Covid19 Part XVI- 21,983 in ROI (1,339 deaths) 3,881 in NI (404 deaths)(05/05)Read OP

Xertz · 04-05-2020 03:50AM

Miike wrote: »

At this point it's too late. It's a waste of time waiting for CT results to show abnormal findings.

I still don't understand the logic of it but I keep getting the impression that when it comes to this the medical world is reinventing the wheel when it comes to statistical methodologies and it really makes no sense to me at all.

You could get into the community, take swabs of representative samples e.g. based on:

1. Geography - region/county, urban vs rural etc etc.
2. Socioeconomics
3. Types of community setting : care homes, schools, long stay hospitals.
4. Types of housing: individual homes, apartment buildings,
5. Types of workplace / types of work.

(Also you can design this so that 1 sample has multiple characteristics so you avoid having to do separate surveys and waste precious tests. Think about it as various Venn diagrams of people in multiple categories.)

If you do something like that, even with say a sample of a few thousand tests repeated at regular intervals, you would be able to feed that data into models and very rapidly build a picture of what is going on.

That's how a lot of statistical research is done for all sorts of other areas from social science, to political polling, opinion polling, and actually probably in some of its most advanced ways for market research for commercial products and services.

Right now all they're getting is a filtered, selection biased sample that is based on people who've pretty serious symptoms being tested to confirm diagnosis. To me that is absolutely meaningless data for the population. It may be useful for medics in terms of showing success of treatments and so on, but if you proposed to use this mythology in a social science or marketing thesis, you'd fail (well you probably wouldn't fail, but you'd certainly be guided towards a methodology that actually worked!)

It's producing data, but it's not data that you could realistically extrapolate anything useful from in terms of what's going on in the population.

Can they not get the CSO in to help with this? They're excellent at this kind of research planning.

Miike · 04-05-2020 03:58AM

Xertz wrote: »

I still don't understand the logic of it but I keep getting the impression that when it comes to this the medical world is reinventing the wheel when it comes to statistical methodologies and it really makes no sense to me at all.

You could get into the community, take swabs of representative samples e.g. based on:

1. Geography - region/county, urban vs rural etc etc.
2. Socioeconomics
3. Types of community setting : care homes, schools, long stay hospitals.
4. Types of housing: individual homes, apartment buildings,
5. Types of workplace / types of work.

If you do something like that, even with say a sample of a few thousand tests repeated at regular intervals, you would be able to feed that data into models and very rapidly build a picture of what is going on.

That's how a lot of statistical research is done for all sorts of other areas from social science, to political polling, opinion polling, and actually probably in some of its most advanced ways for market research for commercial products and services.

Right now all they're getting is a filtered, selection biased sample that is based on people who've pretty serious symptoms being tested to confirm diagnosis. To me that is absolutely meaningless data for the population. It may be useful for medics in terms of showing success of treatments and so on, but if you proposed to use this mythology in a social science or marketing thesis, you'd fail (well you probably wouldn't fail, but you'd certainly be guided towards a methodology that actually worked!)

We're trying to catch people who have symptoms and stop them becoming super spreaders in at risk populations. Not provide accurate statistical analyses. We don't have the capacity to run run "a few thousand tests at regular intervals" for nonclinical purpose. We aren't coping with the current clinical demand on the system.

Social sciences or marketing is not epidemiology. Contact tracing tells us a lot about what's happening.

Xertz · 04-05-2020 03:59AM

Miike wrote: »

We're trying to catch people who have symptoms and stop them becoming super spreaders in at risk populations. Not provide accurate statistical analyses. We don't have the capacity to run run "a few thousand tests at regular intervals" for nonclinical purpose. We aren't coping with the current clinical demand on the system.

Social sciences or marketing is not epidemiology.

It clearly isn't as the data that it's producing is extremely flawed.

If you turned this stuff up in the context of a marketing analysis you'd be fired. I'm not kidding.

There has been continuous attempts to extrapolate details of community spread from this data and it makes no sense whatsoever.

Also, how is is catching 'super spreaders' ?

The people who it's testing are people who have symptoms and are already self isolating in probably 99% of cases, so they're not spreading anything. The super spreaders are people who are mildly or asymptomatic and they're not being tested at all.

Miike · 04-05-2020 04:00AM

Xertz wrote: »

It clearly isn't as the data that it's producing is extremely flawed.

If you turned this stuff up in the context of a marketing analysis you'd be fired. I'm not kidding.

and if you turned up marketing analysis for epidemiological purposes you'd be shot. You're comparing two wildly different fields and approaches.

