The 9am figures not disclosed yet?

What do you think? Are we levelling out as the ‘experts’ seem to be claiming at the 5pm updates and when do you think the deaths will peak?
One key measure, that is presented as a graph in the daily briefing, is numbers of people in ICU beds. If that number begins to fall, the NHS is getting on top of the problem. That is beginning to look promising.

Another important measure will be numbers of people recovered. When recoveries exceed new cases, as Italy has got to, the epidemic is on the same. That figure isn't available.

The third measure is number of tests per day, as that will be key to minimising subsequent outbreaks. Although there were 16000 tests today, only 10500 people were tested. Does this say we are very inefficient in our testing ability as well as being 90000 tests a day short of where we need to be?
 
What do you think? Are we levelling out as the ‘experts’ seem to be claiming at the 5pm updates and when do you think the deaths will peak?

Well, the first thing I should do is give the caveat that I'm not an epidemiologist. I'm just using my skills and knowledge to analyse publicly available data, but I've no professional expertise in the behaviour of diseases, or even health in general.

With that health warning, I would say that the 4-day average for new infections has fallen (i.e. negative growth) for three out of the past four days. The rate of growth has also fallen for seven out of the past eight days.

This is despite the fact that the overall level of testing has increased over the past week, so you might have expected a jump in new infections to have followed that. Therefore, I think there is reasonable evidence that the infection numbers may, indeed, have levelled out.

If we look at new deaths, although they fell slightly today from yesterday, the overall trend is still upwards. You would probably expect the infection rate to reach its peak before that is then reflected in the death rate.

I wouldn't like to say when the peak might actually arise, as it would be pure guesswork on my part. I'm much more comfortable analysing what has already happened than predicting what might.

That said, I would caution against anyone thinking that reaching the peak means that we will be nearly through this. From the experiences of both Italy and Spain, and assuming that they both have passed their peaks, the death rate appears to remain stubbornly high afterwards.
 
Last edited:
Well, the first thing I should do is give the caveat that I'm not an epidemiologist. I'm just using my skills and knowledge to analyse publicly available data, but I've no professional expertise in the behaviour of diseases, or even health in general.

With that health warning, I would say that the 4-day average for new infections has fallen (i.e. negative growth) for three out of the past four days. The rate of growth has also fallen for seven out of the past eight days.

This is despite the fact that the overall level of testing has increased over the past week, so you might have expected a jump in new infections to have followed that. Therefore, I think there is reasonable evidence that the infection numbers may, indeed, have levelled out.

If we look at new deaths, although they fell slightly today from yesterday, the overall trend is still upwards. You would probably expect the infection rate to reach its peak before that is then reflected in the death rate.

I wouldn't like to say when the peak might actually arise, as it would be pure guesswork on my part. I'm much more comfortable analysing what has already happened then predicting what might.

That said, I would caution against anyone thinking that reaching the peak means that we will be nearly through this. From the experiences of both Italy and Spain, and assuming that they both have passed their peaks, the death rate appears to remain stubbornly high afterwards.
Thanks for that, I do appreciate it’s just a personal opinion, but you do keep a tight eye on the stats so you will have a feel for the trends.
One stat I find interesting is that the testing seems to have a 25% positive rate and how reflective that is of the wider population.
 
Thanks for that, I do appreciate it’s just a personal opinion, but you do keep a tight eye on the stats so you will have a feel for the trends.
One stat I find interesting is that the testing seems to have a 25% positive rate and how reflective that is of the wider population.
It's nearly 50% the number of people tested is far fewer than the number of tests. Only people presented to hospitals are tested (more serious symptoms) so it doesn't account for those recovering at home.
 
It's nearly 50% the number of people tested is far fewer than the number of tests. Only people presented to hospitals are tested (more serious symptoms) so it doesn't account for those recovering at home.
Does the number of tests include those who are retested?
On the 5pm yesterday didn’t they say they thought about 10% of the population may have had the virus? I thought that was disappointingly low, however it has been arrived at.
 
