Yes, the contact tracing app did work

So why is nobody interested?

The NHS Covid-19 app at work

Imagine that a cutting edge technology aimed at cutting the spread of Covid-19 had been released during the pandemic, and had been the subject of great controversy. Then imagine that a paper in a prominent scientific journal had shown that this innovation had prevented a million Covid cases, averting around 10,000 deaths. Surely that would be huge news?

Well, the innovation in question was the NHS Contact Tracing app for England and Wales and those figures are from a paper published in Nature this week looking at its impact. Yet somehow after scouring the newspapers and listening to the radio I cannot find the headlines I think this story merits.

Maybe my news judgment is skewiff because as the BBC’s Technology Correspondent I spent months obsessing over the app, the science behind it and the political rows surrounding it. We broke the story of the decision to ditch the first app over concerns that a centralised model would threaten privacy and just wouldn’t work.

When the version built using a system approved by Apple and Google finally launched in September 2020, we reported on the confusion many users felt when trying to use the app. And when the app kept millions at home in what became known as the pingdemic I analysed the battle over whether what had been touted by ministers as a key weapon against Covid-19 had outlived its usefulness.

By the autumn of 2021 it seemed the received wisdom about the contact-tracing app seemed to be that it was an expensive dud, an experiment that had failed. (The misinformation that it had cost £37 billion when FullFact says the real figure was £35 million did not help.)

Now the. study in Nature appears to show that view was wide of the mark. The paper analyses data from the first year of the app’s operation up to September 2021, tracking the rise in downloads and then the falling off as many responded to the pingdemic by removing it from their phones.

The app, which used Bluetooth to record users’ proximity to people who later reported positive Covid. tests, was seen by some critics as less reliable than good old fashioned manual contact tracing. That always seemed unfair to me - Bluetooth may be a pretty blunt weapon in measuring distance but so is human memory. Do you remember how many people stood close to you last Wednesday? And the study’s authors point to one huge advantage of an automated system:

“The app is able to process exponentially increasing numbers of positive tests and/or high contact rates, whereas manual tracing has a reach determined by staffing numbers that varies over time..”

In a blog about the paper one of the authors admits that the decentralised nature of an app which records very little data about its users makes analysing its impact quite a challenge:

”Our estimates include a lot of uncertainty because the app is privacy-preserving by design.”

But there is still quite a lot of data - for instance about the number of people who recorded a positive test after receiving a notification that they may have been exposed to Covid-19.

The conclusions have what is called a wide confidence interval - in other words a broad range of possible estimates. So while the central conclusion is that the app averted 1 million cases and 9,600 deaths the estimate of cases ranges from as high as 1.4 million to as low as 450,000 and estimated deaths could be as high as 13,000 or as low as 4,600.

But even if the app saved “only” 4,600 lives that was surely a major contribution to the battle against Covid. The study ends by making a case that contact tracing apps are here to stay:

“Our study adds to the body of evidence which shows that digital contact tracing apps have major potential for reducing transmission of SARS-CoV-2 when combined with strong user engagement.”

In the case of the NHS Covid-19 app that strong user engagement was undoubtedly weakened by criticism from journalists and politicians that was sometimes inaccurate and unfair. So here’s a conclusion for all of us writing about the science of health - careless reporting can cost lives.