I think often the goals of contact tracing apps are misunderstood. The goal isn't to replace "name and phone number" human contact tracing, but rather to augment it. Iceland's location based approach seems to emphasise this, albeit at expense of privacy.
Where contact tracing apps help is for contact events where people are in relative proximity for sustained periods of time. It seems that prolonged contact is a major factor in spreading the virus. If you're on a train with someone for half an hour, you'd not be able to identify them through regular contact tracing, and they wouldn't be able to be altered. On an aircraft this can usually be done as the airline cooperates with public health authorities to identify those sitting near a given individual on the plane etc.
The impersonal but sustained contact situations are the ones apps can really help with, by raising the number of people able to be alerted to their risk of exposure. Traditional tracing handles people you know. App-based tracing, if it gets adopted, can help with the people you don't know but are sitting near, or otherwise around for a prolonged period.
Iceland is quite unique, only one real large population center, limited public transport, tiny venues, multi-generational homes and overall a tiny population so it doesn’t seem to be the prime candidate for contact tracing apps since it’s effectively the best case scenario for traditional contact tracing.
South Korea is this context similar to an island . While it is peninsula the only land border is with North Korea and highly guarded , effectively you can only enter by flying in just like an island.
Not sure about the Russia and China routes but the Busan-Shimonoseki ferry is pretty impractical. 13 hrs and about as expensive as a flight. More if you consider train costs from Shimonoseki.
I feel like this is just a coincidence. The UK is a collection of islands too and isn't doing very well - the only land border they have is with another part of an island... Ireland.
Germany which is doing much better has land borders with 9 (?) different countries, of which most are still open for commercial traffic.
Thailand is definitely not an island. Borders with Myanmar, Laos, Cambodia and Malaysia. And significant population at 69 million. Yet it is "winning" (according to https://www.endcoronavirus.org/countries), with single digit daily confirmed cases for a couple of weeks and zero new cases several days this past week. This despite having huge mainland Chinese tourist traffic before the lockdown and first confirmed case on January 13. And despite often chaotic implementation of ubiquitous masks, border closures, business closures, travel restrictions, quarantines, tracing, social distancing - all the recommended things were not done perfectly but were done quickly.
Look at the population and population distributions of GB, NZ, and AUS. It becomes pretty clear that the uk stands apart from the other two in these respects, and it is common sense that this makes it more susceptible.
Regardless of how badly we're handling it, our risk was vastly different to the others.
It's not just the fact there's 66 million people in a country smaller than Michigan either. The UK is a major transport hub. Heathrow is one of the busiest international airports in the world, both for visitors to the UK and as an international hub. London is one of the world's top tourist destinations. It's also very close to continental Europe, with tens of millions of people travelling between them by train and boat every year. The government's response was badly botched, but a fairer comparison is New York not New Zealand.
> Look at the population and population distributions of GB, NZ, and AUS.
The disparity in population distribution is no where the near the massive disparity seen in COVID 19 outcomes.
Australia has a population of 25 million which makes in more than one third the size of the UK with its population of 66 million.
Also 10 million of that 25 million live in just the two major cities of Sydney and Melbourne which I'm sure are just as densely populated as cities like London, Liverpool etc.
To date UK COVID deaths total 34466 or 508 deaths per million.
By comparison Australia has 98 or 4 deaths per million.
There is more than population distribution at play here.
You're right that comparing population centres is more helpful, but London has a total population around half that of the whole of Australia with density over 4500/km^2. Even if you expand to the whole of south east England, there's a population density of 450/km^2
Our biggest cattle station (think "cow farm" for non-Aussies) is almost 10% of the land area of the UK. We have ~20 cattle stations larger than 10,000 square kilometers. You'd only need to aggregate the top 15-17 biggest farms to get the same amount of land area as the entire United Kingdom. (Admittedly, far far less productive land, most of which probably qualifies as desert, but still...)
I (probably badly) was pointing out that vast chunks of Australia, to maybe 4 or 5 decimal places of precision, have zero people per square kilometer. We only have 25-ish million people, but fully half of them live in just three cities, Sydney, Melbourne, Brisbane.
And mine too - look how much of the country is devoid of towns with at least 10,000 people in them... I just slurped that table into a spreadsheet, there are 24,860,107 people represented in a coloured dots in that map. From that site's total population of 25,463,119, that leave 603,012 people spread across that vast unmarked unmarked grey interior...
Its a very different comparison. NZ and AUS are very similar to GB culturally and politically.
Its very notable that most countries that experienced SARS and MERS (burning it into their cultural and political will) are dealing with COVID very well.