It's nearly 50% the number of people tested is far fewer than the number of tests. Only people presented to hospitals are tested (more serious symptoms) so it doesn't account for those recovering at home.
Less than 2% of tests where positive until we started only testing hospital patients.
 
Was that about 2% of the tests? Every man and his dog were tested in those days. People returning from Italy etc. With no symptoms.

The hospital positives have gone up from 25% to 50% as people are not presenting themselves with mild symptoms. Also, the number of people tested is also listed as well as the number of tests.
 
Thanks spanishman. Do you know the source for those figures? Are they the official statistics?

I use the stats on wikipedia for each country, as they have a page for every country with their timeline in the same format. They use worldometer as their source.

I did check my figures against the ones the British government are using when comparing the UK position with Spain, and they're almost identical, so I think they must be using wikipedia too!
They come from this Spanish site. From a university bloke somewhere in Spain. I will add a Google translation of any bits I think are usefull in a couple of minutes.

https://smreputationmetrics.wordpre...situacion-y-tendencia-tasa-de-multiplicacion/

These are his Italy stats

https://smreputationmetrics.wordpre...-situacion-y-tendencia-tasa-de-multipicacion/
 
For a few days, I have been publishing the daily data on the trend of the expansion of the coronavirus epidemic in Spain on Twitter, using the variable of the multiplication rate, in the hope that the official data published every day may have easier and clearer reading and understanding by the general public.

The trend measurement measure that I have selected is the multiplication rate of each variable every four days (of registered cases, ICU admissions, deceased). In other words, by what number have today's total data been multiplied compared to four days ago. It is important to emphasize that this is total data.

If the multiple reaches a value of 2.0, it means that the number of cases or deaths has multiplied by two in four days. If it is 1.5, they have increased by 50% in four days. This multiple indicates the goal to be achieved to have the epidemic fully controlled. If it takes a value of 1.0, today's total cases are the same as those from four days ago: in four days there has been no new case, or new deceased.

The contribution that the presentation of the official data gives in the form of multiples or multiplication rate is both in the value that this multiple takes, and especially in the trend that this multiple experiences. If it decreases day by day, although the absolute numbers increase and even reach maximums (which the epidemic dynamics happens almost every day), the expansion of the epidemic is slowing down. When the multiple goes below 2.0 it is in all probability as a direct consequence of the personal and legal measures of social distancing.

In another entry in this blog I explain in more detail the role of the multiplication rate in the communication of coronavirus crises (access here).

In this blog post I will also publish updated data on the multiplication rate for the epidemic in Spain every day.
 
There is a longer descriptive article:

https://smreputationmetrics.wordpre...en-la-comunicacion-de-crisis-del-coronavirus/

UPDATED: have added the diagrams in the translation.

Here is a Google translation:

I propose in another way the usefulness (importance?) Of presenting the evolution data of the epidemic showing relative data such as the multiplication rate (but also the daily percentage increase or the doubling time), in addition to communicating the figures daily and total cases, for crisis communication adapted to the challenges of its public and private management.

For those unfamiliar with the term, the multiplication rate is the number of times the total number of reported (or deceased, or hospitalized) cases is multiplied in a given period of time. The interval that I have chosen in the charts that I disseminate is 4 days (I will explain why 4 days in another space). It is important to highlight that it is a matter of relating the increase in total cases, from the beginning of the epidemic. For example, the total cases registered in Spain on April 1 were 102,137. Four days earlier, on March 28, there were 72,248 cases registered. So, the case multiplication rate as of April 1 is Ct / Ct-4; 102,137 / 72,248 = 1.41. The cases have multiplied by 1.41, they have increased by 41%. If it is 2.0, they are multiplied by two every four days. If it is 1.0 (exact), in the last four days there have been no new cases (Ct / Ct-4 = 1; Ct = Ct-4), the expansion of the epidemic has been entirely cut off, as regards cases of declared new infections.

The epidemic, all epidemic diffusion moves in cycle: implantation, development, growth, containment, deactivation, or in the technical way that they receive from epidemiology experts (among whom I am not). The dynamics of the cycle, reflected in the evolution of daily cases, is the elephant within the boa.

