It finally switched to using Apple and Google's decentralized model, and said the app would be available later in the year, perhaps as late as winter.
Contact-tracing apps in the wild
In the meantime, other countries were building their own contact-tracing apps, including Ireland, which commissioned local software company NearForm to work on it. The company had begun working on contact tracing back in March, also focusing initially on building a centralized app, but it switched to using Google and Apple's service as soon as it became available.
According to NearForm's technical director, Colm Harte, both the company and the Irish health authority quickly saw the benefits of moving to a decentralized model supported by Apple and Google, even though it meant collecting less data for epidemiologists to work with.
Apple and Google's platform solved the privacy issue and the iPhone issue, and it also meant the app would be properly calibrated for all Android phones. "iOS devices are all pretty consistent, because it's the same hardware," said Harte in an interview. "But from an Android perspective, there's literally tens of thousands of combinations of OS versions and hardware versions, so Google took on that problem."
The switch to Apple and Google's model after initially taking a centralized approach wasn't overly tricky, Harte added. "From a tech perspective, it kind of meant that we ended up throwing away maybe a few thousand lines of code," he said. "We were happy enough to do it, because you could see the upside of where we would end up."
Ireland's app was up and running at the start of July, and Nearform's application for Northern Ireland was available by the end of that month. As the virus has progressed, the devolved nations of the UK -- Scotland, Wales and Northern Ireland -- have increasingly taken the handling of COVID-19 into their own hands. Northern Ireland was the first to find its own app rather than rely on the UK government's effort, but it wasn't the last.
This week NearForm's app went live in Scotland, where I'm based. The app was quick to download and involved an equally quick and straightforward setup process. It's now happily ticking away in the background on my phone. If I cross paths with anyone who's later diagnosed with COVID, it'll alert me, when that person uploads his or her positive-test code.
Some people are concerned that the app running in the background will drain the phone battery faster than usual, but I haven't noticed any issues with this in the first 36 hours I've been using the app.
Next up, NearForm's app will be coming to Pennsylvania and Delaware, with more US states to be announced shortly.
For Ireland and Northern Ireland, it was important for their apps to be interoperable, said Harte, as so many people cross back and forth across the border every day. The same is likely to be true of US states.
Contact-tracing apps don't necessarily have to be built by NearForm to be interoperable, Harte explained. As long as apps are built on top of Apple and Google's service, then "technically it becomes a lot more straightforward" to make them work together.
According to Harte, throughout the coronavirus crisis, teams building different apps in different countries have been in close communication to share what works and what doesn't. NearForm has now open-sourced its software and is working closely with Linux Foundation Public Health to ensure that those who need it are able to access it.
And that app the UK raced to build, and promised would be one of the first to be available? The government still hasn't released it publicly. People living in England and Wales will be able to download it on Sept. 24, according to an announcement on Friday.
"My team have worked tirelessly to develop the new NHS COVID-19 app and we are incredibly grateful to all residents of the Isle of Wight, London Borough of Newham, NHS Volunteer Responders and the team that went before us; the learnings and insight have made the app what it is today," the managing director of the NHS COVID-19 app, Simon Thompson, said in a statement.
Are they working?
One of the early concerns about contact-tracing apps was how effective they'd really turn out to be, and it's still too early to say for sure. To truly determine their impact, experts will need to study them long term rather than just over the course of a few months.
The conventional wisdom has already shifted away from the apps being a solution to them being a tool used to augment a wider contact-tracing strategy. "The manual contact tracing is actually the piece that drives everything," said Harte.
Some contact-tracing apps, such as Norway's effort, have already been shut down due to privacy issues. But Harte says NearForm's app is off to a good start.
With over a million downloads within the first 36 hours, the company considered the launch of the Ireland app a success. Within days people were testing positive and uploading random IDs, and other users were getting alerted as a result of those uploads.
On Friday, Scotland's first minister, Nicola Sturgeon, thanked the more than 700,000 people who'd already downloaded the Protect-Scot app in the first 24 hours. Getting as many citizens on board as possible is a priority for governments, because the more people who use the apps, the more effective they'll be.