Another day, another U-turn. The government has abandoned the “world-beating” centralised contact-tracing app it spent three months and millions of pounds on developing, to integrate it with the toolkit for a decentralised tracing app, developed by Google and Apple. It will not be ready until the autumn, or even the winter.
To summarise: the NHS app is good at distance, bad for privacy, to the extent that its Bluetooth-reliant software can’t run effectively on Apple devices, due to those devices’ privacy protections. The Apple/Google app toolkit, meanwhile, affords greater data privacy, but at the expense of accurate distance measurement.
There is some sympathy with Matt Hancock’s push for a state-developed app that overcomes these accuracy issues and some frustration at a perceived lack of cooperation from Apple. But only some sympathy, because need for the U-turn was entirely foreseeable months ago and those intervening months have cost the government money, time, and political credibility.
But while the outcry continues over yet another U-turn, it is worth stepping back to observe the bigger picture, which suggests the main problem with the test and trace programme is not the tracing, but the testing.
“Independent SAGE”, the group of experts who have assembled to “shadow” the recommendations made by the government’s scientific advisers, have released a report highlighting concerns over “extensive data gaps” in the tracing system as a result of “haphazard” testing, which, it concludes, makes the current system “unfit for purpose”. It suggests there are missing cases in the community, and that a failure to collect data at every point of the scheme prevents a robust evaluation of its performance, among other detailed criticisms.
Some of the report’s recommendations have already been implemented since its release two days ago: Dido Harding announced yesterday that test results are now being lined up with GP records, in a welcome step to address the widespread concern that the tracing scheme is insufficiently integrated with local healthcare providers. But there are myriad further concerns, and it is worth noting Dido Harding’s own acknowledgement that the scheme will not be fully functioning until September.
It’s easy to get excited about another, very obvious, government U-turn over an app developed by the very politician who once named an app after himself. But the real problems with the test and trace system are myriad, small failures in data collection, insufficient local healthcare involvement, under-performance by private contractors, among other things.
These are not as sexy, as simple, nor as accessible to the lay reader as a single, obvious app disaster. Nevertheless, these are the things that desperately need to be scrutinised and put right.