Also "speaking for Australia", or at least my small inner-west Sydney microcosm of it - which is all my own anecdata really covers...
We are quite a way from being "almost back to BAU". This weekend we allowed cafes, restaurants, and food serving areas of bars to reopen, but only to 10 people at a time (so many have not reopened, since only 10 customers is completely unviable financially for many places). Bars (for drinking) and entertainment venues are still closed. The cafe nearest me is playing with fire right now - they have 8 outdoor tables, all more than 2m apart, but had al of them filled with 2, 3, or 4 people around each - easily 20-25 people seated while I was there. And there were still people crowding around the door clearly not "social distancing" waiting for a table to open up, and ignoring the cafe owner's request to spread themselves out to comply with the health advice...
The plan is to get back to "everything open as normal" by the end of July, but it'll only take a serious "second wave" for that plan to be found unworkabe, I suspect...
> Both were early to restrict international arrivals. The UK lagged behind.
This was a big factor. When Australia started closing its borders (on Feb 2), it was being condemned by the WHO and China for doing so [1]:
"WHO director-general Tedros Adhanom Ghebreyesus said in Geneva that despite the emergency declaration, there is "no reason for measures that unnecessarily interfere with international travel and trade"."
I was tracking conferences in the UK at the time (I run a website with conference listings), and one of the UK conferences kept quoting that WHO statement as a reason to insist their conference was still happening and people should fly to the UK for it... for another six weeks, right up until the UK lockdown. Meanwhile conferences in the rest of the world were cancelling themselves "out of an abundance of caution". The UK was one of the countries that took the longest for their conferences to cancel (Germany was the other).
Did we really "start closing our borders" that early?
I know we'd restricted direct flights from China or the Wuhan area earlier, but citizens and permanent residents (and their families) could still arrive. We were still allowing international visitors in mid March - they only cancelled the F1 GP in March 12th, the day it was supposed to star, when all the international drivers/teams/fans/media-circus were already in Melbourne.
That's fair - I meant that we'd started closing our border to China. Apparently Australia blocked tourists from China from February 1 and required other travellers from China to go into quarantine.
I'm still trying to find when the UK implemented the same restrictions on China that Australia did. Apologies for a Daily Mail link, but it suggests that a week later the UK was still struggling to implement restrictions on travel from China due to EU rules [1] (but then, the Daily Mail would say that). Deutsche Welle suggests that a Europe-wide ban didn't happen until 17th March, and even then the Schengen still remained open [2].
As for the F1 GP, I'll just paraphrase our WA Premier: "They do things differently on the east coast" ;)
More amusingly the Insane Clown Posse canceled their yearly event of themselves without any whining. Makes you wonder about the competence of a lot of people running things.
I will just speak for Australia... but Elimination was not the initial goal. It was just to ensure the hospitals could cope. It is true that Australia never entertained the idea of herd-immunity. I think some of the state governments later moved towards an elimination idea when it appeared possible.
I think that is not likely going to be possible now, since people were beginning not to accept the lock down like they had in the past. When you have a small number of cases in the hundreds for the whole of Australia, people had already started to change their behavior even before lock down restrictions had began to be lifted.
No Australian government has made any kind of indication that elimination is a goal, and nor was herd immunity a goal. The goal was time to prepare the country, and to minimise the death toll.
As it happens, it has been functionally eliminated in SA/NT/WA with no new cases in May and zero active cases. But the expectation is that new cases will arise as travel restrictions ease but contact tracing will be sufficient to control them.
> Both Australia and NZ are almost back to BAU, again, the UK will lag behind.
Not even slightly. We (Australia) only just (this week) had some states start to loosen up the initial list of restrictions, so some businesses and social activities can restart in limited fashion. But we're still on "stage 3" restrictions - with stage 4 being the most restricted.
In my own state (Victoria), we're going slower as we've recently had some new infection clusters show up, which we don't want to get out of hand.
NZ is also postponing app tracing as our cases hit zero, and our PM recommends old school ppl power. (200 tracers on standby with more being trained just in case)
NZ and Iceland may be unique, but I am extremely skeptical about tracking apps being useful at all elsewhere, in comparison to large teams of trained contact tracers.
I'm skeptical of the idea with regard to civil liberties, but I don't see why one would doubt it's efficacy. It's a tool to make the contract tracers more productive, and I expect it will be fit for purpose.
The biggest problem with contact tracing apps is that contact tracing in general is not very effective when the case rate is high. Its ideal use is to deal with the first case arriving in an isolated area. The sixth wave in the heart of NYC, not so useful.
This means that the scalability advantages of tech don't synergize well with the uses of contact tracing. If you don't have very many cases to worry about, a dedicated team of humans will do just as well. And not have nearly the same long-term privacy issues.
