Ground Zero demonstration of Catalonia



Preamble

Many comments indicate that the fact that we mix an interesting analysis with a hypothesis about the non-actions of the Govern, reduces the strength of the manifesto.
Our sole objective is to analyse what happened in the 29-F demonstration and its consequences.

During the work on this analysis, the coincidence of dates between the communication of the Govern and some key points have seemed suspicious to us, and we have decided to publish our doubts.

We want to make our reflection on a possible contagion on 29-F reach the maximum number of people, and we understand that our hypotheses 'prevent' many people from seeing this purpose, and think of the whole as 'pamphlet'.
Each one is responsible enough to make his own opinion on the non-actions of the Govern.
So we have withdrawn all reference to hypotheses, and noted only the particularities of the communication of the Govern on these key dates.

We have decided not to make linguistic correction of our analysis, aware of our approximate Spanish ;-)
So this way it's obvious that we are neither Spanish nor Catalan, but European residents in Catalonia.
It allows us to maintain a certain distance, in a matter that takes on many people an ideological aspect, and that's not our objective.



At the end of March, we 'squared' some essential data, pointing to a massive contagion in the manifestation of 29-F (29th of February) in Perpignan and its consequent explosion of deaths throughout Catalonia, 17 days later .
On March 31st, we released our manifesto: CatCovidTransparencia_Manifiesto

It took us over a month to research and complete the analysis we present today.

There are 40 pages, 20 graphics and thousands of data.
The entire database is Open-source on Github, verifiable and with links to the sources:
CatCovid_Data

Among other data, we have compiled the historic of deaths for March by Health Regions.
We are the only ones who have those datas , ... with the catalan government who is still not making it public till today.

We compare Catalonia with 82 European regions and 264 countries.

The essential is summarized in the first chapter: (if you have to read only this ...)

A1. Summary


Now that we enter a less urgent phase, we think it's time to ask ourselves why 'it has gone so badly', why Catalonia is still having an high deaths count every day.
In this analysis, you will find a large part of the answer.

In the following 15 chapters, we demonstrate that :

• a synchronised explosion of deaths happened on march 17th in ALL Catalonia.
this synchronisation of ALL Catalonia, from march 17 to 27th, created a world record of death grow in 10 days.
• that this exceptional synchronisation comes from a massive precise contagion in the 29-F Demonstration.

We think that if the Government made public the data of deaths by Health Regions from March 17 to 27, we could demonstrate this link with official numbers.
Those are the numbers we compiled and present today.

In our analysis, we conclude that the 29-F demonstration was the main factor in the atypical epidemiological trajectory of Catalonia.

In ALL this analysis, every time we point to a negligence, omission or responsibility of Gencat, we make it very clear that it is an hypothesis.

Our ONLY aim is transparency and democratic accountability.
We DO NOT want to punish a collective.
Any input, clarification of data, and analysis is welcome.

CatCovid Transparency

You can sign our petition at:
Change.org/p/cataluña-covid-transparencia-de-la-gencat

You can re-tweet our post:
https://twitter.com/CatCovidTransp/status/1258829825122369538



Democratic accountability and transparency


How can it be that we are the only ones who have looked for the data that reveals the synchronized explosion of deaths in all of Catalonia? (Analyzing 137 articles in the local press, it was not so difficult.)
Girona, Lleida, Tarragona and Osona had their first death on the SAME day March 17th, and Barcelona exploded from 6 to 23 deaths.

Note for our english readers :
The Catalan government (Gencat or Generalitat or Govern) was one of the main promoter of this huge demonstration in Perpignan the 29th of February (29-F).
It was very important in their political agenda, to show that the nationalist movement was still able to gather many supporters, 2 years after the independence intent of October 2017.
Therefore, it's a political suicide for the catalan government to admit and announce to its followers and the population of Catalonia, that the event it promoted weeks before turned out to be a biological bomb.


As we explain on March 31, it is the responsibility of Catalan society and its 'actors' to investigate the possible consequences of the 29-F demonstration .

