The government’s contact tracing app, COVIDSafe, has been plagued with functionality and privacy issues. Now it has another damning strike against it. A new academic working paper says it’s not equipped to streamline the work for manual contact tracers. Its lead researcher even believes a 100 per cent uptake of the app wouldn’t make it work any better.
A group of researchers from the University of Queensland and Auckland University of Technology have released a working paper on digital tracing apps and COVIDSafe’s effectiveness. It concluded the latter is just not doing a great job.
At the core of the findings is a discussion of speed when it comes to detecting a coronavirus case and then informing any contacts. The researchers determined speed was crucial to a contact tracing app’s success and with COVIDSafe’s design, it simply isn’t possible.
Uptake percentages aren’t useful in discussing COVIDSafe’s effectiveness
The Department of Health confirmed 6.2 million users had downloaded and registered with the app. It’s not known how many of those users are actively using it each day. However, it means around 25 per cent of the total population has, at least, listened to the government’s messaging.
Still, it’s got a way to go before it reaches the magic 40 per cent uptake figure the government initially touted but has since backed away from. That figure itself is shy of Oxford University’s National Health Service (NHS) group’s 60 per cent model, which it believes is the benchmark for tracing apps to assist with the reduction of cases and ultimately, deaths.
“A 100 per cent uptake of the federal app will not do anything.”
But Professor Rhema Vaithianathan, an expert in social data and analytics at the University of Queensland and the lead researcher of the paper, believes the uptake is irrelevant as the app’s flaws are fundamental.
“In my opinion, a 100 per cent uptake of the federal app will not do anything,” Professor Vaithianathan said to Gizmodo Australia in a call.
"I don't find the uptake discussion in Australia very meaningful because I don't quite know how you measure what the right number is of a thing that doesn't increase speed."
Professor Vaithianathan explains that figure, which comes from Oxford University's NHS group, is based on digital tracing apps offering instantaneous detection — something COVIDSafe does not offer.
"True digital contact tracing has the ability of the app itself to notify you if you have been with someone while they were infectious," Professor Vaithianathan said, explaining European authorities were looking to apps that offered instantaneous, automatic cascading notifications to contacts.
"The worry I have, in Australia, is that that feature is not available so all these people have been downloading that federal app, thinking that somehow they'll be notified if they've been in contact [with someone infectious].
"They'll be notified but it will be by a manual tracer exactly like you would be notified if you didn't have an app."
The modelling is a bit complex but the paper quantifies it as best it can. By reducing the days between a person finding out they're positive and then informing their recent contacts, you can drastically limit further transmission. It's something manual contact tracing can do but would be made faster with a true digital contact tracing app, unlike COVIDSafe.
An overhaul of COVIDSafe might be the best way forward
The reason COVIDSafe isn't able to provide this is due to its design. It aims to find the gaps in people's memories when they're tasked with informing authorities of everyone they've been in contact with. While a positive case might remember all the family and friends they've seen in recent days, they might not know who was standing in line behind them at the supermarket. In theory, that's where COVIDSafe should come in.
“If they’re willing to change [to Apple and Google’s approach], then the change will need to be a fundamental overhaul.”
While COVIDSafe's design is a crucial flaw, according to the working paper, it's also plagued by functionality issues that stop it from providing its intended use.
iPhone users have complained of its limited ability to work in the background when the app wasn't open. This has since been fixed but other issues regarding the Bluetooth functionality on Apple devices remain.
Apple and Google's API framework, released in May, is a possible solution for those Bluetooth issues but it also offers a decentralised system that could potentially pave the way for automatic notifications.
It's still not known whether the Digital Transformation Agency (DTA) is planning to integrate the update wholly but it has admitted to it's in talks with the tech companies.
Doing so, however, would require a complete overhaul of the app's foundations, according to Associate Professor Vanessa Teague, a cybersecurity expert looking into the app.
"The surface appearance [of COVIDSafe] might not change much, but everything about the information flow would be different," Professor Teague said to Gizmodo Australia in May.
"If they're willing to change [to Apple and Google's approach], then the change will need to be a fundamental overhaul."
Offering users additional information would help drive engagement with the app
Given the speed COVIDSafe was created, Professor Vaithianathan understands the challenges the government faced creating new technology amid a global pandemic. Still, with millions of Australians downloading an app that was sold as crucial but has shown little evidence of it, she thinks the government could have been a bit more transparent and agile about it in the first place.
"If they play fast and loose with some things — like telling us download the app and it'll allow us to go out and go to the pub — then when they tell us, you know, to wear masks, or do other things, we might not believe them anymore," Professor Vaithianathan said.
The solution seems simple, however. The app needs to offer something Professor Vaithianathan calls a 'credence good' — a reason for people to check in with the app and get information from it too.
"The last thing you want to do is get people to download the app and realise you need another app."
Professor Vaithianathan explains this can be offered in three ways. The first is to enable automatic notifications of coronavirus exposure. It could also provide a de-identified way to self-monitor how many people you've been in contact with in a 15-minute block to help manage social distancing. The last is a digital passport, which changes colour and shows shop owners whether you're supposed to be in isolation or not.
"A government should think about all the different uses of the app because the last thing you want to do is get people to download the app and realise you need another app," Professor Vaithianathan said.
It's these sorts of features that are being considered or rolled out overseas — the U.K.'s NHS digital contact tracing app is an example — that she believes should be considered for Australia too as things begin to re-open.
With authorities sparsely using COVIDSafe data in the six weeks it's been available, it's hard to tell how valuable it would be if larger outbreaks were to occur. While it's a blessing for the country to have had such few coronavirus cases since its release, it's also been a curse for judging the app's worth.
As COVIDSafe edges further out of the community's memory and usage, its effectiveness could fall even further. Giving users something to come back for, such as stats, information or self-monitoring features, could be the band-aid solution it needs until an overhaul is considered.