That seems unlikely to me given that no jurisdiction has unlimited resources, not everything is captured on camera, patient's memories aren't perfect, and there may be incentives to lie or otherwise under-report certain contacts when things aren't completely anonymous. (For example, in many countries patients are likely to avoid revealing contact with a drug dealer.)
In comparison, DP-3T (and also the joint Apple-Google framework based on an earlier revision of it) will capture any significant contact provided that both parties are carrying appropriate devices. It also maintains anonymity, so it removes nearly all incentives to mis- or under- report. It's not a replacement for teams of people, but it should provide significant improvements in areas where the majority of people are carrying such devices.
> And not have nearly the same long-term privacy issues.
Neither DP-3T, the Apple-Google framework, or the TCN protocol have any significant privacy implications (beyond whatever is already associated with carrying the physical device doing the tracing, and to the best of my knowledge of course).
Right but the hard part isn’t making the app anonymous it’s getting people to carry it. Which is significantly harder when no one cares. Iceland has less than 40% uptake in a reasonably educated population that is taking the pandemic seriously. Extrapolate from there and things don’t look great.
The post I responded to specifically objected to efficacy as well as privacy concerns so I limited my response to those points.
As to Iceland's 40% uptake, their app isn't anonymous - in the event you test positive you share your full location data with the people doing the contact tracing. It's also not built into the operating system so there's more friction for user uptake.
To be clear, I'm not saying contact tracing is useless. I'm saying its effectiveness scales inversely with the case rate.
At very low case rates, the overhead and required adoption rates for an app-based tracing system are major disadvantages against human contact tracers. As case rates go up, that overhead starts to pay off and apps make sense. But as case rates continue to go up further, you get less and less information from contact tracing. At some point, when case rates are high enough, it basically becomes a single bit of information: did you go out? Yes? Then you were probably exposed. At that point, the app isn't adding much value!
I believe that in many places around the world dealing with COVID-19, we are closer to the latter scenario than a lot of people think. No, we're not literally in the scenario above (it was an exemplar), but the general point that case rates may be too high for contact tracing to "save us" is still worth making.
Contact tracing doesn't need to be perfect, it just needs to be enough to kick the replication factor down from >1 to <1. I would be surprised if the factors you mention would reduce the effectiveness of contact tracing by more than, say, 5%.
Iceland has also has a large team of contact tracers. Which
has been pretty key along with our large scale testing of the population. The fact we’re still below 40% is a good indicator that this idea is pretty bad.
One thing the app might be used for is to allow visitors in without quarantine. Prove you have tested negative recently and use the app so if you do get sick we at least have a vague idea of where you were.
The logs that businesses are keeping in NZ are quite funny. The whole ‘write something on this bit of paper’ process is basically opt in as far as I can tell.
Every place I've been has a pen and paper and hand sanitiser next to it so you can sanitise before/after.
If the place you went to doesn't have the sanitiser part then they're probably breaching their public health obligations. You can ask them to put out sanitiser or report a breach.
As an aside, I'm impressed in NZ that there's a central place to report issues but then issues are dispatched to the right agency to follow up. That makes it super easy to report issues.
I just went to the shopping mall near me for the first time since it opened. I notice it was essentially random as to which stores required you to 'sign in' and how they did it. Some had nothing at all.
That’s just the law of small countries (just made up that term). It’s easier for smaller entities to be an outlier. You’ll find about the same positive as negative ones, though.
If you were to split a large country such as the USA or China into parts with a third of a million people, chances are that, for any of these, there would be a parts that beats Iceland.
There may be exceptions such as the highest usage of hydro-power usage, as they have a rather unique geology. Vatican City won’t be beat on popes per km², either (it has just over two)
I would doubt that. If I remember the Iceland outbreak map correctly, they had two isolated outbreak centers, far away from each other. One in Reykjavik center and the 2nd in a small town outside. I would definitely like to know if my village is suddenly affected or not.
They already did the very first traditional contact tracing, worldwide. We know a lot from this data. And they also did tracing via the mutating generations. Full scale.
Iceland is whatever you call the opposite of that. Grandma very rarely lives with her family, she lives by herself -- hopefully with her partner -- until she is unable to and then moves into a system that ranges from living-alone-with-help to typical old folks homes, and will gradually move towards more care as her health allows.
So that's wrong in the original comment, but the rest seemed pretty spot on to me, and doesn't change the argument much I think.
Contact tracing is not needed - identify the people at high risk and protect them from the virus. People at low risk should be expected to circulate the virus and develop immunity. Contract tracing among low risk defeats the purpose.