Faced with its main outbreaks that occurred in the same period (carnival, congregations, events, ...), France, Germany, Korea and others, took immediate action, and have already published complete scientific studies.
They considered these outbreaks very serious, although they do not reach a quarter of what we estimate to be the 29-F outbreak.

As European citizens, we find it very worrying , that NO catalan epidemiologist, scientist, journalist, politician, even question the issue of 29-F. At least publicly. (Apart from e-Noticies and El Confidencial)

Is there a form of self-censorship and acceptance of "not now" about confronting the Govern over the consequences of the demonstration?
The priority is not the health of the population?
Will epidemiologists and journalists from 'outside' have to come to investigate the 29-F, because the 'actors' of the catalan society are unable to ask for a democratic behavior of transparency?

Although we received thousands of visits, after our manifesto, we found that, apart from e-Noticies (Conexión Perpiñán. 21000 views), no contacted catalan politician, journalist or scientist has responded .
So, for the record, we will keep a chapter with eventual comments of the persons in charge.

On the other hand, this analysis is in Spanish and translated into English, French and German (in translation).


Force to the health workers, to the older people and everybody, in that difficult moment.


Notes 48h after the presentation of the analysis


We want to thank Xavier Rius from e-Notícies for giving (much) visibility to our analysis:
The Perpignan rally spread the pandemic throughout Catalonia . 50510 views. versió en català

38 days after our March 31 Manifesto:
Perpignan Connection . 21670 views. versió en català

On March 31, we tried to communicate with some catalan medias or politics, but none responded. Apart from e-Notícies.

We want to communicate that of the first many visits to this page, the average duration of the visit has been almost 6 minutes.
It is the most important thing for us, it means that many people have taken the time to read our reasoning. Many people also visited our Open-source database.

As specified in the manifest, our goal is, and remains:
• Request an official investigation into a possible contagion on 29-F.
• Ask the Generalitat to publish the deaths by health regions from March 17 to 27.



We have received many comments and questions that we find legitimate.
A few contradictions from people who had not read the analysis, and very few insults.
As was inevitable, social media groups misuse our analysis. We can't avoid it.

So we have created a chapter at the end to clarify some points regarding that:
This analysis is not a report, and part of it is analysis and part of it is hypothesis.
We also explain who we are and how we have developed this document:

E4. Notes and context





Index



Introduction:

A1. Summary

A2. Current situation

A3. Contagion chronology

A4. How we discovered the contagion


Uniques datas:

B1. Exceptional epidemiological leap

B2. Deflagration of 10 days without comparison

B3. Synchronized explosion throughout Catalonia

B4. Distribution similar to the buses


Impact assessment:

C1. Mass contagion on 29-F

C2. Curve modeling

C3. Estimation of deaths related with the 29-F

C4. Summary simulation of contagion


Political and health action:

D1. Easy confirmation with an investigation of 104 deaths

D2. Chronology of the possible non-action of the Catalan Government

D3. Action of other governments against outbreaks


Annexes:

E1. People contacted in Catalonia

E2. International Contacts

E3. Contact and references

E4. Notes and context





Introduction



A1. Summary

ALL the results in this summary are extensively detailed and calculated in the following 15 chapters.
And what are estimates and hypothesis are this, estimates and hypothesis.

We explain in chapter A2_b why the numbers of deaths in hospitals and ICU are the base data for our analysis . All of these numbers come from official sources.
And why we relate a death rise to a previous contagion rise.

In our analysis, we sometimes use the entity Cat-sin-Igua (Catalonia without Igualada).
It is simply the deaths in Catalonia, minus the deaths in Igualada , 2 numbers that come from the Gencat communications during the month of March.
It allows us to 'isolate' the small Igualada outbreak, that ocurred earlier. (La Conca d'Odena with 70,000 inhabitants).

Thus, with Cat-sin-Igua , we can more accurately measure the impact of the demonstration on the rest of Catalonia .
As you will see below, the outbreak of contagion at the demonstration was, for Catalonia, a much larger outbreak than the small Igualada outbreak.

