The NHS COVID-19 app

[JGD]

The NHS COVID-19 app is launching 24 September.

The new NHS COVID-19 app, available to download for free in England and Wales from 24 September, is the fastest way to see if you're at risk from coronavirus. The faster you know, the quicker you can alert and protect your loved ones and community.

The app has a number of tools to protect you, including contact tracing, local area alerts and venue check-in. It uses proven technology from Apple and Google, designed to protect every user’s privacy.

See details on this NHS site:

https://covid19.nhs.uk/

This link takes you directly to an explanatory video:

https://youtu.be/fp0kHcr_Pi4

ETA

If you choose to go directly to Apple's App Store, you can search for:

NHS COVID-19

..and download the app directly to your iPhone.

It requires the latest IoS version.

This download worked as at 23:45 on 23 September.

[stuart]

And now we are two ...

Stuart

[Parker1970]

I downloaded and installed it this morning when I saw it on BBC news.

[Robin Orton]

I'm temperamentally a late adopter. It sounds a bit complicated to me. I gather it includes a 'QR code', a term that always causes my bowels to turn to water. I shall wait a bit and see what others have to say about it.

[JGD]

Dinnae worra yer heart, Robin.

You do not need to have a QR code, or wear one or give one to anyone else. Or affix one to your face mask or safety hat (tin-foil models not recommended), if you are so inclined as to wear one.

The app, once set up on your device sits in the background, silently scanning other devices and the environment around you.

It will only alert you if something changes - like a change in the area status - or if your device is notified you have been in contact with a person who has been tested as a positive case.

The QR code ONLY comes into play when it is on display in a public premise, library, community centre, post office, pub, restaurant, council office, gym, vets et al in such circumstances as you find yourself in close proximity to people not in your bubble in enclosed rooms or buildings.

Even then you do need not do anything more than use your app (which uses your devices camera) to scan that code - nothing else.

Of course, if it is your practice in this pandemic not to visit such places you will never encounter or have to point your device at a QR code.

[chrisj1948]

I downloaded the app yesterday morning. It seems simple. Since I don't visit pubs or restaurants all I have to do is notify it if I get Covid 19, or wait for it to tell me that I am at risk.

[stuart]

Goodness me. Battery consumption only 0.38% - impressive.
Though I do have other Bluetooth connections running in parallel.
[Android 10]

Stuart

[stuart]

I've copied this post from Clausy @ se23.life 'cos I think it deserves the widest possible circulation. And to prove some geeks can communicate with style and clarity!

There’s a great Facebook post from the guy who worked on the team that built it: (David Bonsall) - I’ll paste it in here (hope that’s OK) rather than linking to Facebook as some people might not have that.

Last 24 hrs I’ve provided the polite version of the following to C4, ITV,BBC, Radio 5, Sky News, Times Radio, LBC. I thought I’d answer a few frequently asked questions more directly, while our communications officer isn’t watching.

Is it going to steal my data?

No its completely private. All your data stays on the phone, if you delete the app, you delete the data. Facebook, on the other hand…
Is it going to track me?

No it doesn’t know where you are. Uber, Dominos pizza, google maps, on the other hand…
Is Dominic Cummings…

No
Why has it taken so long?

It was hard. We tested it, made it better, tested it again, made it better, now it’s ready. And we helped Google develop their system. And we beat most states in America
How many people need to download it to work?

Two. If you download it and I download it and I get infected and I notify you and you don’t visit your gran, Granny lives
What happens if we all download it?

Maths shows it could stop the epidemic.
But Dominic Cummings…

Just no
Why did the Isle of Wight app not work?

Actually it did. It looks like it stopped their epidemic. There were problems with iPhones.
Are there not still problems with iPhones

Yes. The few people with iphones earlier than 6S should write Apple and complain. If you have a later version you should download the app, stop the epidemic, and save the old folks with outdated smartphones.
What about the fines for not following the notifications? Those aren’t fair!

What about the local outbreak and deaths you could cause by not following them? also what about the £500 you receive if you get notified and can’t work from home?
Yeah but “False Positives”…

The only way you can get a notification is by being close to someone who tests positive for COVID-19 for a prolonged period.
Yeah but its not 100% accurate at measuring distance so there are false positives

Do you think the virus cares if it’s 2 meters or 2.3 meters?
Yeah but…

-If I asked you to remember all your 2meter 15minute contacts in the last 3 days would you give me an accurate answer? Where’s your tape measure and your stop watch?

So is it more accurate than manual contact tracing?

-Yes probably. And it can notify people you don’t know.

How else is it better?

well it’s a lot more private - you don’t need to give personal details about you or anyone else
Anything else?

it uses computers so it’s fast enough to find people BEFORE they infect others…which is sort of the point.
Anything else?

it doesn’t require millions of netflix subscriptions.
Anything else?

It’s a quick way to check in to bars (I like bars) and it’s a faster way to book a test without navigating a website designed by a Vogon
There are no tests?

Fair point. Not my department. But will keep the pressure on.
Its not a silver bullet is it? It’s just a cherry on the cake?

No. If we all download it, it’s A GOLD-PLATED MISSILE, which relatively speaking cost one-thousandth of a cherries worth of cake, compared to another lockdown.
Is it going to work?

not unless you download it…give it a go. It’s really rather good.
(And well done to our group at Oxford, Alan Turing Institute, department of health, NHSX, Google, Pivotal, Zuhlke, and beyond who’ve all contributed…it’s been a long time coming…and it’s something to be proud of. Well done!)

