The Pingdemic

The newspapers over the last few days have been full of stories about the Pingdemic: the number of people who are being pinged by the "NHS"* Track and Trace app and told to isolate as they've been in contact with someone later tasted positive with Covid19. A couple of samples are this Guardian story about holidays being cancelled (link) and today's Guardian story about empty supermarket shelves because delivery drivers are isolating (link). There are loud cries in the tabloid press that the app should be made less sensitive as it's obviously a problem - see the BBC coverage of today's papers (link).

Question: if the app is telling people to isolate as they have been in contact with Covid19, is that not the app doing its job, for the first time since it was established? And isn't this an indication that we are currently in the next wave, third? fourth? fifth? of the virus, and really should be shutting down again, not opening up regardless and ignoring the pings warning us of danger?


* in inverted commas because it's called the NHS track and trace app, but it's an initiative of this Government and just called that to try and encourage people to trust it.

In full disclosure, I don't have the app, partly as we've been shielding forever, and partly I don't have blue tooth switched on on my phone so it wouldn't work. Partly battery drain, mostly I decided blue tooth was a bad idea when commuting by tube, because I didn't want to broadcast to a carriageful of commuters.
«1

Comments

  • chrisstileschrisstiles Shipmate
    Yes, the 'pingdemic' is entirely the consequence of a series of actions that led to a large number of infections in the population.
  • KarlLBKarlLB Shipmate
    *Test and trace

    And yes, this was inevitable. Since the government opened up on the basis that hospital admissions and deaths were being suppressed by vaccination, while infections were still high, it follows that you are going to see a massive increase in infections. Which increases the number of people who are going to have had contact with someone with a positive test.
  • Apparently a large number of Co-Op stores are affected.

    I've not braved either of our two local Co-Ops this week (being keen to keep out of the way of maskless covidiots), and this news gives me even less incentive to visit!
  • Gill HGill H Shipmate
    Hugal is self-isolating as he got pinged. We know when it happened - we were sitting outside a coffee shop and he went inside to order. He checked in on the app, placed his order and was outside again in less than a minute. We drank our coffee outside so I didn’t check in as we were nowhere near anyone else.

    There is no other place we have been where he was the only one to check in - so we figure it must have been there.

    However, prior to that he was on a train for 90 minutes a day, and at work in a department store for 8 hours a day. Surely higher risk, but no opportunity to check in.

    This is my gripe with the app. It doesn’t seem fit for purpose.
  • BoogieBoogie Shipmate
    @Gill H, it’s not just about checking in. The app logs all nearby app users and pings you if any of them test positive within a certain period.
  • Fawkes CatFawkes Cat Shipmate
    Once upon a time I was an IT type (but technology and me went our separate ways the best part of twenty years ago) but this sounds very much to me like the app is working as designed - which from an IT perspective is not normally considered to be a bad thing.

    If there is a scientific, economic or political reason why less sensitivity is needed then it's for the scientists/economists/politicians i.e. the customers who asked for the app to ask for a change to be made - and if they don't want a change, then it's for them to explain to the end-users why no change is happening.
  • chrisstileschrisstiles Shipmate
    Gill H wrote: »
    However, prior to that he was on a train for 90 minutes a day, and at work in a department store for 8 hours a day. Surely higher risk, but no opportunity to check in.

    This is my gripe with the app. It doesn’t seem fit for purpose.

    As @Boogie says , assuming he had the app active during this time, then the ping could also be due to a contact with someone who tested positive during his commute or workday.

    In which case the app would be entirely fit for the purpose for which it has been used.

    Which leads to the point that @Fawkes Cat alludes to - you can't solve a political/economic/public health (delete as appropriate) issue purely with technology.
  • @chrisstiles - and everyone else on the train had the app enabled at that time.

    Interestingly, in a conversation last night I said that I didn't have the app because we'd been shielding and also I have blue tooth switched off so it wouldn't work. To which one of the people I was talking to realised that although they might have the app, they too had disabled it as they also had blue tooth switched off.
  • BoogieBoogie Shipmate
    Thoughts of a past health minister - in the Guardian today.

    https://tinyurl.com/twyhzt7y

    He has a point.
  • chrisstileschrisstiles Shipmate
    @chrisstiles - and everyone else on the train had the app enabled at that time.