Clarifications

• We are not saying that the entire epidemic in Catalonia comes from the demonstration. We assess the impact of the demonstration .
• We are not saying that the city of Perpignan was massively infected by 29-F. The demonstration was a fairly closed event .
Our aim is not a race between Cataluña and Madrid. Both are going through an exceptional situation, each one with its own factors.
We are not an official investigation, although we have a solid scientific background. Our purpose is to disclose verifiable data and analysis to the public. We request an official investigation.
• We have decided to adopt a 'general public' presentation of the data. In order to reach the maximum number of people, the results of some calculations beeing sometimes a bit complicated. (Chapter C2)
Our estimates are ... estimates, but they are not speculations . They come from mathematical calculations and statistical analysis, and whenever a doubt arises, we choose the scenario that is the least favorable for our hypothesis.

Five indisputable data


Two data come strictly from:
• numbers of deceased and ICU given by the Govern . ( CatCovid_sources & Gencat_notes-premsa )
• compared to 5000 days / deceased data from 264 countries/regions ( CatCovid_Data / World_series & JHU-CSSE ) and 83 European regions ( CatCovid_Data / EU-Reg_deaths-hosp_series )

On March 17th, Catalonia achieved 2 World Records :

• Highest Increase in deaths in 1 day (128%) . Going from 18 to 41 †.
This increase is confirmed by the UCI jump on March 9. Going from 5 to 18 ICU.


• Highest rise over the next 10 days (2500%) . Going from 41 to 1070 †.
Catalonia reaches over 1,000 deaths faster than New York, Madrid and Lombardia.
All the rest of the countries/regions of the world do not reach half the speed of Catalonia (below 1200%).

Figure 2

Compiling 137 articles from the local press :
• Synchronization of the whole of Catalonia in the explosion of deaths on March 17th.
Although it seems incredible, Girona, Lleida, Central and Tarragona have their first death on day 17.
Barcelona explodes on the 17th too (from 6 to 23 †). First dead in Ebre and Perpignan on the 19th.

Graphic 3

Chart 4. Links: Press / First_dead

Gathering information from the local ANC on 429 buses of the 583 official ones:
• Distribution of the first 320 deaths similar to the distribution of ANC buses.
El Vallés Occidental (70 buses), Barcelona (57), Osona (34) have many buses.
Tarragona and Ebre have few buses.

Graphic 5

According to studies published in The Lancet (see article: Covid-19 periods )
• 17 days is the average period between infection and a sharp rise in deaths.
It is exactly the time between the demonstration of 29-F and the explosion of deaths that begins on March 17th.

Graphic 6


In our analyzes and simulations, we used the entity Cat-sin-Igua (Catalonia without the deceased of Igualada), to measure more precisely the impact of the demonstration on the tragic trajectory of the deceased in Catalonia.
Cat-sin-Igua is out of reach:
350% increase on March 17 . That is, it is 3x times stronger than any other record in another country / region.
3650% rise for the next 10 days .

Conclusions summarized

Next, in this analysis:

We prove that:
The Perpignan demonstration of 29-F was one of the largest coronavirus accelerators and dispersers in the world.
The most terrible thing is that it was massive and occurred at the beginning of the epidemic in Catalonia, multiplying its effects .
The 10-day blast brought Catalonia to over 1,000 deaths.

We prove that :
The synchronization of all Catalonia is the cause of the deaths grow record in 10 days.
And that this outstanding synchronization comes from the 29-F demonstration.


We estimate that (hypothesis):
Note: We see that some internet sites published what is an estimate as a statement. We have clearly explained, from the publication, of this analysis, in chapter C3, that it is an hypothesis for a simulation.
The contagion explosion of 29-F could be the origin of more than half of the deaths in Catalonia.
A deflagration with long-term effects , which propelled Catalonia in a critical dynamic, until today, the beginning of May . And that it exploded in a segment of the elderly population, very fragile against the coronavirus .

We propose:
An easy confirmation of the massive contagion on 29-F with an investigation of 104 deaths.
Of the 104 official deaths from March 17 to 20, with a proof that at least 12 deaths had a DIRECT relationship with the demonstration, our hypothesis would be confirmed.

Possible non-action by the Govern

We have decided to record the dates in which, in our opinion, the Government could have detected a massive contagion on 29-F and its communication and reaction this day.
We are aware that all governments experienced an exceptional situation that makes it very difficult at the time to "read" the data.