If we're ever going to get our lives back, and not be on indefinite timeline of "maybe there will be a highly effective vaccine" we actually need the virus to circulate among people who can safely get it, get over it, make antibodies which historically - before the advent of highly effective vaccines - is how pandemics ended.
Total isolation policies prevent broad population immunity. We know from decades of medical science that the population develops immunity while acquiring antibodies.
> "identify the people at high risk and protect them from the virus. People at low risk should be expected to circulate the virus and develop immunity."
Well, sure. That is what has been happening anyway. But it's hard to keep the low risk people away from the high risk people all of the time. Or are you really not going to ever visit grandma again, potentially for years?
So, even if you're low risk, it helps to know if you've been exposed so you can isolate yourself and protect the high risk people.
As someone in a low risk group (I’m 30 without co-morbidities), the risk of surviving but having an awful time is highly underrated. That’s part of why I stay at home now, I’d prefer to not experience that.
General pontifications aside, I’m sorry that you’re sick, and hope that you recover quickly and cleanly.
I think I have it. Going on 4th week with mild symptoms (severely itchy hands/arms, constant headache, muscle fatigue recently) that don't match published ones. I was probably infected going to grocer 6 weeks ago and took 2 weeks to manifest. The itch/red spots move around like every 8 hours ( I don't scratch it). What happened to you? Did it get in your lungs? What does a relapse behave like? I'm just trying to figure out what might happen.
They still wont test until you are near death. CDC and Kaiser tell me to stay home. Plague has at least 2 forms: one if its in your lung and another in the skin. I'm wonder if this might not be similar.
In Australia we tried to keep aged care homes isolated, and it was catastrophic - something like 25% of deaths were from a couple of cases that were undetected and made it into aged care homes.
In Sweden this was explicitly their plan and here's what happened:
“It’s very difficult to keep the disease away from there,” he said. “Even if we are doing our best, it’s obviously not enough.”
But he said: “We are not putting anybody’s lives above everybody else’s lives – that’s not the way we’re working.”
Tegnell said in late April that at least half of the country’s deaths had been in nursing homes.
“We really thought our elderly homes would be much better at keeping this disease outside of them then they have actually been,” he told Noah
As someone working with large health non profits on COVID tools, whose head is deep into the reporting and news around it, and who has 3 school age kids in NYC, I am of the opinion that there is no low risk population, no "safe to get it" group.
There are groups where symptomatic incidence is lower. Sure. But JFC some of those symptomatic cases are awful, with lifelong impact.
Someone used the Russian Roulette metaphor early on and it continues to resonate with me. We see over and over again surprising case histories all across the demographic, age, and health history map. Too many that are completely unexpected and for which there are presently no predictive metrics.
We are slowly learning more about what this virus can do, and we have a sense of what it will do when presented with a known vulnerable population. It is remorseless.
But even for "low risk" populations it is exactly like Russian Roulette. You and 9 others might be at the peak of known clinical health measurements and it may leave those others unscathed- that we know of- while it will destroy your lungs or your liver or cause you to have a stroke.
With all respect to the goal of broad population immunity, for my family- we stay carefully isolated and adopt all defensive protocols, we wait for a vaccine, and we advocate for govt entities to provide the liquidity to keep safely running the various life sustaining machines that our interdependent world depends on.
> we advocate for govt entities to provide the liquidity to keep safely running the various life sustaining machines that our interdependent world depends on.
There is a very real risk that won't work. If it worked, why would we only do it in emergencies? Government may as well provide funding for essential services all the time.
The government would simply be handing out money to the people working in essential services who are doing real (and quite risky, apparently) work while. You would be receiving the benefits without doing anything. That is almost the 'paid in exposure' meme that artists have to put up with.
The economic system is crafty, it will figure out that there is dead weight somewhere and start to optimise around it. This might be unprecedented so the failure modes are beyond me - but it won't be pretty if it goes on like that. Gluts would probably not be the end of the world but severe shortages are possible.
In Australia we had an official government inquiry into why the price of milk was so persistently low. No other problems. Just low milk prices. That is what happens when a free markets run the food delivery system. That won't happen if the government steps in and starts making decisions about what is fair and who needs which items.
Food is essential. Water and electricity are essential. Going to the movies or eating out is nice but certainly not essential. There's nothing clueless about wanting to keep the essentials running while limiting the non-essentials.
I don't pretend to know where the correct balance lies between injecting liquidity and reopening things but there's certainly nothing clueless about doing so in a calculated manner.
There absolutely is essential work, and at this time, until we know more, until we have both some predictive power and some reliable treatments at the individual level, this work is hazardous, barely indistinguishable from war. Those doing that work should have all of the support and benefits that come from choosing that sacrifice, putting themselves in harm's way. That we are not treating those roles that way is barbaric.