Xertz · 04-05-2020 04:06AM

Miike wrote: »

and if you turned up marketing analysis for epidemiological purposes you'd be shot. You're comparing to wildly different fields and approaches.

So explain me :

1) What is the purpose of this data gathering?
2) What are we trying to achieve here?
3) What use is the data that is being gathered put to?

Basically we are taking someone who has symptoms, telling them to self-isolate, so they don't spread anything anyway, then taking them to a testing centre, testing them positive / negative and sending them home in most cases.

At best all it would seem to do is alert someone they have coronavirus and ensure that they take their symptoms seriously or get medical assistance / monitoring so they're hospitalised at the right point, should they need to be.

Beyond that, what exactly is it achieving?

It's not finding people who are super spreaders because those people are not sick or are mild sick.

Members of my own family had mild systems and were turned away / had tests cancelled.

So it is absolutely not telling us anything about how many of those people are out there in the community or where they are likely to be.

The only thing I can see that has worked well in this has been the mass restrictions as that makes sense and has reduced spread dramatically and thus deaths.

However, we still know very little about the actual scale of this in the community because the testing simply could not be telling us anything about that with the current methodology.

Statistics are statistics no matter what type of research you're carrying out and there are some fundamental issues with how this is being conducted that are the same whether you're looking at how many people have coronavirus or how many people have a preference for chocolate biscuits.

If anything, market research or complex social science research is FAR more complicated as you're looking at multiple variables like preferences etc. This is just a case of find one variable : do they have / don't they have coronavirus and then link it to geography or some useful other variables like type of work, type of housing, congregated settings, socioeconomic status etc.

If you had a model of all of that, and you understood the risk areas, then you can put mitigation in place where it is actually needed and relax things were spread is low for example.

Miike · 04-05-2020 04:07AM

Xertz wrote: »

It clearly isn't as the data that it's producing is extremely flawed.

If you turned this stuff up in the context of a marketing analysis you'd be fired. I'm not kidding.

There has been continuous attempts to extrapolate details of community spread from this data and it makes no sense whatsoever.

Also, how is is catching 'super spreaders' ?

The people who it's testing are people who have symptoms and are already self isolating in probably 99% of cases, so they're not spreading anything. The super spreaders are people who are mildly or asymptomatic and they're not being tested at all.

We both keep editing our posts to make things clearer and I'm on mobile so I'm at a disadvantage here ha! :pac:

It's STOPPING people becoming super spreaders, by, for example: Catching healthcare staff (who have sustained interactions with the public, the already sick public) who may be pre/asymptomatic or nursing home residents who are in a very risky environement etc. I didn't stop it's catching super spreaders, it's stopping that kind of spread happening. This is partly the purpose of testing criteria.

Re attempts to extrapolate community spread: This is done in conjunction with contact tracing data and trends being fed into models. This gives a clear picture of the type of spreading that's happening and where.

Turtwig · 04-05-2020 04:08AM

Miike wrote: »

The R0 of this virus was not 2 without control measures in place. If the virus was here in an unidentified and uncontrolled state the R0 is much higher. R0 is not a constant.

We don't know how many people are asymptomatic. We have rough estimates that none of the research agrees on and as such cannot remotely being to calculate the true fatality rate of the virus. Also, it doesn't take several chains of transmission for a virus to become lethal. Cases of pneumonia of unknown origin would be subject to heavy retrospective scrutiny at this point, especially once this was declared a PHEIC and then again when declared a pandemic.

It only takes one chain, one patient for a virus to be lethal. I could have explained my point better.

Suppose someone dies of unusual pneumonia. While it's possible that's your very first case of the new virus. If the fatality rate is 1% it's more probable there is somewhere around 1 - 100 infected people out there. Likewise, if a disease is only severe in a quarter of cases then the.first manifestation of severe disease probably isn't your first case.

In this way, the more chains of transmission you have the more visible the disease becomes.

Regarding asymptomatic carriers: fully agree.

Turtwig · 04-05-2020 04:13AM

Miike wrote: »

We both keep editing our posts to make things clearer and I'm on mobile so I'm at a disadvantage here ha! :pac:

The time traveling is way more pervasive on regular browsers

(Hence why I'm on mobile - also might be watching westworld)

Xertz · 04-05-2020 04:25AM

Miike wrote: »

It's STOPPING people becoming super spreaders, by, for example: Catching healthcare staff (who have sustained interactions with the public, the already sick public) who may be pre/asymptomatic or nursing home residents who are in a very risky environement etc. I didn't stop it's catching super spreaders, it's stopping that kind of spread happening. This is partly the purpose of testing criteria.
e.