    Well, or one other person who received a positive test also had the app enabled.
  • ThunderBunkThunderBunk Shipmate
    This is the effect of the Euro 2020 experiment. Something of a failure I would say. I wonder whose bright idea that was......
  • PomonaPomona Shipmate
    If it was over-sensitive then surely it would be pinging people who couldn't possibly have been in contact with someone who tested positive? I have basically not left the house since the hot weather started and haven't been pinged at all (and wasn't getting pinged before that either), which to me suggests that it's not just randomly glitching pings.
  • ISTM that this is a consequence of having inconsistent policy. Either you open up, acknowledge that infections will rise, and no longer try to do contact tracing, or you keep restrictions in place, keep contract tracing and try to keep infections down. I think either policy could be defended, but it makes no sense to have half of each.
  • orfeoorfeo Shipmate
    edited July 22
    There are loud cries in the tabloid press that the app should be made less sensitive as it's obviously a problem

    This is really no different to Trump's complaint that if you test too much you find too many cases.

    The fundamental problem is going in with a mindset about the answers you'd like to see or are prepared to accept, and then if you don't get an answer within an acceptable range then the method must be wrong.

    This sort of thinking is hardly unique to COVID-19, it's just a lot more serious than in some other situations.

    You CAN of course change the methodology, so long as you understand that this means you're creating a greater risk of infections spreading. A risk you can reduce by having everyone not physically move around so much...

    Here in Australia the most recent complaint was actually the exact opposite - that the federal government's app (which in truth has been almost completely superseded by the app in each state anyway) wasn't sensitive enough for the Delta strain because it still only responded to around 15 minutes of proximity.

  • chrisstileschrisstiles Shipmate
    Pomona wrote: »
    If it was over-sensitive then surely it would be pinging people who couldn't possibly have been in contact with someone who tested positive?

    If it was over-sensitive (as opposed to randomly glitching) it would be picking up people who were much further away, or in situations where people couldn't infect each other - for instance neighbours on either side of a wall, or people sat in adjacent cars in traffic. That this doesn't appear to be happening suggests that it is largely working correctly (the proximity functions are provided by Apple/Google rather than the app itself), and the 'excess' pinging is because we went into the unlocking phase during a time when the number of infections was high.
  • Gracious RebelGracious Rebel Shipmate
    edited July 22
    But apparently there have been cases where the pings were because of somebody in a neighbouring house, behind a wall.

    I confess I have been a bit of a coward over this....while I still use the app to check into venues (including church!) I turned off the proximity checking feature at the start of the pingdemic. Figured it still leaves me in a better position than not having the app, and avoids any moral dilemmas about reacting to a ping.

    I have heard that many people have quietly deleted the app completely now.
  • Marvin the MartianMarvin the Martian Admin Emeritus
    I thought the problem was that fully vaccinated people (who one would assume can be in proximity to the infected with impunity, or else what's the point of a vaccine in the first place) are still being told to isolate.
  • orfeoorfeo Shipmate
    edited July 22
    I thought the problem was that fully vaccinated people (who one would assume can be in proximity to the infected with impunity, or else what's the point of a vaccine in the first place) are still being told to isolate.

    It's not impunity. No vaccine is completely effective. It's reduced risk, which is not the same thing.

    See also: wearing a seatbelt is not a complete guarantee of not being killed in a car crash. Do you therefore say "what's the point of a seatbelt"?

    No one person's vaccination guarantees that person freedom of infection. But if most of a population is vaccinated, the chance of an infection continuing to spread through that population is greatly reduced. The real question is not what's the point of your individual vaccination, the question is what's the point of having as many people as vaccinated as possible.
  • orfeoorfeo Shipmate
    edited July 22

    A classic illustration of measuring the wrong thing. 61% of a small number is still less than 5% of a huge number.

    If you ever get to the point that 100% of the population is fully vaccinated, the percentage figure will "rise" to 100%... probably of a very small total in comparison.

