New nhs app

[Lindarose]

Hi all
I’ve been reading about the new app being considered to track people who have been either in contact with or have symptoms of Covid.
This bit I am confused by -
People who have been in sustained proximity with someone who may have Covid-19 could then be warned and advised to self–isolate, without revealing the identity of the infected individual.
Does this mean nhs workers working on wards with positive cases have to self isolate?
I know it’s not even happened yet but wondering. Just in case!
Maybe I’m misreading things so interested in any views.

 

[grovesy]

I would not trust, to be honest, my GPS on my Garmin Fitness just told me I had done lap but I was not near where i started my walk. My husbands location Tracker just said he was home and he was at the shops.

 

[mikeydt1]

without community testing this will be a total waste of space to anyone who wants to use it. it requires user input from the person using and if they haven't even been tested then it could cause issues for the person concerned.

 

[Bruce Stephens]

Yes, it looks like somethingism: they can't seem to do the useful things (mass testing, contact tracing, supplying basic stuff to front line staff) but they realise they need to be seen to be doing something and a smartphone app is something and maybe a bit sexy (even if useless without the actually useful things).

Contact Tracing COVID-19 Infections via Smartphone Apps - Schneier on Security

Google and Apple have announced a joint project to create a privacy-preserving COVID-19 contact tracing app. (Details, such as we have them, are here.) It’s similar to the app being developed at MIT, and similar to others being described and developed elsewhere. It’s nice seeing the privacy...

www.schneier.com

 

[Ralph-YK]

I would not trust, to be honest, my GPS on my Garmin Fitness just told me I had done lap but I was not near where i started my walk. My husbands location Tracker just said he was home and he was at the shops.

I think there was something about bluetooth, and the phones being able to sense each other. Though I quess this means the person who's been infected, that people are being warned about, needs the app on their phone.

 

[Bruce Stephens]

I think there was something about bluetooth, and the phones being able to sense each other. Though I quess this means the person who's been infected, that people are being warned about, needs the app on their phone.

Yes, as far as I understand it (I haven't read any of the designs for this very carefully) it's Bluetooth LE based. Bluetooth LE because that's low power and fairly low range. (Though it'll go through thin walls, floors, etc.)

Some more details (of the Google/Apple proposals): https://www.theverge.com/2020/4/10/...-tracing-bluetooth-location-tracking-data-app, https://www.fivestars.blog/code/how-covid-19-contact-tracing-works.html

I presume whatever the NHS app does will be similar (if not identical with) what Google/Apple are already doing. Seems daft if not, since they're going to be releasing it in their operating systems.

 

[Andy HB]

I must admit to being very skeptical about this app. If a person knows or suspects they have the virus, why are they walking around potentially infecting others? I agree that it is just a bit of "somethingism".

 

[trophywench]

I'd be instantly afraid of Bluetooth for the simple reason that's how my pump and its handset communicate with each other and I'd want to know immediately before I even considered such a thing that it wouldn't interfere with el pumpo. We DO have to turn the Bluetooth off as we pass the boarding gate at an airpor tas it is.

 

[Bruce Stephens]

I'd be instantly afraid of Bluetooth for the simple reason that's how my pump and its handset communicate with each other and I'd want to know immediately before I even considered such a thing that it wouldn't interfere with el pumpo.

I presume they've tested that! Presuming you walk around with it in crowds (well, used to do so, anyway) you'll have encountered lots of bluetooth signals, and not just the LE that they're going to be using.

 

[Bruce Stephens]

Here's a story covering the various applications that have been used together with the Apple/Google one:

So how do the coronavirus smartphone tracking apps actually work and should you download one to help?

Bluetooth, GPS, cell towers, code scanning: what’s the best way?

www.theregister.co.uk

 

[mikeyB]

As that article says, many apps - weather, my central heating controller, the phone itself as examples. When I’m on my mobility scooter, my phone tells me that it has switched off alerts while I’m driving. Your phone constantly pings the nearest mobile phone tower to get the best signal if you want to make a call, or receive one. That’s how cops can tell if you’re telling the truth about your location. It’s also why drug dealers throw away cheap phones on a daily basis.

We are passively watched all the time, so adding another passive app is just a way of focussing that to reveal interactions between phones. I’d rather not install such an app. Do you really want to reveal everything you do to the government? It’s the thin end of a wedge to those of a totalitarian mindset. Viktor Orbán would love it.😉

 

[Bruce Stephens]

Do you really want to reveal everything you do to the government?

Which is why the Apple/Google approach is trying so hard to work usefully without needing to release that information to anyone. They're not using location at all, and are trying to hide identifying information as much as they can (so it's just individual phones that have the necessary information to work out what's needed to advise that user).

Something like what the other countries have been doing would be much simpler, but would involve lots more fine-grained location information and things being kept centrally (and I think it's likely that governments would be slow to give up that).

I'm still doubtful that enough people have suitable phones that would make this particularly useful, even if we pretended our government was going to increase testing sufficiently for it to be really helpful.

 

[Ralph-YK]

(Though it'll go through thin walls, floors, etc.)

I've a small Bluetooth speaker that'll work in different parts of my flat to where my tablet is. Through walls and hall. Don't know if it would work next door. That's made me think. Would the system highlight people who haven't met. Because they were on opposite sides of a wall ect.

I'd be instantly afraid of Bluetooth for the simple reason that's how my pump and its handset communicate with each other

My defib can be read and controlled remotely. (Hospital will pair up a machine when I visit, and get info from the defib, reprogram & do tests.)
Defib came with a warning not to use a phone on the same (left) side of my body. There was advice to keep away from magnets, and not hand around security systems (like in shops).

 

[Bruce Stephens]

Would the system highlight people who haven't met.

Yes, almost certainly. Similarly, systems where you have to scan a QR code every time you go on public transport, go in a restaurant, etc., (as used in one of the countries) is going to connect people who really didn't connect meaningfully. Such things might still be useful (they'd need to look at the modelling to see what error rates would make it not useful).

For that matter one of the really annoying catchphrases they're using is "No test is better than a bad test.". All tests have false positives and false negatives. If you're insistent on getting tests that have a zero error rate then you're going to have problems sourcing such tests.

Obviously if you tell someone a test shows they definitely, for sure, aren't infected (or have had the infection and are now immune), you'd better be reasonably sure that's right. But being able to tell someone they're probably infected and should quarantine themselves for 14 days is fine even if there's a pretty high chance that's wrong. (And we're admitting that by not even bothering to test such people.)

And in some contexts, giving someone a negative test result even with a not that accurate test would be fine. (There's modelling suggesting if you could do 22 million tests a day in the US (so testing the entire population every couple of weeks) even if your test isn't that good that would still be helpful. Even if you tell an infected person they're clear, you'll most likely test them as infected in 14 days if they still are.)