There should be a plan and an expectation, just not published. We need to follow up on what other countries are doing who are ahead of us, then make decisions based on evidence.
Got something to say or just want fewer pesky ads? Join us... 😊
Main Coronavirus / Covid-19 Discussion Thread
- Thread starter GingerBeerMan
- Start date
There should be a plan and an expectation, just not published. We need to follow up on what other countries are doing who are ahead of us, then make decisions based on evidence.
I have just updated my calculation of death rates (deaths per cases) in each country from the John Hopkins data. For the first time this week the ratio of cases per deaths has changed. The change is dramatic. The difference between the ratio today and yesterday has increased in 25 of 36 countries. It has gone down in only one and stayed the same in 10.
My interpretation of this is that the epidemic has peaked. By this I mean that the increase in deaths is now outstripping the increase in cases. Why is that important? Regardless of the variation in reported deaths and reported cases in each country, I don't expect the virus to become more virulent, so I expect the rate of death to remain constant. But the time when you get diagnosed is nearly always before the time when you die. So disease data emerges earlier than death data. If the deaths to cases ratio is changing in favour of deaths I take this to mean that the rate of increase of cases is falling (not that the rate of deaths is increasing - which requires the virus to become more virulent). There are still more cases appearing every day but the rate of this increase has peaked. The fact this is happening in so many countries, all of which capture (and fiddle with) their data differently and the only variant is Morocco, convinces me this is a real thing. To see a signal where there is so much noise is startling.
I invite being slapped down, but I find this to be encouraging.
Here are the data
My interpretation of this is that the epidemic has peaked. By this I mean that the increase in deaths is now outstripping the increase in cases. Why is that important? Regardless of the variation in reported deaths and reported cases in each country, I don't expect the virus to become more virulent, so I expect the rate of death to remain constant. But the time when you get diagnosed is nearly always before the time when you die. So disease data emerges earlier than death data. If the deaths to cases ratio is changing in favour of deaths I take this to mean that the rate of increase of cases is falling (not that the rate of deaths is increasing - which requires the virus to become more virulent). There are still more cases appearing every day but the rate of this increase has peaked. The fact this is happening in so many countries, all of which capture (and fiddle with) their data differently and the only variant is Morocco, convinces me this is a real thing. To see a signal where there is so much noise is startling.
I invite being slapped down, but I find this to be encouraging.
Here are the data
LamieRobertson
Not awoke
I thought Hancock was evasive first time round and could of just said the countries Preston was referring to were on a much tighter lockdown in the first place hence what they r doing isn’t relevant to us yet
Second time Peston he asked ..saying we don’t want a date which is true ..but an inkling of how..
The problem I see is if it becomes a media ‘thing’ then it almost making ‘relaxing lockdown’ a dirty topic
LamieRobertson
Not awoke
I agree
LamieRobertson
Not awoke
Weststander
Well-known member
How can there be a timetable, approximate dates given, when we’re at the mercy of nature?
We all know that restrictions will be eased at some stage, but 700+ people are still dying each day, as well as many others away from hospitals.
It sounds like a tedious political ploy, to enable a pointless “but you said” at a later date.
Generous of you.
So your ratio comparison tells us (for each country) that either the number of new daily cases is down, or the number of new daily deaths is up, or a combination of the two.
It were just that deaths were up (and cases about the same), I don't think that's necessarily a good thing. It might be that case numbers aren't also up simply because of the limit of the number of tests a country can do.
If it's because the number of new cases are down, surely that's a good thing, and indicative that the peak of infections has passed.
In summary, aren't you just better off looking at whether the number of new cases is falling, rather than using a ratio that takes deaths into account?
Exactly and I hope it's seen as just political posturing. It certainly isn't helpful and Starmer hasn't really started off on the correct footing. It's obvious to most that we needed to isolate and social distance yet the moronic still believe it doesn't apply to them. If the government set the timeline, they would be held to account if had to change and also if it failed. Therefore grown up thinking should preside over political scoring. The shortest notice possible will be the best way as this will stop the idiots from jumping the gun.
The stupidity of some hold no bounds, many appear to be reporters asking questions.
Looking at the data in many of the third world countries there doesn't appear to be a stark difference to everyone else. This surprises me considering the close proximity of people, their medical facilities and lack of hygiene/sanitation
It is what it is. You can process data anyway you like (as long as you do it evenhandedly). So, if you can show me greater clarity from a different assessment, that would be interesting. f the outcome is the same by an independent method, it would suggest we are both on the same track.