    Lies, damned lies and statistics, etc etc.
  • BroJamesBroJames Purgatory Host, 8th Day Host
    True, but it’s important to read the whole article
    However, this does not mean that the vaccines do not work, or are dangerous.

    None of the current vaccines are 100% effective against Covid-19, but they do make a big difference. The same PHE briefing also reported that two doses of the vaccine are 96% effective against hospitalisation with the Delta variant.

    The UK’s vaccination programme has prioritised ensuring that the elderly and vulnerable were vaccinated first, meaning these people are also the most likely to have received their second vaccine dose. However, this group is also the most at risk from Covid-19, and no current vaccine removes all risk.

    Writing for The Guardian, statisticians Professor Sir David Spiegelhalter and Anthony Masters said that someone aged 80 who is fully vaccinated essentially is at the same risk from Covid-19 as an unvaccinated person of around 50.
  • At that time we had about 50% of adults vaccinated.
  • orfeoorfeo Shipmate
    edited July 22
    BroJames wrote: »
    True, but it’s important to read the whole article

    I did. I'm agreeing with them that the quoted statistic is completely misleading. It's misleading for at least 2 reasons, and I highlighted one of them.

  • KarlLBKarlLB Shipmate
    At that time we had about 50% of adults vaccinated.

    But most of the deaths were in the higher age groups where the proportion of vaccinated adults was much higher.
  • orfeoorfeo Shipmate
    Indeed, if the vaccination amongst older people was HIGHER than 61%, then that would mean that vaccinated people are underrepresented in deaths.
  • Marvin the MartianMarvin the Martian Admin Emeritus
    orfeo wrote: »
    See also: wearing a seatbelt is not a complete guarantee of not being killed in a car crash. Do you therefore say "what's the point of a seatbelt"?

    If there were people saying that seatbelts aren't a 100% guarantee against dying in an accident and therefore nobody should be allowed to drive faster than 5mph whether they have a seatbelt or not, then you're darn right I'd be asking what exactly they think the point of the seatbelt is.
  • RicardusRicardus Shipmate
    The proximity sensor has always worked on a balance of risks. IIRC you have to be in close contact for 15 minutes before you are pinged. That doesn't mean that there is zero risk if you are in contact for less than 15 minutes, it's just that that lower risk is considered outweighed by other factors.

    If the risk of a double-jabbed person becoming infected after 15 minutes' close contact is similar to the risk of an unvaccinated person becoming infected after 5 minutes' close contact, then prima facie it seems reasonable to treat them the same, i.e., no ping.

    (And yes, that is for policy-makers to decide, it's not something the app can magically decide by itself.)
  • Leorning CnihtLeorning Cniht Shipmate
    edited July 22
    Pomona wrote: »
    If it was over-sensitive then surely it would be pinging people who couldn't possibly have been in contact with someone who tested positive?

    I think in this context, the claim of "over-sensitivity" is a claim that you're being pinged for being 20 or 30 feet away from an infected person, and that might be over-cautious.

    But look at the rates. The UK currently has 45,000 new cases per day - almost as many as the Christmas peak. That's 0.1% of the population per day being infected. Assuming people are in public for a few days, on average, whilst they're infectious and before they're pinged, or fail a Covid test - so 0.5% of the people you meet are infectious. If you ride a bus with 50 people on it, you've got to do that four times on average before you've shared a bus ride with an infected person.

    I thought the problem was that fully vaccinated people (who one would assume can be in proximity to the infected with impunity, or else what's the point of a vaccine in the first place) are still being told to isolate.

    @Marvin the Martian - have you read anything at all about either Covid or vaccines in the last 18 months? Because this comment suggests that you haven't.

    The Pfizer vaccine appears to be 88% effective against symptomatic disease from the Delta variant, and 94% effective against symptomatic disease from the Alpha variant. For the AstraZeneca vaccine (a significant fraction of UK jabs), the figures are 67% and 75% respectively.

    For both vaccines, the effectiveness against severe symptoms is quite a bit higher, and the effectiveness against being asymptomatic but transmitting the virus is somewhat lower (I don't have great numbers for that, but perhaps someone else does.)