Healthcare worker screening is really more part of health and safety screening in hospital environments / care homes though and should be separate from any statistical testing for population analysis. That's essential to safety.

Getting the volume of tests up is extremely important, but so is using them correctly to build a serious model of population spread and given that the virus has a huge issue with being asymptomatic for a lot of us, it absolutely needs a part of the research to be based on a much broader sample selection approach.

Again, you don't need a vast amount of tests to do this.

You could use say 5-10% of the total available tests to do something like that and still get enormous amounts of data if you just set out criteria where each sample point has multiple variables.

With a bit of clever design of how you're selecting who you're surveying, you can pull a hell of a lot of data out of a relatively small sample e.g: you'll have the ability to pick those being surveyed using criteria that basically create multiple overlapping Venn diagrams i.e. each person is actually a member of a number of different groups in your sample.

Data is data is data is data whether it's medical, socioeconomic, creepy stuff Google knows about you and there are ways of looking at linkages, drilling down and even using complex computer models and AI to really extract a hell of a lot from it.

I would rather see this done to maximum efficiency to get the highest resolution picture we can get.

I guarantee you, when the US and particularly states like California get past the initial crazy issues with Trump and so on, they will apply advanced data analytics to this stuff with all of the firepower used by the likes of Google and will get it beaten very effectively.

If we keep going the way we are we'll still be queuing people up finding out what we already know!

COVID-19 elimination is, until we get drugs and vaccines, a social science / data analysis issue primarily. It's about suppressing it with clever use of social measures and that's why we need to be bringing in the big guns on analysis, not just relying on the medical side of things.

Miike · 04-05-2020 04:43AM

Xertz wrote: »

Healthcare worker screening is really more part of health and safety screening in hospital environments / care homes though and should be separate from any statistical testing for population analysis. That's essential to safety.

Getting the volume of tests up is extremely important, but so is using them correctly to build a serious model of population spread and given that the virus has a huge issue with being asymptomatic for a lot of us, it absolutely needs a part of the research to be based on a much broader sample selection approach.

Again, you don't need a vast amount of tests to do this.

You could use say 5-10% of the total available tests to do something like that and still get enormous amounts of data if you just set out criteria where each sample point has multiple variables.

With a bit of clever design of how you're selecting who you're surveying, you can pull a hell of a lot of data out of a relatively small sample e.g: you'll have the ability to pick those being surveyed using criteria that basically create multiple overlapping Venn diagrams i.e. each person is actually a member of a number of different groups in your sample.

Data is data is data is data whether it's medical, socioeconomic, creepy stuff Google knows about you and there are ways of looking at linkages, drilling down and even using complex computer models and AI to really extract a hell of a lot from it.

I would rather see this done to maximum efficiency to get the highest resolution picture we can get.

The kind of surveillance data you're looking for is going to achieved with antibody testing, not with PCR. This will use the methodologies you're familiar with as you've outlined.

We need to identify cases of disease in the higher risk groups, this is the best clinical (not statistical) application of PCR testing within our current capacity and scope.

I can't answer all the questions you've posed above while on mobile, sorry

JanuarySnowstor · 04-05-2020 07:44AM

I wonder is the hot sunny weather a factor afterall
Europe cases are falling steadily but it's coinciding with some quite hot weather!! If it's characteristics are similar to flu it won't stand a chance with temps over 30C

Brianmwalker · 04-05-2020 08:31AM

JanuarySnowstor wrote: »

I wonder is the hot sunny weather a factor afterall
Europe cases are falling steadily but it's coinciding with some quite hot weather!! If it's characteristics are similar to flu it won't stand a chance with temps over 30C

Yeah it's the good weather, nothing to do with the lock downs. Ffs

Brianmwalker · 04-05-2020 08:45AM

JanuarySnowstor wrote: »

I wonder is the hot sunny weather a factor afterall
Europe cases are falling steadily but it's coinciding with some quite hot weather!! If it's characteristics are similar to flu it won't stand a chance with temps over 30C

Yeah it's the good weather, nothing to do with the lock downs. Ffs

Jim_Hodge · 04-05-2020 08:54AM

JanuarySnowstor wrote: »

I wonder is the hot sunny weather a factor afterall
Europe cases are falling steadily but it's coinciding with some quite hot weather!! If it's characteristics are similar to flu it won't stand a chance with temps over 30C

No mention of the affects of lockdowns? Plenty of cases in countries warmer and sunnier than we've have in an Irish April. 33c and higher in the middle east recently and cases are still growing in UAE and others.