The other reason I'd say 'no' is that apart from anything else, the rate of change of cases on a day to day basis, (dc/dd) on its own is too variable (plus it would take too long to work t out for all those countries - but please don't let me hold you back). Also, as people have said ad nauseam there are problems with reporting that differ from country to country that means the raw numbers of deaths and cases are heterogenous. I explained my calculation in other posts, but the spreadsheet shows what I have done. I explained why the data cannot mean that the liklihood of people dying when they get sick has increased. The strength of the 'signal' is because it comes from data of heterogenous provenance and yet is unequivocally one-directional (Morocco aside).
The test of this will be the next set of data tomorrow.
Last edited:
Weststander
Well-known member
I’ve heard epidemiologists put that down to the lack of travel of populations.
NYC, London, Paris, Milan, Brussels, Madrid (sounds like a company boasting about its branches), as well as having high population density, had a huge daily turnover of international travellers in and out. Spreading the virus in January, February and March.
LamieRobertson
Not awoke
Sorry no can do ..it’s not part of the recommended daily exercise
My MP has just written to me to confirm that the furlough cut off date has been extended to 19th March, which is a relief for me. This is confirmed here:
https://www.gov.uk/guidance/claim-for-wage-costs-through-the-coronavirus-job-retention-scheme
https://www.gov.uk/guidance/claim-for-wage-costs-through-the-coronavirus-job-retention-scheme
If you have a spreadsheet with the numbers you've posted, you simply add columns for number of new cases for each day, and a column to see if that it going up or down. I don't agree that it's too variable. Your ratio will also vary, as it's directly linked to the number of new cases.
The differences in reporting makes it difficult to compare one country with another, but I'm not suggesting you do that. I'm also not suggesting you look at the number of deaths at all.
Sure, I just think you'd be better looking at the number of new cases in each country, and leave it at that.
Weststander
Well-known member
There’d be a long queue on those on the infamous Ignore List, ready to give him a slap.
oldboyroy
Well-known member
So in effect at moment we the UK are having 1 in each 8 cases texted here are dying so today's 4000 odd we can expect another 500 odd of them dying...,this is shocking but reality and added to people dying outside hospital this wretched disease is is unrelenting isn't it☹
Machiavelli
Well-known member
Isn't the rate of (confirmed) cases falling simply on the basis of the fact that most countries on your list have been in lockdown for three or four weeks, and we really ought to expect this to be the result, especially given the incubation period?
I think I should go back and see if the rate of increase in cases ordeaths has started to drop. Remember a lot of the discussion that has been had is about how we can't believe the data presented. I find it slightly ironic that because it is being said that the rate of increase may be falling (without actual evidence) people* are prepared to accept this, even though it is based on the same 'dodgy' reporting. *I don't mean you.
There are data on the numbers of confirmed cases and the rate of increase of these can be inferred from the slope of the line between days and cases. But that isn't sufficiently clear as an indicator of how we are doing because at the moment the slope is still approaching a peak asymptotically (apart from in few countries where the numbers of new cases is now less and less from one day to the next). Basically in steady state the ratio of new cases to deaths should be fixed, and when it changes something is happening. The only thing that can be happening now is a genuine fall in new cases across the board.
My number is just another measure, but it is so dramatic I thought it worth reporting. I think it does match the trends in new cases but I think it is more consistent and this a better index. The reason is it accomodates all the local variation in reporting quality and honesty. But you're right - I should really test this, so will have a look at the data again...
I read somewhere that Zambia makes 20 tests a day, dont know about the situation elsewhere but that probably tells a story.
Also the median age in EU is 42.9 years compared to 15-30 in most third world countries.
OK so I have gone back and looked at the difference in new cases between the 12th and 13th of april, and the difference between the 14th and 15th. I can't do the 13th and 14th because some numbers we not updated. But if I look at the difference between the first and second columns of numbers, doing it for these days give a better chance of picking up a slowdown if there is one. Remember with my previous calculation it was all positive (slowdown) except Morocco.
Now, simply looking at changes in reported new cases on a pair of consecutive days, we see a slowing down in new cases in 9 countries, but a speeding up of new cases in 27. Well, that is NOT good news. What does this mean? Remember my premise was that the reported numbers are inaccure from one country to the next, BUT that the inaccuracy is country specific, so as long as each country reports deaths and new cases using a consistent (albeit inaccurate) rubric from one day to the next, it may be possible to work out what is actually going on by my method. My rule of thumb is when a correction, applied fairly, removes variability and reveals a pattern, the correction is useful. Ergo, I submit my correction is useful and it is saying there is a genuine slowing down of ndw cases across the world. BUT this needs to be verified over the next few days.