    Assuming that a vaccinated person can "be in proximity to an infected person with impunity" is stupid.

    What the vaccine does for you is reduce your risk. The vaccine is why although case rates are about as high as they were at Christmas, death rates are less than 10% of the Christmas numbers. If you have had the AstraZeneca vaccine, then if you are exposed to someone with Delta-variant virus, you are about one-third as likely to develop some Covid symptoms as you would be had you not been vaccinated, and your symptoms will almost certainly be significantly milder than they would otherwise have been.

    We know that before people started to take any anti-Covid measures, R was somewhere around 5. We know that the Delta variant is more than twice as infectious, which would suggest that it should have R of about 10 under the same conditions. If R is 10, then to reduce effective R to 1, you need a 100% of people to have a vaccine that is 90% effective against transmission. The vaccines aren't as good as that, and not everyone is vaccinated, so you must also use other methods to reduce R (masks, maintaining distance, limiting contact with people, ...)

    This really isn't complicated.
  • RicardusRicardus Shipmate
    We know that before people started to take any anti-Covid measures, R was somewhere around 5.

    Do we? This BMJ article (link) gives the R0 for original Covid as a bit under 3. This BBC article (link, last paragraph) also says 3. This would make your final numbers come to 60% efficiency with 100% takeup.
  • DoublethinkDoublethink Shipmate
    I don't think turning off Bluetooth solves a moral dilemma, I think it is just making a poor choice.
  • Ricardus wrote: »
    We know that before people started to take any anti-Covid measures, R was somewhere around 5.

    Do we? This BMJ article (link) gives the R0 for original Covid as a bit under 3. This BBC article (link, last paragraph) also says 3. This would make your final numbers come to 60% efficiency with 100% takeup.

    Yes, it would. R is strongly behaviour-dependent: I think my memory of 5-ish might have come from an early Chinese study, so perhaps with more crowding than a UK average.

    In England, currently just over 30 million people are fully vaccinated, which is about 54% of the population.
  • TelfordTelford Shipmate
    edited July 22
    Have all these people died due to covid or was the main cause of death something else and they just tested positive for covid.
  • Telford wrote: »
    Have all these people died due to covid or was the main cause of death something else and they just tested positive for covid.

    Oh, not this again!

    I can't tell you that there isn't someone who died of some other cause and got counted as a Covid death, but I can tell you that any such people are statistically irrelevant. We are not overrun by people who happen to get hit by a bus whilst infected with Covid-19.

    If you're talking about elderly, infirm people who might have multiple contributing conditions, then yes - you probably count all of them, but you probably count all of them because Covid contributed to their death. Take a person in poor health, and give them some kind of severe respiratory illness. It's going to make things worse. It's going to make them die faster.

    More than half of people killed in car accidents are wearing seatbelts. You understand, I'm sure, that this doesn't mean that seatbelts don't work, or are a bad idea.

    If you think of the vaccine like a seatbelt, you're not far wrong. It makes things better. It reduces the severity of an incident you might be involved in. But if you put on a seatbelt, and then drive like an idiot, you can still kill yourself.
  • Curiosity killedCuriosity killed Shipmate
    edited July 22
    I suspect that @Telford has heard Lord Sumption's comments on Radio 4 and other media in the last week - Full Fact checked his statistics here (link).
    The daily data on the number of people who have died after a positive test does include some people who died for other reasons. However, we also have data from death certificates, which records whether or not Covid itself was the “underlying cause”.

    This shows that up to 2 July this year, 124,082 people died with Covid as the underlying cause of death in England and Wales alone.

    Other things said by Lord Sumption that were refuted were that (2) people would have died anyway in the next year - in fact people who caught Covid19 on average lost 10 years of life, and that (3) the people who died had comorbidities, which again is a tiny effect.

    The other thing Lord Sumption omitted in his rosy picture was Long Covid - another 14% of those who contract Covid19 are suffering with the effects over 12 weeks later, many of which are debilitating or life changing effects - link to ONS study.