Lol, that's a poor, contextless insult, and pointless as well because you read into my comment things I did not say and do not believe.
Note that I said "safely". Essential roles are as hazardous as war and deserve the support and equipment that that entails. We are not doing that, and it is barbaric.
I'm not going bother with your dismissal of a government providing liquidity. It is clear you have no idea what that means, or how the financial system works. I can only recommend that you learn.
Good luck to you, elves of hills, brooks, standing lakes and groves...
This is a ridiculous comment, totally disconnected from reality. The risk of a seriously bad outcome from SARS-CoV-2 infection are far less than 1/6 (Russian Roulette) for most people. Hiding in our homes until a vaccine appears at some unknown point in the future is obviously not a viable strategy.
If we drive the economy into another Great Depression that will kill many people as well. Federal government liquidity injections are not a solution. At some point if people aren't out working to create value then society will disintegrate.
Appreciate your opinion, and agree the ratio isn't 1/6, but it's not de minimus and there is no safe population.
There is essential work, those jobs are wartime/highly hazardous jobs, and deserve all of the support and gear that actual war fighters get. That isn't happening.
Finally, Fed liquidity is a tool with far greater power than we are utilizing. To not exercise that power, and to not equip hazard roles- is simply barbaric.
No. In NYC, in the teen and pre-teen set, we have thousands of confirmed/symptomatic infections, hundreds of hospitalizations, dozens of deaths, and dozens of strange case histories. That's what's reported. I speculate based on anecdotal conversations with other parents there is much still unknown/unreported.
We don't have a model for how to understand exposure and immunity dynamics for this population. My personal sense/estimate is that between 10% and 20% of the child population has been sufficiently "exposed" to now be considered "safe" but I would not bet my children's lives on that measure. I am eagerly awaiting the day I can get my kids tested for antibodies, but even if they test positive I can't consider them or us to be "safe" until we know a whole lot more.
At any rate, given just the reported numbers, a conservative back of the envelope risk-of-significant-impact from "exposure" in the teen/pre-teen set is somewhere between 1 in 1,000 and 1 in 10,000.
Which says there been three coronavirus related teen+preteen deaths with no underlying conditions (all of which were teens IIRC) in NYC and another 6 with underlying conditions.
Do you have a more reliable source?
Also, 1 in 1000 is close to the estimated infection fatality rate for the general population (estimated between 0.1% and 0.5% normally) and kids are supposed to be a couple of orders of magnitude less at risk than the population average. So your numbers all seem utter nonesense to me.
I've never even heard of a preteen death of covid.
> Hiding in our homes until a vaccine appears at some unknown point in the future is obviously not a viable strategy.
Yes, this is the point I am trying to make. But most of the demographic here (who already work from home or easily transitioned to doing so) likely don't find it difficult and are mostly unaffected.
Iceland has done quite fine through this with contact tracing as one of the main tools in the toolbox. I think a blanket statement of "not needed" needs something more behind it.
Contact tracing is a world wide privacy issue, and especially in the US. I'm talking more generally here, especially that the overwhelming sentiment here on HN towards contact tracing is indifference or praise. We should not be praising further encroachments on our privacy, especially when contract tracing is hardly an effective way to mitigate the disease
We don't have evidence due to disease being so new, but there's enough reason to suspect we'll develop at least short term neutralizing immunity.
But yes, caution is necessary. We don't have enough evidence to push forward a national strategy that's based on people developing immunity without a vaccine.
What's worrisome is that some people are testing positive again, after being diagnosed as positive. Crew on the CVN-71 Roosevelt have tested positive, been quarantined for 14 days, then tested negative twice (two days apart), but then still testing positive a few days later.[1]
The possible scenarios for this I can come up with are:
1. No immunity after initial infection
2. Bad test methodology (contamination etc)
3. Bad tests. The article doesn't mention the test type, so perhaps they were initially misdiagnosed (false positive), then infected upon return to the ship. The implication from this would be that the virus is still prevalent in the crew on board the Roosevelt...
In South Korea, they discovered that all of the many potential reinfections they had detected were due to faulty (overly sensitive) tests. There are no confirmed cases of reinfection and many thousands of people on 'front line' jobs who have recovered form the disease, returned to work and never been reinfected. In a study done on monkeys, scientists were unable to achieve reinfection.
Risk vs reward. If there is immunity, then developing herd immunity sooner rather than later is important (even more so if there's immunity but it's only short-term). If there's no immunity, we're, excuse the expression, fucked anyhow.