BackgroundArticleImage1.jpg

All right. The problem with the cycle is that appearances are deceiving, but not systematically, but in two critical phases of the cycle: in the initial propagation (Phase 1) and in the zone where the maximum number of cases is reached, the famous peak (Phase 3).

In the initial development part (Phase 1), the appearance is that things are going well, because we are talking about tens or a few hundred cases, with no deaths. Insignificant, negligible thing. And so it has been despised country after country, despite having the message of terror from other countries that were advanced in the internal development of the epidemic. The appearance is that everything is going well, that there are so few cases that everything is controlled, because they are detected above. But the reality is that things are going wrong, because inaction allows the bug to spread and multiply at will, at its natural rate. No one slows it down. And it is the background noise, which is almost invisible, but it is real: contagion multiplies every two to three days. Against small, insignificant total numbers, yes. They are the snowball that will make it more or less giant in the end. The reality is bad.

Then when the cases are fired (Phase 2) there is no longer a divorce: the appearance is really bad, and so is the reality. The cases continue to multiply, which are already beginning to be noticeable, and the first measures are taken to limit contagion, but the effects translate into nothing. At this stage it is very easy to take containment measures against the spread of the epidemic, because everyone sees the catastrophic consequences. But the great thing would have been to take those steps before it explodes in the face, the fastest in phase 1.

In phase 3 there is again a problem of divorce between appearance and reality, because the containment measures, private and public, are already beginning to have a real impact in reducing the rate of virus transmission, but the appearance is Dantesque, because they rise cases. This divorce between appearance and reality is also disastrous, because it consumes energy, the hope of those who suffer the consequences of confinement, casts doubt on the measures and puts pressure to take more measures like crazy.

But the situation is about to turn around. And in the last phase (Phase 4), new cases fall day by day, and then a little later, that of the deceased. Things seem to get better, and they actually do.

Well, it is very difficult to see that things go wrong in the initial phase, and that they begin to go well in phase 3 of change, if one looks at the graphs of daily data and accumulated data (both in linear or logarithmic format) .

BackgroundArticleImage2.jpg

BackgroundArticleImage3.jpgBackgroundArticleImage4.jpg

To illustrate the dynamics of the epidemic, I take the case of the epidemic and closed in the first wave, from Wuhan-Hubei. The first graph is the one of the new daily contagion cases reported, from the day that there are 100 counted cases. The jump in data observed from the 24th is artificial: it is because from that day on, not only the cases of the macro city of Wuhan, but also those of the region, Hubei, are added to the statistics.

BackgroundArticleImage5.jpg

It is difficult to see, in any of the 3 graphs, that the situation in Phase 1 is bad and that the situation in Phase 3 is good (something if you see it in the first daily data graph).

Instead, that information, both that things are going wrong in the initial phase (Phase 1), and that things are improving in transition phase 3, they are in the multiplication rate charts and the like.

That is, the multiplication rate (and similar relative measures) manage to match well, for each of the stages, the visual information with the reality of the dynamics of the epidemic, of reality, whether it is already apparent or hidden. . Therefore, it illuminates the experts' speech and makes it understood.

Let's go back to the multiplication rate graph. In phase 1 (orange), although the cases are few, the indicator alerts you that the situation is out of control: the cases multiply by 4! every 4 days. It is a clear and resounding message, which should help to take action.

In phase 2 (red), of already visible and constant growth of cases and deaths, people understand that 1000 is bad, and that if you have 3,000 four days later, we have a serious problem. Politicians and the public are ready to take action. Increasing absolute numbers help convey a sense of emergency and therefore urgency.

Does the multiplier drop in this explosion phase? Yes, because people already started shaking hands less and began washing their hands more. But the message of the rate is not that it is going down, it is that the multiple is still between 2 and 3. If there is no reaction, the total figures will continue to multiply, explosively.

And it reacts. In individual behavior and in public measures of mobility restrictions and social contact. And the days go by, and it seems that it is useless because the dead continue to grow. But yes, it happens. The multiplier already falls below 2. It is Phase 3 (yellow). It is in this phase 3 where the multiplication rate again plays a role that can be key in crisis communication: to explain that yes, that the measures have and their expected impact, despite appearances. In this phase you have to drill less with the death records. Not because the death toll and its increasing number is a lie, but because that is only part of the reality of what is happening. Because it is really happening: the change of rhythm announces a change of cycle.