    The other thing that isn't being reported widely is that over a third of people who were discharged from hospital as recovered are returning to hospital within 140 days seriously ill with long term effects of Covid19 and another 10% died after discharge -link to BMJ article, so the 28 day figure is undercounting all these people who die of the effects of Covid19.

  • TelfordTelford Shipmate
    Telford wrote: »
    Have all these people died due to covid or was the main cause of death something else and they just tested positive for covid.

    Oh, not this again!

    I can't tell you that there isn't someone who died of some other cause and got counted as a Covid death, but I can tell you that any such people are statistically irrelevant. We are not overrun by people who happen to get hit by a bus whilst infected with Covid-19.

    If you're talking about elderly, infirm people who might have multiple contributing conditions, then yes - you probably count all of them, but you probably count all of them because Covid contributed to their death. Take a person in poor health, and give them some kind of severe respiratory illness. It's going to make things worse. It's going to make them die faster.

    More than half of people killed in car accidents are wearing seatbelts. You understand, I'm sure, that this doesn't mean that seatbelts don't work, or are a bad idea.

    If you think of the vaccine like a seatbelt, you're not far wrong. It makes things better. It reduces the severity of an incident you might be involved in. But if you put on a seatbelt, and then drive like an idiot, you can still kill yourself.

    I apologise for you having to make the post and I thank you for your explanation.
    I suspect that @Telford has heard Lord Sumption's comments on Radio 4 and other media in the last week - Full Fact checked his statistics here (link).
    The daily data on the number of people who have died after a positive test does include some people who died for other reasons. However, we also have data from death certificates, which records whether or not Covid itself was the “underlying cause”.

    This shows that up to 2 July this year, 124,082 people died with Covid as the underlying cause of death in England and Wales alone.

    Other things said by Lord Sumption that were refuted were that (2) people would have died anyway in the next year - in fact people who caught Covid19 on average lost 10 years of life, and that (3) the people who died had comorbidities, which again is a tiny effect.

    The other thing Lord Sumption omitted in his rosy picture was Long Covid - another 14% of those who contract Covid19 are suffering with the effects over 12 weeks later, many of which are debilitating or life changing effects - link to ONS study.

    The other thing that isn't being reported widely is that over a third of people who were discharged from hospital as recovered are returning to hospital within 140 days seriously ill with long term effects of Covid19 and another 10% died after discharge -link to BMJ article, so the 28 day figure is undercounting all these people who die of the effects of Covid19.
    Thanks for that. I did not hear the piece on Radio 4
  • orfeoorfeo Shipmate
    orfeo wrote: »
    See also: wearing a seatbelt is not a complete guarantee of not being killed in a car crash. Do you therefore say "what's the point of a seatbelt"?

    If there were people saying that seatbelts aren't a 100% guarantee against dying in an accident and therefore nobody should be allowed to drive faster than 5mph whether they have a seatbelt or not, then you're darn right I'd be asking what exactly they think the point of the seatbelt is.

    Meanwhile here in the real world, there are different places with different speed limits. There are school zones with lower speed limits. There is now a lower speed zone near my work because of a very high rate of pedestrian traffic.

    You seem to think that all that matters is whether YOU have a great seatbelt and fucking fantastic air bags, and that it doesn’t matter where you’re driving. It doesn’t matter how many lanes there are on the road, whether there are sharp turns or anything like that.

    And hell, I bet that when we abandon metaphors and go for actual real-life driving, it wouldn’t occur to you that you might need to adjust your expectations because of the WEATHER.
  • The UK app should be part of a proper contact tracing and testing programme. Part of the problem is that there isn't such a programme, and so the automated system based on a simple metric (15 minutes in close enough proximity for bluetooth contact, or check-in to the same establishment) is all that's available. In a proper system, instead of a ping there would be a trained public health professional on the phone who would ask for details of where you've been the previous few days which will be cross checked against where the known infected person was, and also questions about whether you wore a mask or washed hands etc. It's almost certainly the case that if you were walking in a park and happened to be just in front of someone else for 15m who was infected then the risks are low compared to 5m on a bus where the infected person wasn't wearing a mask, the first would ping and the second wouldn't. A conversation with a knowledgeable public health professional could tease out the details and provide a rational basis for a risk assessment. Of course, getting a test never hurts and probably should be available to everyone contacted.