Not in Kansas · 04-05-2020 09:00AM

fritzelly wrote: »

As for the other poster worrying about having spread it to here colleague - as Dr Holohan says, no blame for spreading it can be put on anyone, it like most viruses are easily spreadable. You could spread flu to someone and it kills them but if they didn't catch off you they could have caught it off someone else.

That's an insane thing to say. Of course blame can be apportioned if people go into work day after day knowing they are sick. Asymptomatic virus carriers have no knowledge or control over their influence but those with symptoms do, and choosing to go into work knowing you are sick is idiotic and shows a blatant disregard for your co-workers and their families.

And that applies just as much to people who - prior to Covid 19 - behaved this way with Flu symptoms. Surely those days are gone forever. You would be treated like a social pariah turning up to work sick now and hopefully would be sent straight home.

bilston · 04-05-2020 09:01AM

The battle for a vaccine could become interesting in the months ahead. There doesn't appear to be a united global effort.

It appears the EU, Britain, Norway, Canada and Japan are working together, but worryingly the US and China are going it alone.

Countries will hopefully realise that for life to return to normal then you won't just be able to hog any vaccine for yourself.

easypazz · 04-05-2020 09:03AM

JanuarySnowstor wrote: »

I wonder is the hot sunny weather a factor afterall
Europe cases are falling steadily but it's coinciding with some quite hot weather!! If it's characteristics are similar to flu it won't stand a chance with temps over 30C

Indirectly it is I think. It started raining the other day and everybody out walking ducked into the shop and forgot social distancing.

The good weather certainly allows people spread out more.

Hector Savage · 04-05-2020 09:19AM

Glorious weather here the last few days, have been on the beach daily with the kids, the mediterrean looks so clean, one silver lining in all this is this summer the med will really get a chance to regenerate - it will be a marine biologists dream.

Thankfully people don't seem as tetchy and paranoid as they were last weekend and thankfully I didn't witness any arguments over space this weekend.

However I did notice people are getting complacent , we keep our distance on the beach but I notice groups of people fairly close together, and families seeming to be meeting up.

Lets see this in June/July when it gets really hot.

Azatadine · 04-05-2020 09:21AM

Hector Savage wrote: »

Glorious weather here the last few days, have been on the beach daily with the kids, the mediterrean looks so clean, one silver lining in all this is this summer the med will really get a chance to regenerate - it will be a marine biologists dream.

Thankfully people don't seem as tetchy and paranoid as they were last weekend and thankfully I didn't witness any arguments over space this weekend.

However I did notice people are getting complacent , we keep our distance on the beach but I notice groups of people fairly close together, and families seeming to be meeting up.

Lets see this in June/July when it gets really hot.

Hector, I've never seen you so non apocalyptic.

sadie1502 · 04-05-2020 09:30AM

forgodssake wrote: »

Agree somewhat with poster saying Covid 19 was here since oct/Nov last year. As a nurse of 21 years I have never before seen the number of pneumonia related admissions before. Also on a side note I haven't needed /taken antibiotics since 2010 . Last year I had 2 courses for throat/chest infections which absolutely floored me.

I too had a chest infection could not shake it couldn't catch my breath a cough so dry it was like a bark. Spent nights awake for over a month two antibiotics and inhalors because I couldn't breathm I'm not asmathic. Wouldnt be over weight. The girl I work with end up in hospital with pneumonia for weeks. Were in our 30 / early 40. But this thing absolutely flooded me. Steroids antibiotics inhalors. Never was on a steroid or inhaler before. Convinced it was here since last year my boyfriend and friends think the same as they know how sick I had been.

Non solum non ambulabit · 04-05-2020 09:33AM

Interesting analysis. Healthcare workers provide a great barometer of this diseases potency.

https://mobile.twitter.com/higginsdavidw/status/1257029491177795594

Over 5600 health care workers infected with only 5 deaths. So 0.1% CFR.

Brianmwalker · 04-05-2020 09:38AM

JanuarySnowstor wrote: »

I wonder is the hot sunny weather a factor afterall
Europe cases are falling steadily but it's coinciding with some quite hot weather!! If it's characteristics are similar to flu it won't stand a chance with temps over 30C

Yeah it's the good weather, nothing to do with the lock downs. Ffs

Hector Savage · 04-05-2020 09:38AM

Azatadine wrote: »

Hector, I've never seen you so non apocalyptic.