And I remain quite convinced that transmitting that idea, that the dynamics are changing, is really difficult to transmit with data and graphics in absolute terms. And that "trust us, we are better" does not work either. People need to see. And what the multiplication rate tells us, or the measures of percentage increases, is that you can see it. That this tool exists, that it helps to detach itself from appearances. That the rate is not a trick of manipulation, but on the contrary is a way to touch reality.

We are all learning everything with this pandemic, its impact and its management. Cross fertilization of knowledge and talents is taking place in a manner and with prodigious speed and flexibility. It must also be done in how to communicate the crisis.

Francesc, are we in Phase 3? I think so. It is precisely what epidemiologists say (pending that yes of the evolution in Catalonia). I do not know how far we are from Phase 2 and how close to Phase 4. As we are probably in Phase 3, it is time to put the indicators of relative variation of the epidemic into action, because they can help the population (and the press and political leaders) to understand what is happening, so as not to become more psychologically overwhelmed.

And there will come a phase 4 (green), in which the multiplication rate will get closer and closer to 1.0, to the goal (I remember that the rise here is because they added cases from the entire province). But in this phase 4 the multiplication rate will no longer play any strategic role, because the appearance will align itself again with reality: each day the new cases and deaths are lower than the previous day. We can abandon the crutch. It is on those days when tweeters will begin to appear that they will respond with a “small bull**** about the multiplication rate, just look at the data in the daily case graph to know that we are out of the crisis. Small experts ”. When those tweets arrive, we'll celebrate. Meanwhile, I continue to think that those who are reporting the crisis, in addition to giving the terrible daily and accumulated data on cases and deaths, should provide information on developments in relative terms.
 
Last edited:
There is a longer descriptive article:

https://smreputationmetrics.wordpre...en-la-comunicacion-de-crisis-del-coronavirus/

You can see the diagrams referred to in the translation in this article.

Here is a Google translation:

I propose in another way the usefulness (importance?) Of presenting the evolution data of the epidemic showing relative data such as the multiplication rate (but also the daily percentage increase or the doubling time), in addition to communicating the figures daily and total cases, for crisis communication adapted to the challenges of its public and private management.

For those unfamiliar with the term, the multiplication rate is the number of times the total number of reported (or deceased, or hospitalized) cases is multiplied in a given period of time. The interval that I have chosen in the charts that I disseminate is 4 days (I will explain why 4 days in another space). It is important to highlight that it is a matter of relating the increase in total cases, from the beginning of the epidemic. For example, the total cases registered in Spain on April 1 were 102,137. Four days earlier, on March 28, there were 72,248 cases registered. So, the case multiplication rate as of April 1 is Ct / Ct-4; 102,137 / 72,248 = 1.41. The cases have multiplied by 1.41, they have increased by 41%. If it is 2.0, they are multiplied by two every four days. If it is 1.0 (exact), in the last four days there have been no new cases (Ct / Ct-4 = 1; Ct = Ct-4), the expansion of the epidemic has been entirely cut off, as regards cases of declared new infections.

The epidemic, all epidemic diffusion moves in cycle: implantation, development, growth, containment, deactivation, or in the technical way that they receive from epidemiology experts (among whom I am not). The dynamics of the cycle, reflected in the evolution of daily cases, is the elephant within the boa.



All right. The problem with the cycle is that appearances are deceiving, but not systematically, but in two critical phases of the cycle: in the initial propagation (Phase 1) and in the zone where the maximum number of cases is reached, the famous peak (Phase 3).

In the initial development part (Phase 1), the appearance is that things are going well, because we are talking about tens or a few hundred cases, with no deaths. Insignificant, negligible thing. And so it has been despised country after country, despite having the message of terror from other countries that were advanced in the internal development of the epidemic. The appearance is that everything is going well, that there are so few cases that everything is controlled, because they are detected above. But the reality is that things are going wrong, because inaction allows the bug to spread and multiply at will, at its natural rate. No one slows it down. And it is the background noise, which is almost invisible, but it is real: contagion multiplies every two to three days. Against small, insignificant total numbers, yes. They are the snowball that will make it more or less giant in the end. The reality is bad.