    Though contact tracing through public health professionals would be swamped with the level of infection currently experienced by the UK. Keeping infection rates low reduces the work-load on the contact tracers, allowing the numbers to inflate exponentially makes contact tracing effectively impossible - though as we don't have contact tracing anyway that's maybe not a concern to the government.
  • RussRuss Ship-mate
    The UK app should be part of a proper contact tracing and testing programme...

    ...In a proper system, instead of a ping there would be a trained public health professional on the phone who would ask for details of where you've been the previous few days which will be cross checked against where the known infected person was, and also questions about whether you wore a mask or washed hands etc.

    Alan, I think what you're talking about here is a three-stage process.

    Which first identifies a set of people potentially at risk from having been in proximity to the infected person.

    Second gathers more information about each individual contact-event to judge the risk of transmission to that contact-person.

    And third applies some sort of utility-maximising framework to see what level of quarantine of that person is warranted.

    The first step requires a high level of surveillance of the population - the government knowing where everyone is every minute of the day.

    The second step requires real quantified knowledge of the transmissability of the virus under different conditions. (As opposed to an "every little helps" mindset).

    And the third requires an accepted trade-off between risk of infection and the costs/pain of preventive measures.

    Not sure how close we are to having any of those.

  • Gill HGill H Shipmate
    Don’t get me wrong, I am happy to do my bit (and I stand corrected about the app - I didn’t realise it worked when you were not checked in to somewhere).

    But I remain confused about my legal obligations (as opposed to moral ones).

    Hugal got pinged on the app late one night and the message said to self-isolate for 9 days. It said ‘press to accept’ which he did.

    There was no follow-up - no email, no invitation to log in to the Test and Trace website to share details, no text, no phone call.

    Am I right in thinking you only receive these things if you are contacted by Test and Trace (by email or phone) but not if you have only been pinged on the app?

    If you have only been pinged, are you legally obliged to self-isolate (in England? in Wales where we are?)

    Thoroughly confused now!
  • BoogieBoogie Shipmate
    I’ve no idea.

    But I do know that quarantine is closely monitored. We had a visit the next day and he insisted on seeing us both (I was still in bed!) and our passports. After that we got phone calls every day.

    Some essential businesses have now been exempted from the ping. https://www.bbc.co.uk/news/business-57937342
  • RicardusRicardus Shipmate
    Gill H wrote: »
    Am I right in thinking you only receive these things if you are contacted by Test and Trace (by email or phone) but not if you have only been pinged on the app?

    If you have only been pinged, are you legally obliged to self-isolate (in England? in Wales where we are?)

    AIUI in England there is no legal obligation to self-isolate if pinged, but for obvious reasons the government is keeping quiet about it. The reason being:

    1. The legal framework only references Test and Trace as the source of your instruction to self-isolate. (This also means that if your kid is sent home from school because someone in their bubble tested positive, there is also no legal obligation, because you were contacted by the school and not Test and Trace.)

    2. Each app only holds its contacts locally, it doesn't share anything centrally (with Test and Trace or anyone else). So Test and Trace has no way of knowing if you've been pinged or not.
  • orfeoorfeo Shipmate
    Gill H wrote: »
    But I remain confused about my legal obligations (as opposed to moral ones).

    Which is not a good situation. There definitely should be clarity on that.

  • chrisstileschrisstiles Shipmate
    orfeo wrote: »
    Gill H wrote: »
    But I remain confused about my legal obligations (as opposed to moral ones).

    Which is not a good situation. There definitely should be clarity on that.

    This kind of thing is happening often enough that it's now hard to say that lack of clarity isn't a tactic of governance. It avoids the government being pressured to do more to support people isolating, while allowing them to blame individuals for the rise in infections.
  • RicardusRicardus Shipmate
    IIRC, in this case they wanted to make it a legal obligation, but Apple and Google wouldn't let them access the data they would need to enforce it, and Mr Hancock's finest couldn't figure out how to build a working app without Google and Apple.
  • Leorning CnihtLeorning Cniht Shipmate
    edited July 23
    Of course, getting a test never hurts

    Well, that depends, as well. If you have a test which generates a lot of false negatives, and you get a negative result on such a test, and on the basis of this negative result, you declare yourself Covid-free, and so engage in risky behaviour that you wouldn't have done in the absence of a test, then the test can cause harm.