• As soon as March 9 , a major ICU (Intensive Care Units) jump occurs throughout Catalonia (from 5 to 18). It is a clear indication that a significant contagion had occurred on 29-F, since it corresponds to the period between contagion and ICU of The Lancet study. .
This day was the only day that the Govern did not communicate the ICU number.

From March 12 to 16 , there is a very rapid rise in ICU in all of Catalonia (from 23 to 66 ICUs), when in Igualada the ICUs rise only from 3 to 8 ICU. For us, this rise is the wave of deaths that arrives.
We think that the Govern could see this rapid increase these days.
It would explain the total change of comunication of the Govern between day 11 and day 13.


• On March 17, a synchronized explosion of the first deaths occurs throughout Catalonia (jump from 18 to 41 †, with 23 new deaths ). With one of the world's largest increase in all data recorded so far.
The death jump is so strong, that the ICUs go down.
The Govern publishes an 11-line statement , with no mention of the sharp rise and first death in every health region. It only communicates about the Igualada outbreak (2 new deaths).
Gencat_Comunicat_17-March

The Govern stopped giving details on the place of death for 1 month . As of today, it does not give the data for this period.

• In the following 10 days, Catalonia experienced one of the largest worldwide acceleration of deceased ever recorded . The Govern did not comment anything.

• On March 20, 3 days after the explosion, he announces that he has investigated the Igualada outbreak.
He claims the contagion comes from an 80-person dinner on February 28, when the next day 600 protesters from Igualada (12 buses) spent hours together, in what studies describe as a mouthpiece of contagion.

• 1 month later, on April 16, the govern starts again to give the numbers of deaths by health regions.


The explosion summarized

We are the only ones that have compiled the data on deaths by Sanitary Regions (RegSan) from March 17th to 27th, ... with the Gencat not making them public till today.
We have put this database on Open Source, with links to the 137 local press article reporting the deceased in each hospital.
CatCovid_Data / Cat-RegSan / Units

On March 17, we see how the curves of:
Barcelona, ​​Girona and Central-sin-Igua take off. Girona increases by 100% on the 23rd.
Lleida-Pireneu and Igualada climb very fast too.
While Tarragona, Ebre and Perpignan, although synchronized, appear less affected.


Let's try to illustrate what happened ...

After the explosion of contagion in the demonstration, during the 10 days of the blast ,
Catalonia behaved, like a "dense unified Virtual City" , similar to New York or Madrid.

• The ' big neighborhoods of this virtual city' were the areas of Catalonia where more protesters were coming from , such as el Vallés, Barcelona, ​​Osona, Girona, Bagés . (See graph 5. ANC - Buses). Not necessarily the most densely populated.
• The ' small neighborhoods of this virtual city' were the areas in Catalonia of where LESS protesters were coming from , such as Tarragona and Ebre .

The contagion was so strong and punctual that "those big virtual neighborhoods" had a rise in deaths in the first days comparable to entire cities (much denser and larger) such as NY, Madrid or the city itself Barcelona.

(We have lined up around 3 to 4 deaths to compare the rise of the 'neighborhoods'. Barcelona is aligned at the beginning of its explosion 6 †, to compare with NY and Madrid, the 2 other largest deaths rise in the world)

in 5 days, the ' small neighborhoods ':
• Ebre goes from 4 to 9 †
• Tarragona goes from 3 to 16 †

in 5 days, the ' big neighborhoods ' of the demonstration:
• Girona goes from 4 to 37 †
• Central Catalonia goes from 3 to 30 †
• Lleida goes from 3 to 27 †

• Barcelona goes from 6 to 132 †

in 5 days, at the same period of 'their' epidemic:
• Madrid goes from 8 to 81 †
• New York goes from 5 to 84 †

This behavior and acceleration of the 'interior' Catalonia in 5 days is totally 'abnormal' and impossible in an epidemic in areas with less / low population density.

Barcelona (RegSan) with a density 5 times lower, exceeds New York City by far. (Barcelona 2008 hab/km2, New York City 10647 inhab/km2).