Then when the cases are fired (Phase 2) there is no longer a divorce: the appearance is really bad, and so is the reality. The cases continue to multiply, which are already beginning to be noticeable, and the first measures are taken to limit contagion, but the effects translate into nothing. At this stage it is very easy to take containment measures against the spread of the epidemic, because everyone sees the catastrophic consequences. But the great thing would have been to take those steps before it explodes in the face, the fastest in phase 1.

In phase 3 there is again a problem of divorce between appearance and reality, because the containment measures, private and public, are already beginning to have a real impact in reducing the rate of virus transmission, but the appearance is Dantesque, because they rise cases. This divorce between appearance and reality is also disastrous, because it consumes energy, the hope of those who suffer the consequences of confinement, casts doubt on the measures and puts pressure to take more measures like crazy.

But the situation is about to turn around. And in the last phase (Phase 4), new cases fall day by day, and then a little later, that of the deceased. Things seem to get better, and they actually do.

Well, it is very difficult to see that things go wrong in the initial phase, and that they begin to go well in phase 3 of change, if one looks at the graphs of daily data and accumulated data (both in linear or logarithmic format) .

To illustrate the dynamics of the epidemic, I take the case of the epidemic and closed in the first wave, from Wuhan-Hubei. The first graph is the one of the new daily contagion cases reported, from the day that there are 100 counted cases. The jump in data observed from the 24th is artificial: it is because from that day on, not only the cases of the macro city of Wuhan, but also those of the region, Hubei, are added to the statistics.

It is difficult to see, in any of the 3 graphs, that the situation in Phase 1 is bad and that the situation in Phase 3 is good (something if you see it in the first daily data graph).

Instead, that information, both that things are going wrong in the initial phase (Phase 1), and that things are improving in transition phase 3, they are in the multiplication rate charts and the like.

That is, the multiplication rate (and similar relative measures) manage to match well, for each of the stages, the visual information with the reality of the dynamics of the epidemic, of reality, whether it is already apparent or hidden. . Therefore, it illuminates the experts' speech and makes it understood.

Let's go back to the multiplication rate graph. In phase 1 (orange), although the cases are few, the indicator alerts you that the situation is out of control: the cases multiply by 4! every 4 days. It is a clear and resounding message, which should help to take action.

In phase 2 (red), of already visible and constant growth of cases and deaths, people understand that 1000 is bad, and that if you have 3,000 four days later, we have a serious problem. Politicians and the public are ready to take action. Increasing absolute numbers help convey a sense of emergency and therefore urgency.

Does the multiplier drop in this explosion phase? Yes, because people already started shaking hands less and began washing their hands more. But the message of the rate is not that it is going down, it is that the multiple is still between 2 and 3. If there is no reaction, the total figures will continue to multiply, explosively.

And it reacts. In individual behavior and in public measures of mobility restrictions and social contact. And the days go by, and it seems that it is useless because the dead continue to grow. But yes, it happens. The multiplier already falls below 2. It is Phase 3 (yellow). It is in this phase 3 where the multiplication rate again plays a role that can be key in crisis communication: to explain that yes, that the measures have and their expected impact, despite appearances. In this phase you have to drill less with the death records. Not because the death toll and its increasing number is a lie, but because that is only part of the reality of what is happening. Because it is really happening: the change of rhythm announces a change of cycle.

And I remain quite convinced that transmitting that idea, that the dynamics are changing, is really difficult to transmit with data and graphics in absolute terms. And that "trust us, we are better" does not work either. People need to see. And what the multiplication rate tells us, or the measures of percentage increases, is that you can see it. That this tool exists, that it helps to detach itself from appearances. That the rate is not a trick of manipulation, but on the contrary is a way to touch reality.

We are all learning everything with this pandemic, its impact and its management. Cross fertilization of knowledge and talents is taking place in a manner and with prodigious speed and flexibility. It must also be done in how to communicate the crisis.