    Which is really "don't make inappropriate use of bad tests", but we've already talked about how people in general think in very binary ways.
  • BoogieBoogie Shipmate
    “Sage adviser claims ministers trying to get as many as possible infected with Covid.”

    This sounds quite likely to me. They wanted to do this from the get-go.
  • Of course, getting a test never hurts

    Well, that depends, as well. If you have a test which generates a lot of false negatives, and you get a negative result on such a test, and on the basis of this negative result, you declare yourself Covid-free, and so engage in risky behaviour that you wouldn't have done in the absence of a test, then the test can cause harm.

    Which is really "don't make inappropriate use of bad tests", but we've already talked about how people in general think in very binary ways.
    Which is part of why there needs to be a trained public health professional involved in the discussion over testing. Probably multiple PCR tests would suffice - there's good evidence that 2 tests 3-5d apart with 2 negative tests is close to being as certain that someone is not carrying the virus as we're likely to get. Of course, in the event of a ping then that would still mean self-isolation until both tests come back negative - so it may let people out of self-isolation a couple of days early.
  • Probably multiple PCR tests would suffice - there's good evidence that 2 tests 3-5d apart with 2 negative tests is close to being as certain that someone is not carrying the virus as we're likely to get.

    Agreed.
    You can probably make a good case for "isolate for 4 days, then take a single PCR test" as being sufficient, but you can't escape the necessity of isolating for a few days after a possible exposure to give things time to develop.

    So whatever happens, someone who has been exposed to Covid needs to isolate for at least a handful of days.

    The open question is how aggressive defining "exposure" should be - there's an obvious tradeoff between being aggressive about this, which minimizes Covid spread, but requires large numbers of healthy people to spend a week stuck at home - or attempting to minimize the inconvenience and economic loss from having people stuck at home, at the cost of more Covid spread.

    Of course, if the people who were "pinged" were also tested for Covid, and records kept, then we'd know where we stood, because we'd have the data.

    But I think in broad terms, you'd want people to isolate if there was a reasonable chance that they'd been infected. I suspect that the optimal threshold is probably somewhere around the 10% mark, meaning you should expect that on average 10% of the people told to isolate will develop Covid.

    Unvaccinated "you" could spend 15 minutes having your hair cut by a masked infected person, and not reach that threshold. Spend a couple of hours in a bar or restaurant with the same person, and you're over threshold.

    So what this tells me is that, if 10% is the threshold, then vaccination by itself (especially with the AZ shot) isn't sufficient to evade a requirement to isolate, but it would significantly increase the amount of contact that should be required with the infected person to trigger an isolation.

    The ping algorithm is in principle fairly sensible: it uses bluetooth signal strength to estimate distance to another phone, then computes a sum of the time spent near another device, weighted by the inverse square of distance, in order to determine a risk score. The current threshold is set at 2m distance for 15 minutes ( = 1m distance for 3.75 minutes, or 8m distance for 4 hours).

    That's certainly going to generate false pings (imagine, for example, people in adjacent flats / terraced houses spending the evening on their sofas, which back up against the same wall. The wall will attenuate the bluetooth some, but not as much as it attenuates the airflow ;)

    It would seem to me to be an improvement if the ping app would tell you where you got pinged (this can be done without any tracking information leaving the phone - your phone currently records the ping ids of anyone nearby: if it also recorded your location, then when it told you you'd had a contact, it could also tell you where you were at the time: if you got pinged on your sofa, and you know you were home alone, then you could safely disregard it.
  • BoogieBoogie Shipmate
    I think they will run out of tests. 🤔
  • DafydDafyd Shipmate
    Am I the only person who thinks the word "pingdemic" insinuates that the problem is the number of notifications rather than the rate of infections?
Sign In or Register to comment.