After a week, these areas of 'inland' Catalonia quickly returned to 'normal' behavior . The punctual contribution of hundreds of infected in 1 day, was blurred with the absence of a dense and important population.
They continued having many daily deaths, due to the high level of contagion reached in just one day, but the reproductive number was low.

Barcelona (RegSan), as a very dense urban area, continued with a very high contagion (and death) logic.

We think that some consequences of this explosion remain until today . This wave was so strong that it reached many corners of these 'neighborhoods'.
We think that contagion by protesters entered the nursing homes as soon as March 3 . The current situation is partly the result of this early and strong entry.
It could explain that areas like Girona maintain a high level of ICU. Or that remote areas of Osona have an infected rate equal to Madrid and Barcelona.
That as of May 5, Catalonia continues to have close to 100 hospital deaths a day . As much as the rest of Spain, or half of France (8 times bigger).
Surely, this deflagration will have as consequence that Catalonia will take longer to get out of confinement .
And it makes a second wave more likely, because the contagion reached a high level of diffusion.


In Madrid, the rise in deaths took 6 days to extend to the 2 Castillas. And it came, and very strong.

In Catalonia, the rise in deaths is synchronized on March 17th.
In Barcelona and in others sub regions like Girona, Central Catalonia, Lleida and Tarragona.
It did not come from Barcelona to the interior area.

The huge and synchronized outbreak in Catalonia came from the 'unifying center of this virtual city', the demonstration of 29-F.
In this virtual center that was the demonstration, contagious people (coming from all over Catalonia and Igualada), infected many demonstrators, who distributed themselves the SAME day, February 29th, in the 'virtual neighborhoods' of Catalonia .
... and caused, 2 weeks later, the synchronized explosion of deaths, the SAME day, March 17th , in all of Catalonia.

We think that if the Government published the data on deaths by Health Regions from March 17 to 27, this synchronization could be demonstrated with official numbers.

Unfortunately, it can be summed up so simply.
You will find all this extensively developed and calculated throughout this analysis.


Note: This exact same 'abnormal' behavior happened in Bergamo and Lombardia. With the explosion of contagion among his followers at the soccer game on February 17th.
But instead of being spread over an entire region like Catalonia, they were largely spread over a much smaller area, Bergamo.
Bergamo had for weeks, a behavior similar to urban areas much larger.
And Milano (Lombardia) shot out too.



A2. Current situation

Catalonia is the third worst area in Europe

(data as of May 1: CatCovid_Data / EU-Reg_deaths-hosp_series )

It is not true that many other countries are doing it 'wrong'.
Catalonia is experiencing an exceptional epidemic. Third in Europe.
In death rate per million inhabitants. (in hospitals, see next chapter)

There are only 3 serious areas in Europe (in red), which have been overwhelmed :
• Lombardia and Madrid are extreme, dragging their bordering regions (two Castillas).
• Catalonia is the third, very serious. With its own trajectory.

Then come ONLY 5 moderately severe areas (in blue), which have never been overwhelmed :
• Basque Country, Navarra and little Rioja.
• Grand-Est and Paris, with 68% of the severity of Catalonia.
• London and North-West, with 73% of the severity of Catalonia.
(As of May 5, London is still below Catalonia.)

All the rest of the European regions have less than half of Catalonia's mortality rate. (in green and gray)
The vast majority, less than a quarter . (in gray)
Occitanie, next door, does not reach 10% of the rate in Catalonia. Belgium reaches 40% of Catalonia, etc ...
And no, the other countries have not hidden 50% of deaths in hospitals.


With a map of total deaths (including nursing homes), Catalonia is even worse, comparatively.
Catalonia (10401 † per 7.67 M inhab) has a rate of 1350, twice that of Belgium (7703 † per 11.46 M inhab) with 675.

In addition, areas of Catalonia, such as Vallés Occidental, Osonès, Bagès or Barcelona , have a death rate, similar to Madrid, and more than:
• 2 times that of Paris or London.
• 7 times that of Rome, Andalusia, Lyon or Marseille.
• 12 times that of Naples.