Can vaccines even be developed if there's no natural immunity?
If there is short term immunity yes (diphteria, tetanos is an example, polio too).
Being exposed to the same antigen while you have an immunity should reactivate this immunity, that's a reason why a "herd immunity" tactic could work, at least for a time
However if the natural immunity w/o reactivation only last a year or less, you will end up with a new outbreak at the end of the period. And if the immunity period is shorter than the recuperation one, you will have a virus that will be a lot more lethal.
> Total isolation policies prevent broad population immunity. We know from decades of medical science that the population develops immunity while acquiring antibodies.
Look at the bigger picture. In today's immensely innovative world there will almost certainly be a vaccine or anti-viral or some effective treatment within a year or two.
The goal for those of us who are lucky enough (in normal times it can be a curse) to live on an island is not to hide out forever - it's to hide out until that treatment is available.
How clever. Yes, we have high confidence that hydroxychloroquine, azithromycin and zinc are very helpful at treating it. We should let the virus circulate among low risk population, isolate/protect the vulnerable, and focus on treatments that we know have been effective in the past or are showing good results now. This perpetual waiting is insane.
Again, please don't make things up. Stop spreading misinformation! It's already incredibly difficult for laypeople to make sense of the chaotic, frequently changing, and often conflicting reports at this point.
> as much as you'd like to sound cute and demeaning
My intention wasn't to demean but rather to provide a clear and understandable message for other readers who might lack the scientific background to critically evaluate your claims. I was also hoping that you'd take the hint that perhaps you lack the requisite background knowledge to commentate usefully on the topic.
The first two links are the ones with unexplainable results (to put it charitably) by the now infamous Dr. Raoult. The third one is a review which identifies previous successes in vitro and argues for further clinical trials to take place. In vitro studies are useful for identifying drug candidates to test but say absolutely nothing about whether something will work in practice.
Meanwhile, the link I provided in my previous comment neatly summarizes the current clinical data on the topic. In a nutshell, hydroxychloroquine does not appear to be particularly useful for treating hospitalized patients. It might or might not have some mild beneficial effects and it definitely has serious safety concerns; the arrival of additional clinical trial data in the near future should hopefully give us a more complete picture.
I stress again that there are currently no known effective treatments.
> In a nutshell, hydroxychloroquine does not appear to be particularly useful for treating hospitalized patients.
The early reports suggested it was only useful in combination with zinc supplements, which seems to have been largely forgotten or ignored, from what I've picked up on in comments and links across multiple sites over the past month or two.
Two of those are the discredited French study and the other is a metastudy that includes those findings along with lab results (ie, not actually showing it works in people).
You also have to consider the failure rate of the app. I don't know about the Icelandic implementation, but most approaches seem to be based on Bluetooth. There are all kinds of reasons why one might miss a signal sent by device nearby, including interference on the 2.4 GHz band with things like Wifi.
Furthermore, the user has to actually report their infection, which probably not everyone is going to do.
If we multiply all of those terms, the end result will be much lower than 15% even in this scenario. The tracing apps seem more like wishful thinking than something that is going to provide sufficiently broad and reliable coverage to have anything more than a marginal impact.
The entire premise of the app is that P[transmission] = F(distance, duration). I think this is a naive assumption and just the two input parameters have nearly zero predictive power, i.e. the variance of the estimated probability would make any further analysis useless. I'd also argue that the only case where it would be useful is tracing a zombie-level virus when there's willingness to lock down with military a 5 sigma radius around the probable infection. Otherwise it's just a tool for marketing and spying.
Maybe I am just ignorant but the article doesn't list a single reason as to why the automated tracing app hasn't helped much. There is a quote from a person who was involved in the tracing op but no comment from him/her either as to why it hasn't helped much.
This is supposed to be a "technology review". Did the writer bother to ask "WHY NOT"?
PS: Happy to stand corrected if someone else was able to read between the lines and find some reasons.
The main reason is that we were able to control the spread of the virus faster than the need for the app.
At the moment we're going several days at a time with zero new cases.
It makes it a more difficult sell to the general public, when the overall feeling is that we've 'beaten' this.
I get the feeling most of the general public aren't preparing for future outbreaks because, like most places, I think everyone wants to rely on the hope that the worst is over.
Assuming that your argument is correct that it's a little too late to see the benefits of tracing in Iceland, the article's negative connotation to use of the tracing app (or lack thereof) is misleading and discouraging for countries that are still struggling to contain the virus. I find the overall quality of online media, even tech blogs like this one that I used to love, seriously falling being caught up in the age of click bait news cycles.
