Gambling with Testing and Tracing
- Published
- in Health
We regularly read about fraud in sport, whether that’s cricket, football or horse racing, where a player accepts money to affect the result. There are arrests, trials and the culprits either banned or sent to jail. But what happens when a Government Minister takes a bet on a Government policy and then manipulates the data to win it?
It may sound bizarre, but that’s what has happened here in the UK, with Matt Hancock, our Secretary of State for Health. In an interview on LBC with Nick Ferrari, he was asked if he would take a £100 bet on reaching his target of accomplishing 100,000 coronavirus test each day by the end of April. Anyone with a scrap of morality would have answered “I don’t gamble on people’s health”, but not young Matty. Although not an ex-Etonian, like many of this cabinet colleagues, Matty always looks as if he wants to be seen as one of the posh boys who likes a flutter. After a brief hesitation, he accepted the bet on behalf of NHS charities. I’m sure that Nick Ferrari thought he was betting on a certainty, unaware that the result was about to get fixed.
Matt’s hesitation in accepting the bet was probably because he knew full well that he wasn’t going to be able to meet that target. Let’s go back to the exact wording of the target he was betting on. In a briefing on 2nd April, he stated that “I’m now setting the goal of 100,000 tests per day by the end of this month”. “Per day” implies that it’s not a target just for the last day of April, but would need to be met on at least one previous day as well. As April progressed it became more and more apparent that that wasn’t going to happen. However, on the 1st of May he proudly proclaimed that on the 30th April there had been 122, 347 tests, claiming that he’d met the target.
It subsequently emerged that only 73,191 people had been tested. The figure included 12,872 test kits which had been delivered to hospitals and 27,497 home testing kits which had been posted out. It appears that the latter were randomly sent to council employees, regardless of whether or not they had reported symptoms. It was a fudge to hit the target. You may have noticed that the figures above don’t add up. That’s thought to be because around 12% of the 73,191 people tested were tested twice. So, Matt failed on two counts – managing to hit 100,000 for at least two days, and actually performing 100,000 test on any one day. It was the classic case of fixing a result.
Is Matt on trial, or banned from any further political events? No, he’s not. Throughout this pandemic, the Tory Government has highlighted the dedication of the NHS, providing a useful smokescreen to obscure the lack of real policies. That public support is a welcome change from the accusations that used to be levelled at the Conservatives, which were that they were working to privatise it or sell it off. Matt Hancock’s actions suggests that the reality is even worse. Here we have a minister potentially prepared to lose the health service over a bet for Matron’s used PPE.
On the plus side, Matt’s got an app.
I’ve written before about the shortcoming of the contact tracing app which is meant to solve our Covid-19 woes. It uses Bluetooth to work out who you might have been in contact with. If any of the contacts it records have tested positive, it tells you that you need to self-isolate. All of this personal data is stored in a central, Government server. The app has been put together by companies who seem to know little about the way that Bluetooth works in phones, and has ignored work being done by Apple and Google to provide a secure framework that preserves user’s anonymity. It’s been tested on the Isle of Wight for the last month. It’s probably an indication of their lack of faith that the app has not been rolled out for testing to Government ministers and advisors, especially given how many have contracted Covid-19. If it had been, they might have been forewarned about Dominic Cummings’ attempt to spread it to Durham.
A number of issues have been raised from the Isle of Wight trial, which have delayed its deployment to the wider population. Matt Hancock appears to be distancing himself from it by appointing Baroness Dido (who sits on the board of the Jockey Club and obviously knows a bit about betting) to lead the programme of track and trace in the UK.
One of the major criticisms of the app is its poor security, with personal data being held by the Government for up to two years. The Open Rights Group is currently preparing a legal challenge over the amount of data collected and retained. On the Government side, Baroness Dido knows a bit about security, or rather the lack of it. She was at the helm of TalkTalk when they suffered a cyber attack that stole the personal details of four million users. When asked whether the data was encrypted, she replied “The awful truth is that I don’t know”, leading Marketing magazine to run a headline, “TalkTalk boss Dido Harding’s utter ignorance is a lesson to us all”. Asked about the NHS trace and track application in a Covid-19 briefing session this week she suggested that, “I have repeatedly said this is the cherry on the cake, not the cake itself. And what you’re seeing today is the baking of the cake is going reasonably well.” Rather than spending the time after she left TalkTalk brushing up on cyber security, it sounds as if she’s been watching the repeats of Bake-Off. The cake she refers to is the manual track and trace system, which was eloquently explained by Private Eye as:
- Someone you don’t know (who may be a scammer) phones you up, telling you to self-isolate and asking who you’ve been in contact with.
- The person being contacted includes the names of a number of people they don’t like, forcing them to go into two-week quarantine.
- That’s it.
In previous articles I explained that a contact tracing app is unlikely to work if we are expected to get back to using public transport, as the scenario above rapidly results in another lockdown, but facts like that, along with the practical experience in other countries, tends to be ignored by people like Matt Hancock, who think that shiny tech is a quick and easy way to solve problems. Singapore did a much better job with their TraceTogether app, but they’ve found that not enough people are using it to make it effective. That’s largely because it doesn’t work reliably on iPhones or older Android phones, so only 25% of the population is using it. It needs it to be used by around 60% of the population to be effective. As a result, Singapore is now looking at a contact tracing wristband, which also grants access to building and public transport, which may be mandatory for everyone to wear. They may find a copyright writ heading their way from Ira Levin’s estate, as that’s exactly the plot of his sci-fi novel “This Perfect Day”, published exactly fifty years ago. As we enter this debate on tracking citizens and the privacy of personal data, it’s a timely read. (Spoiler alert: It doesn’t end well.) But it’s not all bad news. The Government has been handing out contracts for developing track and trace apps, at developer rates that make your eyes water. Whether they will ever be deployed is debatable. If they continue getting awarded at the current rate, it might make sense for the Treasury to consider developing a contract tracing app, before young Matty gambles away the rest of the nation’s health wealth.
Any comments on the latest range determination
https://covid19.nhs.uk/risk-scoring-algorithm.html
Close < 2m
Medium 4m
Samples every 3.6S
They make total sense. Throughout the current app development, it appears to have been led by developers with no understanding of the physics of how radio works (I believe this has been a concern raised by members of SAGE, who do). For contact tracing to work, you need access to low level aspects of Bluetooth, which manufacturers don’t allow, as it can break Bluetooth and negate their radio qualification. Hence you need to involve the phone vendors. That’s what Apple and Google are doing.
That’s the minimum to allow it to work reliably on Apple and Android phones. However, there are others issues which I’ve covered in previous articles, particularly that of using it in crowded places, such as public transport or the adjacent queues outside my local Primark and M&S. Both would result in almost everyone being tagged as a contact and sent into self-isolation. So, what may work on the Isle of Wight (if properly designed) probably won’t work in London.
Any comments on the very recent Apple & Google ‘framework’ updates in preparation for the App