Grand Est is the worst region in France, because it had an outbreak during a congregation of 2,500 people.
Lombardia, because it had an explosion in the 40,000 people soccer match.
Madrid, ... is an explosion in itself and has to be investigated.
The Generalitat has to assume that it has had an explosion among 75,000 people in the Perpignan demonstration .

We use an estimate of 75,000 people in the Perpignan demonstration. In the end, it doesn't matter if it's 75,000 or 150,000 people. What we analyzed is how many people were infected this day.

Deaths in hospitals and ICU

The only correct comparison data is deaths in hospitals .
(data as of May 1)
Because it is used by WHO , UK (23599), Germany (6575), France (15310), Italy (28236), Belgium (3517), Spain (25950) and all other European countries. < br /> It is the only data they also use to give regional deaths , and to calculate the evolution of the epidemic.
(data as of May 1: CatCovid_Data / EU -Reg / Units_series )

It is this number that we use for Catalonia (5987 † as of May 01).

We did not use the number of only hospitals-checked (5137), because we consider that Catalonia and Madrid have been overwhelmed and have not been able to verify all deaths in hospitals with tests . But that those found in hospitals fall within a count similar to France, Belgium, Germany or England, which have proven almost all deaths in hospitals.
To be correct, Madrid (which does not include non-confirmed hospitals) would have to be re-evaluated from 10 to 20%. But in the previous graph, our objective is to compare Catalonia with the rest of Europe.

Then, in order not to be accused of 'hiding the dead', some countries give a number of estimates of total deaths with very different criteria, including geriatric . France (24563), UK (27510), Belgium (7703) and Catalonia (10401), for example. It is the number that appears in the newspapers.
A phase will come where we can work on more reliable numbers of funeral homes.


We always relate death rises to a previous contagion rise. (it starts around 17 days after contagion)
This data and this close relationship is proven and is the basis of the most extensive latest studies .
Imperial College - Estimating the number of infections

The numbers of UCI are our other reliable confirmation data. (it starts around 9 days after contagion)
The government has just announced that it is the main evaluation data for softening the confinement.
El País - ICU, the only unchangeable requirement

Tangled Gencat deaths count

Since April 17th, Gencat provides a daily table with counts by Health Region, and gives different types of total count of deaths ('Exitus'). ( Gencat-Prensa_25 / 04 / 20_Table )
Gencat meets the challenge of giving 11 different types of total death count, without ever giving the dead in hospitals confirmed with Covid-19.(which is the reference number for all countries)

For our part, we will not enter into this 'game' of the Gencat, and we continue to use the only continuous data that Gencat has provided from the beginning (deaths in hospitals, confirmed and not).

And our analysis does not change, because we work to evaluate the part of the coronavirus epidemic in Catalonia that can be attributed to the manifestation of Perpignan , and because the change in criteria occurred after the period of our analysis.



A3. Contagion chronology (hypothesis)

We are developing a complete simulation of the contagion, before, during and after the demonstration of 29-F.

Meanwhile, our summary timeline for March 31 is still valid, 1 month later.
In addition, it has been validated throughout the month of April with new official data, in particular on nursing homes.

We predicted, at the end of March, that the advanced age of many protesters (55 to 70 years old) led to a significant contagion of nursing homes in the days after the demonstration (from March 3) , when the protesters visited their relatives (80 to 95 years old).

15 days later, our hypothesis is confirmed, many nursing homes were infected in the 11 days BEFORE confinement , from March 3 to 14.
A desperate situation was created in a closed environment, with the age range (80 to 95 years) the most affected by the fatality of Covid-19. With a first official death on March 15 , possibly infected by a protester right after the demonstration.

As a consequence, Catalonia now has a contagion rate and a drip of deaths in nursing homes that does not stop.

You can see this sequence explained, in our timeline for March 31:

Graphic 2



A4. How we discovered the contagion of Perpignan

5 factors have been key.

Analyze the numbers of deceased

We have succeeded in focusing our analysis on the number of deaths in hospitals, only valid data in the Catalan context . The infected data depend too much on the number of tests.