Thanks, it helps that things are finally improving here - healthwise that is, now lets see how the economy recovers

Also, some things and this applies to anyone else who feels negative.

* reduce time on newssites - I just check news briefly in the evenings...
* reduce time on this thread - sorry, but it can get mentally tiring seeing nothing but bad news ...
* Stop watching John Campbell - haven't watched him in almost 2 weeks, I know he's pretty level headed, but again, it's obsessively going over the bad news bad news bad news ... numbers numbers numbers ... f*ck that ...
* read/watch TV - especially comedy to keep mind off it ...
* Stop watching John Campbelll

ChelseaRentBoy · 04-05-2020 09:46AM

easypazz wrote: »

Indirectly it is I think. It started raining the other day and everybody out walking ducked into the shop and forgot social distancing.

The good weather certainly allows people spread out more.

You can't just duck into a shop, all shops have distancing guidelines.

Boggles · 04-05-2020 09:47AM

sadie1502 wrote: »

I too had a chest infection could not shake it couldn't catch my breath a cough so dry it was like a bark. Spent nights awake for over a month two antibiotics and inhalors because I couldn't breathm I'm not asmathic. Wouldnt be over weight. The girl I work with end up in hospital with pneumonia for weeks. Were in our 30 / early 40. But this thing absolutely flooded me. Steroids antibiotics inhalors. Never was on a steroid or inhaler before. Convinced it was here since last year my boyfriend and friends think the same as they know how sick I had been.

Did your boyfriend get sick?

Home antibody tests should be wide spread and affordable in the next few weeks apparently. You should definitely do one.

ChelseaRentBoy · 04-05-2020 09:51AM

Russia now in big trouble.

New York Times - Top Five Countries where cases are increasing

Russia
Brazil
Peru
India
Saudi Arabia

Top Five Countries where cases are stable

United States
United Kingdom
Canada
Sweden
Belarus

Top Five Countries where cases are in decline

Spain
Italy
Germany
France
Turkey

Europe seems to be getting a hold on it but all depends on what happens next with the relaxation of restrictions.

Touchee · 04-05-2020 09:55AM

Not in Kansas wrote: »

That's an insane thing to say. Of course blame can be apportioned if people go into work day after day knowing they are sick. Asymptomatic virus carriers have no knowledge or control over their influence but those with symptoms do, and choosing to go into work knowing you are sick is idiotic and shows a blatant disregard for your co-workers and their families.

And that applies just as much to people who - prior to Covid 19 - behaved this way with Flu symptoms. Surely those days are gone forever. You would be treated like a social pariah turning up to work sick now and hopefully would be sent straight home.

I agree that people with flu or cold symptoms should not be turning up for work. But, the problem is that a lot of employers do not pay for sick leave. Therefore, some people cannot afford to call in sick, unless they really are on their death bed

Wibbs · 04-05-2020 09:58AM

Boggles wrote: »

Did your boyfriend get sick?

Home antibody tests should be wide spread and affordable in the next few weeks apparently. You should definitely do one.

+1 I'm certainly going to do one. The precise symptoms of that dose I had in January would get me a Covid19 test in a heartbeat these days. Ditto for a few people I know who had similar. Including one family who always get the flu jab, already had a flu in early December and had gone on a European break at Christmas and came back with a dose, which I caught from them. Just sounds too coincidental to me. Maybe there were two strains of flu going on? That's the only way I can think to explain it without involving Covid19

laurah591 · 04-05-2020 10:01AM

ChelseaRentBoy wrote: »

Russia now in big trouble.

New York Times - Top Five Countries where cases are increasing

Russia
Brazil
Peru
India
Saudi Arabia

Top Five Countries where cases are stable

United States
United Kingdom
Canada
Sweden
Belarus

Top Five Countries where cases are in decline

Spain
Italy
Germany
France
Turkey

Europe seems to be getting a hold on it but all depends on what happens next with the relaxation of restrictions.

Is Russia in big trouble though? - they have run over 4m tests and identified 140k odd cases with fatalities of less than 1500. Compared to say Germany who are praised for there handling of this virus.

sadie1502 · 04-05-2020 10:04AM

Boggles wrote: »

Did your boyfriend get sick?

Home antibody tests should be wide spread and affordable in the next few weeks apparently. You should definitely do one.

He got the sniffles but nothing to what I had. Will the home testing kits be available soon? I'd definitely get one.

Covid19 Part XVI- 21,983 in ROI (1,339 deaths) 3,881 in NI (404 deaths)(05/05)Read OP

Comments