Analysis: When society begins to open up and the virus is less pervasive, a more nuanced approach towards contact tracing apps may be appropriate, but even that approach brings challenges.

By Dr Jim Buckley and Dr Cristiano Storni, UL

Contact tracing - where close contacts of known, positive, Covid-19 cases are identified, contacted and asked to self-isolate/get tested - is seen as an integral part of the fight against the spread of the virus.

Digital contact tracing apps are seen as a vital component of that effort. These apps typically compliment the manual contact tracing effort by informing people who have downloaded the app, near-instantaneously, of any close contacts they have had with any Covid-19-positive individuals (who have also downloaded the app).

These apps work on a retrospective model that assumes an altruistic user. You download the app and, if you are unfortunate enough to test positive, you can upload your close-contact details so that other users with the app who have recently been in contact with you will be warned.

It's retrospective in that you are already Covid-19 positive when you upload your details or you have already had a close contact with a Covid-19-positive person by the time you receive an alert.

It’s altruistic in that you upload your details for others only when you already have tested positive, with the aim of alerting them, and in the hope that they will will self-isolate, get tested, and alert their close contacts.

This model is one of the safest in terms of protecting your privacy, and is favored in places like Europe – where privacy is explicitly regulated by the EU’s General Data Protection Regulation legislation (GDPR): your list of contacts is anonymous and encrypted, and stays in your phone until it needs to be shared.

We need your consent to load this rte-player contentWe use rte-player to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences

On Radio 1's Drivetime, Drivetime's Fergal Keane reports on the release of the Covid-19 tracing app, and Elaine Burke, Managing Editor, Silicon Republic, and Brendan Jennings, Waterford Institute of Technology & Director of Connect, discuss the public's uptake of the app and the security concerns surrounding it.

Even then, when users are diagnosed with Covid-19, they have no obligation whatsoever to share their close-contact list. If that data is shared, it is shared anonymously and there can be no record of the time of contact, let alone the location (data that is considered sensitive). The downside is that this model does not seem to offer clear incentives to the users themselves.

So, what if we approach the contact tracing problem from the perspective of a selfish user? In this scenario, apps would be developed to proactively prevent Covid-19 from ever reaching the user, and so adoption of the app would be based largely on self-interest.

Such an app might build up an (anonymous) picture of you in terms of the people you meet and the frequency/duration of those meetings. Likewise it would build up a similar picture of those people and the people they meet and so on, resulting in an encompassing picture of your wider social network.

If users also registered their positive Covid-19 results, this picture could be augmented to show how closely Covid-19 is encroaching on your social network and more specifically, on your 'inner-circle’ of friends. The presumption is that, as Covid-19 comes closer, the selfish user will adapt their behavior (become more reclusive) to proactively avoid catching the virus.

This is the approach adopted by a team from CMU, led by Professor Po-Shen Loh, in an app called NOVID where a user gets a bar-chart representation of how ‘close’ the virus is to them in terms of their social network.

Of course such an approach is not perfect. For example, if some of the people in your social network don’t have the app installed or don’t register a positive test, you may be unaware of how close Covid is and may maintain your current, gregarious social behaviour.

Indeed government restrictions argue for more absolute contact-limiting, and this is appropriate in times of high community transmission. But a more nuanced approach may be appropriate when the virus is less pervasive, and society/economies begin to open up.

Likewise data-protection is an issue here: even though this approach captures no personalized user data, and so users can be entirely anonymous, the social-network data would need to be centrally collected and aggregated, potentially providing indirect opportunities to de-anonymize the information.

It should be noted that this is very much not an either/or scenario where we dump existing altruistic Covid-19 contact tracing apps in favour of a NOVID-type approach. Rather, the goal here is to use any means possible to slow the spread of the virus so there is an important role for both types of apps.

We need your consent to load this rte-player contentWe use rte-player to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences

On Radio 1's Morning Ireland, Niamh O'Beirne, HSE National Lead for Testing and Tracing, discusses the uptake in walk-in Covid-19 testing centres and the ramping up of contact tracing.

But even the suggestion that we use both approaches is problematic. Google/Apple have provided a service to app builders that allows them identify close-contacts based on Bluetooth-based phone link-ups, and most existing national efforts (including Ireland’s) use that service.

Yet Google/Apple will only allow one app from the health authority over each jurisdiction/country access to that service and they demand higher privacy constraints than a NOVID-type approach would involve.

Still, this is another complimentary approach worth considering, providing the opportunity to interrupt the Covid-19 lifecycle before Covid-19 interrupts the user’s.

Indeed, if privacy concerns can be overcome, maybe the future lies in a combined app where both the selfish and altruistic concerns of the user can be addressed to improve the uptake of contact tracing apps, increase their retention, and more effectively fight the spread of the virus.

Dr Jim Buckley is an Associate Professor in the Department of Computer Science and Information Systems at UL. He is also a member of the Lero Research Centre for Software. Dr Cristiano Storni is a lecturer in Interaction design, Course Director, MA/MSc in Interactive Media in the Department of Computer Science and Information Systems at UL.


The views expressed here are those of the author and do not represent or reflect the views of RTÉ