If we can keep the infection rate of this app well ahead of COVID-19 it could help lower the tipping point.

Stuart

[JGD]

Goodness me. Battery consumption only 0.38% - impressive.
Though I do have other Bluetooth connections running in parallel.
[Android 10]

Stuart

BH and I both running app on iPhone 7 Plus models. Only significant difference in our movements and circulation is that she attends school every work-day - but generally adheres to bubble rules there.

The iPhones report the apps battery use for her is at 1% whereas mine is 12%.

Intriguing difference for identical kit.

Clausy's piece is a good one - the style used by the author makes it not only readable and understandable - but it also brings a grin to your face.

[Nic]

Ummm, slight problem:

"If your test took place in a Public Health England lab or NHS hospital, or as part of national surveillance testing conducted by the Office for National Statistics, test results cannot currently be linked with the app whether they’re positive or negative. We are working to make this available as soon as possible."

Source: https://faq.covid19.nhs.uk/article/KA-01307/en-us

[JGD]

There are also divergent comments on this point too.

The BBC has it that only "Negative" test results cannot be input.

https://www.bbc.co.uk/news/technology-54307526

Is it possible that if the recording of "negative" results is deemed to be "No further Action Required" condition within the apps design? Because there is nothing to be done with the notification and nothing to alert other users about perhaps?

If, as Nic's link reports, that the source whereby the test was initiated governs whether ANY result, negative or positive, dictates whether the app will accept or process the input seems to be an evident and glaring gap.

Put strongly - a basic fundament has not been delivered and the effectiveness is diminished.

As an example, schools each have a stock of 10 emergency home test packs. If the school issues one, that means the app knows nothing of its issue or use - how are positive results to be input into the model?

It is also evident that the source of test results must be validated. People must be prevented from entering random data that may or may not be authentic thereby resulting in pyramids of users getting false alerts.

The country is pleading for consistency and joined up thinking - it is essential.

[Nic]

It is also evident that the source of test results must be validated. People must be prevented from entering random data...

To link your test result to the app, you need to enter a code provided by the testing service apparently.

[JGD]

To link your test result to the app, you need to enter a code provided by the testing service apparently.

Let us hope that we can put our faith in that fact and that ALL testing services will issue a validation code that can be input into the app.

If that validation link to every testing services is absent at this point in time - let those who make things happen, make the additions and modifications urgently.

[Nic]

https://twitter.com/SkyNews/status/1309 ... 09992?s=20

[JGD]

Further evidence that Nic's link appears to have been very accurate as can be seen in a tweet from the official @NHSCOVID19app Twitter account.



Note the sign-off - "Thanks".

Not "Sorry ".

Smacks of kiss-off doesn't it.

[stuart]

I don't think this shortcoming is as serious as it appears at first sight.

The ONS tests are random and hence will only have a very small hit rate. Much smaller than when COVID-19 is suspected which is covered by pillar 1 (PHE/NHS) and pillar 2 (Independent labs). Pillar 2 are OK with the App - which is around 67% of Pillar 1/2. Of the other 33% - those tested positive in hospitals will, almost by definition, not be wandering the streets triggering (or not triggering) your app.

So a back of the envelope guess is that the app would react to around 80% of infected people who tried to enter a positive result. Given that 'perfect is the enemy of good' that's good enough to get the app out and about before the other tests can be linked. Delay will be fatal. I don't think it detracts from the urgency of using the app if you can and hence I do think the conclusion of the previous post is unfortunate.

But I'm puzzled about why one should enter a negative test.

Stuart

EDIT: This limitation was announced with the App. So no secrets there.

[JGD]

Is it possible that if the recording of "negative" results is deemed to be "No further Action Required" condition within the apps design? Because there is nothing to be done with the notification and nothing to alert other users about perhaps?

But I'm puzzled about why one should enter a negative test.

Stuart

Whilst not conclusive - logic (oh no, I hear you say) might appear to be at work - it is not essential that negative results be input.

[JGD]

Besides whilst we quietly and fairly constructively debate these points, here is another group with an entirely different outlook.

"None of this for us" appears to be their motto.

https://www.bbc.co.uk/news/uk-54309603

Soon to be candidates for inclusion in the Darwin Awards perhaps?

https://darwinawards.com/darwin/

[JGD]

Announcment on Sky News - great news if this gap has been fixed.

Coronavirus: NHS tracing app problem that left tens of thousands of tests unlogged has been fixed, government says

Can it be fact-checked quickly?

https://news.sky.com/story/coronavirus- ... s-12082178

[JGD]

So a back of the envelope guess is that the app would react to around 80% of infected people who tried to enter a positive result. Given that 'perfect is the enemy of good' that's good enough to get the app out and about before the other tests can be linked.

It came after the app's developers admitted it had not been able to link more than 60,000 coronavirus tests carried out in England on Friday - just under a third of the total - to its systems.

Once more, there is divergence in the reports on the matter of the proportion of results that could not be processed by the app before it was "fixed".

29% of tests (not reported specifically to be positive) does seem to be a significant proportion however.

It is to be hoped that this very speedily produced "fix" addresses the problem as comprehensively as is implied.

On a side note - it is good to see that the safer of our local establishments are early adopters and have the "Venue Check-in" QR on display.