Francesc, are we in Phase 3? I think so. It is precisely what epidemiologists say (pending that yes of the evolution in Catalonia). I do not know how far we are from Phase 2 and how close to Phase 4. As we are probably in Phase 3, it is time to put the indicators of relative variation of the epidemic into action, because they can help the population (and the press and political leaders) to understand what is happening, so as not to become more psychologically overwhelmed.

And there will come a phase 4 (green), in which the multiplication rate will get closer and closer to 1.0, to the goal (I remember that the rise here is because they added cases from the entire province). But in this phase 4 the multiplication rate will no longer play any strategic role, because the appearance will align itself again with reality: each day the new cases and deaths are lower than the previous day.

Thanks Spanishman. That’s an interesting read (if you’re a bit nerdy like me anyway).

He’s using multiplication rate over 4 days to show trend, whereas I use the average over the past 4 days. They show slightly different things but both measure trend and, as discussed elsewhere on this thread, it is the trend which is far more important than any individual number.

By any metric, Spain’s trend on both infections and deaths is now downwards. Long may that continue.

However, as I mentioned to HolgateCorner, reaching the peak is not the same as defeating this virus. The death rate appears to remain high for a significant time after the peak has been reached.

Not certain of the reason, but I suspect there are significant numbers of very ill people remaining in critical care units.
 
Thanks Spanishman. That’s an interesting read (if you’re a bit nerdy like me anyway).

He’s using multiplication rate over 4 days to show trend, whereas I use the average over the past 4 days. They show slightly different things but both measure trend and, as discussed elsewhere on this thread, it is the trend which is far more important than any individual number.

By any metric, Spain’s trend on both infections and deaths is now downwards. Long may that continue.

However, as I mentioned to HolgateCorner, reaching the peak is not the same as defeating this virus. The death rate appears to remain high for a significant time after the peak has been reached.

Not certain of the reason, but I suspect there are significant numbers of very ill people remaining in critical care units.
Ta for that. Have added the diagrams into the translation. For the benefit of the nerdy few that might work their way through it.
 
As of 9am on 10 April, 316,836 tests have concluded across the UK, with 19,116 tests carried out on 9 April. Some individuals are tested more than once for clinical reasons.

256,605 people have been tested, of whom 73,758 tested positive.

As of 5pm on 9 April, of those hospitalised in the UK who tested positive for coronavirus, 8,958 have died.


13,543 people tested.
 
Last edited:
Today's headline analysis:

• Number of new infections in 24-hour period grows by 5,195 from 4,344 increase yesterday
• Rate of growth in 4-day average for new infections increases by 8.1%, having decreased by 8.3% yesterday
• Highest number of new deaths in 24-hour period (980), up from 881 yesterday
• 4-day average for new deaths grows by 17.8%, up from 9.3% yesterday and the fourth consecutive day of growth
• We continue to track further above Italy on a days since 20th death basis. We have followed broadly the same trajectory for the past 26 days and have been tracking above Italy for the past 9 days.
• We are now approximately 13 days behind Italy’s numbers (from 13.5 days behind yesterday and 14 days behind two days ago)
 
NB - on the positive tests above, I have used the 5,195 figure from the notes in the tweet by DHSC. That is the figure which excludes commercial swab testing for key workers and their families. I have done so partly for consistency with previous figures and partly because I can't get their numbers to add up otherwise.
 
NB - on the positive tests above, I have used the 5,195 figure from the notes in the tweet by DHSC. That is the figure which excludes commercial swab testing for key workers and their families. I have done so partly for consistency with previous figures and partly because I can't get their numbers to add up otherwise.
I was surprised how many NHS positives there are. I thought a lot might be just similar symptom isolation.
 
I was surprised how many NHS positives there are. I thought a lot might be just similar symptom isolation.

Yes, it does seem high. Also, the tweet states that a total of 73,758 people have tested positive which, they say, is a daily rise of 5,706. However, yesterday's total was 65,077 which is a difference of 8,681 so I think they must have redistributed some of those across previous days.
 
Back
Top