Something I've been thinking about during lockdown is that, while we are for the most part staying at home when possible, if you get sick, you only have, maybe, 4 or 5 contacts to trace, on average. If we release lockdown, the number of average contacts to trace goes way way up.
Without any actual data to hand I would also guess that infections would skew towards people who come into frequent contact with others as part of their job. If true, it would make taking precautions at, for example, the grocery store all the more important because those most likely to be infected are also those who you're most likely to come into contact with in turn.
This app tracks users’locations using GPS data and according to the article "allows investigators—with permission—to look at whether those with a positive diagnosis are potentially spreading the disease"
Without also being able to know the location of people without a positive diagnosis, the app probably doesn't help that much.
But I don't think this necessarily applies to apps being developed for other countries where Bluetooth is being used to measure person to person contact directly.
This article lacks any real insight into what isn't working or why.
I've heard that the app was useful to refresh people's memory when doing contact tracing, but as the cases went down quickly after the app was released then it might not have been a gamechanger.
Couldn't you DDoS this infrastructure by having beacons masquerade as users and then have people quarantined because they were marked as near an infected individual when in reality they were near a rogue beacon that spams generated identifiers? This certainly can't be constitutional in any Western system of government.
Even if the identifiers were encrypted with a public key that only the tracer service had the private key to, couldn't a bad actor register as a bunch of actual people and then have them show as coming in contact with infected people?
No. All this app does is record a user’s location for the last two weeks. Then, after a positive diagnosis, that location data is examined as an aid to help determine who that person may have come into contact with.
Beacons, identifiers, constitutions, and public keys don’t come into it.
Those beacons would be quickly identified for being motionless, or if they are carried by a person then it's basically someone lying on the app about being infected. This could be verified by medical testing before publishing the ids.
Sure, but equivalent tracing teams in some of the bad us cities would have two to three orders of magnitude higher case load per team member. The "traditional methods" like making phone calls as described in the article do not work at that scale. Yes testing and many other things are more effective, and tracing is less useful if you've effectively got rates down to manageable levels, but at unmanaged levels, automation _could_ really help. New York has more new cases per day than Iceland has handled in total.
The next issue is whether to count positives who are new arrivals placed into quarantine of 2 or more weeks. Countries will be reluctant to blemish their clean virus-free records even though quarantine may be effective and necessary to have limited allowable travel. Seems we should only count people who test positive after passing through quarantine if they were first cleared as negative and released.
Slightly offtopic, but it bothers me that the tech community has remained silent or supportive for contact tracing technology, when there were massive outcries for the past several years regarding other kinds of surveillance-enabling technology like face recognition. In general I would prefer that contact tracing technology not exist.
I would argue that the fact that Apple and Google are working together to develop a contact tracing protocol goes to show the general level of dissatisfaction with the way contact tracing is done in other parts of the world and the privacy implications.
It's important to keep in mind that not all surveillance technologies and methodologies are created equal. The reason that the Apple/Google contact tracing technology has received so much praise is because it's more or less decentralized and all of the data is cryptographically anonymized. Any data that could be associated an individual changes periodically, so it's borderline impossible to track someone with it. There is no central database aside from the list of ID's that are marked sick. There is no location data associated either. In fact, it's blocked by the OS.
There are ways to implement contact tracing so that unnecessary information is not shared and so that you are in control of what gets shared and when. This is not the case for (say) face recognition on a CCTV feed.
I'm sure there are ways. But can you ensure this specific app, recommended (or mandated) by your government, puts you in control and doesn't try to sneak a quick one past you? What would be the incentive for your government to ensure that instead of making the population much more observable and thus governable under the guise of taking care of your health? It's literally doing work against government interests, and the governments aren't usually very good at that. Same with major companies which just happen to be also major accumulators of stockpiles of tracking and surveillance information about you. Maybe this one they'd lay down their instincts and truly will roll out a perfectly privacy-preserving solution. But could you trust them to do that?
As an example, the one used in Singapore is open source and implements a public standard (bluetrace.io), so you don't even need to trust them, you can check it yourself. And before you ask, yes, people have decompiled the app to verify this.
I also find it odd that so many people who supported a physical lockdown baulked at the privacy implications of an app.
Obviously everyone has their own set of value judgments, so I can't say such people are wrong.
But my view is that an enforced lockdown wreaks far greater damage to society than the potential privacy implications from a tracking app. If it's an either-or choice, I'm choosing a tracking app over lockdown every single time. It's a no-brainer, and I consider myself a reasonably privacy-conscious individual.
Some countries had no lockdown instead using an app and testing. I'd say it's worth it, considering it works privately until someone is found to be infected.
More and more I'm getting to the point where owning a mobile phone just isn't worth the tradeoffs. The negative attributes are starting to outweigh the positives.
I think I'm just going to stop paying to own one... All I really need is a laptop.
This paper * suggested, under some pessimistic assumptions, that it would require "near-universal app usage and near perfect compliance" This is due to the large fraction of spread from asymptomatic/presymptomatic carriers.
My review of the paper is that their assumptions are overly optimistic and that contact tracing will be largely ineffective even at universal adoption.
Contact tracing (of the human-powered kind) is obviously hugely important in reducing the scale of outbreaks in the early stages.
However digital contact tracing has a fatal flaw: Bluetooth cannot be used to reliably estimate proximity in dynamic, real-world scenarios as objects (especially human bodies) absorb huge amounts of the signal.
In many scenarios this can make two people sitting next to each other look like they're 10-20 metres away compared to line-of-sight equivalents (just by having a phone in a pocket, handbag, or even next to a head taking a call). You can easily see this using, for instance, Apple's Bluetooth Explorer tool as part of Xcode developer tools [1] (or any of the bluetooth signal strength tools in the Play / App Stores).
You don't have to rely on DIY tests from the internet though. While they're extremely static tests, the Singapore TraceTogether team did some field studies highlighting the significant variability across hardware [2]. Their tests ended in a plea for factory calibration data from hardware manufacturers.
The Singapore team has talked about false positives in depth as a major issue (one was someone in a different apartment, because bluetooth goes through walls), which is why they set a hard, low RSSI value to reduce false positives - this means a lot of true positives will be missed too.
The key Australian dev revealing significant issues in Australia's COVIDSafe app also acknowledged the major limitations of BLE. [3]
The problem of course is you have a situation where you cannot determine if a contact is epidemiologically interesting, because accuracy in real-world situations is really down to the 20-30 metres of Bluetooth range, even over longer time-frames.
This means you either have a huge caseload for human tracers to sort the signal from the noise (and this relies on the memory of all participants) or you have some kind of automated system (such as amber alerts that the NHS talks about) and the challenge there is that no-one knows if they're really interesting epidemiologically, as no-one can tell where each party was in the context.
A recent series of talks by bluetooth experts is extremely informative.
In one, an expert discusses all the significant sources of error which creates the huge variability you can see in DIY tests at home. [4]
There are other great talks in that video, but Jen Watson - who leads a team at MIT engaged in advanced signal processing - delivers a good brief talk of the issues, hoping to use statistical analysis - using detection theory of fluctuating signals to estimate interesting contacts. [5]
The takeaway from all this though is that it's a hard problem, and in Watson's talk she quickly moves on to thinking about additional future capabilities (such as features in upcoming Bluetooth standards) that might help improve the resolution.
This does leave us with a large current problem though. Tracking apps have been thoroughly oversold with little evidence of usefulness, and in the case of the UK and Australia government authorities have refused to publish the algorithms they are using to determine proximity from an RSSI value and a phone model.
There is nothing sensitive about this apart from the fact it may reveal the system is not useful for the stated purposes.
If there's a GPS signal then it may be possible to use the differential phase satellite between two phones to get relative position down to a meter or so (kind of like a super crappy rtk taking advantage of the pps stabilițy). Not sure the gps devices on phones expose enough information for this. Differential rssi of cell/wifi networks is another indicator but probably not very accurate.
This reassures the hypothesis that you probably won't catch the virus from a brief contact in a grocery store or similar if you keep your distance.
Infections happen within families, and spread from one family to another by friends/coworkers. The infected people should be easily able to list all exposed out by name.
An exception is of course mass indoors public events that are forbidden now. This is where a contact tracing app would probably prove to be the most useful.
Incidental contact is unlikely to matter in most cases. However, multiple people at different tables where infected at a restaurant from a single customer. So, it doesn’t take that much.
Yes Public transport is probably the best use case for contact tracing. A combination of location history and fare card data will work for buses. For trains it's a bit trickier since you want to resolve that you're in the same car, and Bluetooth would work well in this case.
Where contact tracing apps help is for contact events where people are in relative proximity for sustained periods of time. It seems that prolonged contact is a major factor in spreading the virus. If you're on a train with someone for half an hour, you'd not be able to identify them through regular contact tracing, and they wouldn't be able to be altered. On an aircraft this can usually be done as the airline cooperates with public health authorities to identify those sitting near a given individual on the plane etc.
The impersonal but sustained contact situations are the ones apps can really help with, by raising the number of people able to be alerted to their risk of exposure. Traditional tracing handles people you know. App-based tracing, if it gets adopted, can help with the people you don't know but are sitting near, or otherwise around for a prolonged period.