Cat-sin-Igua

(dead in Catalonia less dead in Igualada)
Although the communication from the Govern was more than scarce, I provide data on the outbreak in Igualada. Very soon it allowed us to separate the reliable data from Igualada from the rest of Catalonia , and thus calculate Cat-sin-Igua .
It is very important to isolate the data of the specific Igualada outbreak , to analyze more precisely the impact of the demonstration on the development of the epidemic in the rest of Catalonia.

Collection for each Sanitary Region (RegSan)

For each RegSan in Catalonia, we have been able to reconstruct the number of deaths per day, analyzing data from 137 articles in the local press .

Compare with other regions of similar size

We compare Catalonia and Cat-sin-Igua with the dead of Madrid, Lombardia, Occitanie, Euskadi, Gran-Est, Hubei, so 82 European regions and 264 countries/regions.

Calculate daily increase

With these data, we create logarithmic curves every day of the evolution of the deceased, comparing these regions with Catalonia and Cat-sin-Igua.
We also calculate the curves of the daily increase (in%) of deaths (the growth rate) .



Uniques datas



B1. An exceptional epidemiological leap

Ten days after confinement (March 24th), it still seemed strange to us, this peak of increase in deaths on March 17, of 128% in Catalonia and 350% in Cat-sin-Igua , given that it did not exist for other highly affected regions such as Madrid or Lombardia.
And also it had not happened with a jump of 2 to 5 deads, but from 18 to 41 deads, which eliminated the possibility of a variation due to small numbers.

We decided to check if there was a similar peak in other countries.
And the graph of the WORLD RECORD of 264 countries appeared.

... Something very strange had happened this day in Catalonia.


Mathematical note: Obviously, at first, we 'softed' the curve, applying weighted average calculus, but even so, the peak was exceptional. Until we realized it was not a statistical deviation. It was the precise record, in 1 day, of a massive jump.

An exceptional death jump

Cat-sin-Igua goes from having less than 1 death a day, to having dozens a day, after march 17th.
In the following curve, with the same data presented in logarithmic number of deaths , we can see 2 fundamental data.

Compared to 264 countries/regions, Madrid and Lombardia:
• Catalonia and Cat-sin-Igua have an exceptional bend turn .
• Catalonia JUMPS from a low epidemic to a high epidemic.



Confirmation with ICUs

ICUs (ICU in English, 'molt greu' for Gencat) are the only other reliable data , with deaths in hospital.
Gencat sources for Catalonia and Igualada (and Cat-sin-Igua), and we added the ICU numbers we collected by sanitary regions: CatCovid_Data / ICU


UCI jump on March 9 or 10

Catalonia jumps from 5 ICUs on March 8 to 18 ICUs, 2 days later. A break that we have interpolated in the graph (because Gencat does not give ICU this day) and that gives us an increase of 220% (from 5 to 16).
An exceptional jump similar to the death toll on March 17.

These 10 days between contagion (in the demonstration) and the ICU jump correspond to the data from The Lancet studies that we have collected. (see next paragraph)

On March 9, Gencat does not communicate the UCI number. It is the only day that they have not communicated it.
Source: Comunicat_2020-03-09

UCI fast rise from 12th to 16th March

In 4 days, the ICU number quickly goes from 23 to 66 ICUs (outside of Igualada), and Igualada also goes from 2 to 8 ICUs. From the Gencat communications, we can separate the data.

For us, it shows that a 'wave' of deaths was arriving due to the demonstration.

UCI confirmation of March 17

This day comes the explosion of deaths.
The jump in deaths in Cat-sin-Igua (and Catalonia) is so important (from 6 to 27 †) that more ICU beds are released than are created .
We have not found in any other region or country a drop in ICU (and so big!) in the midst of the epidemic's rising phase.
Our estimate (continuing a continuous curve over the descent) is that 19 ICU beds were released this day (78 estimated - 59 real).
These 19 released ICUs correspond to the 21 new dead jump.

It is confirmed that the jump of deceased is exceptional, in 1 day and around 20 deaths.



B2. 10-day blast without comparison

It wasn't just a 1-day death jump.
We saw how, with a week of delay, the curve of Catalonia rose more vertical than Madrid, which itself had risen faster than Lombardia.

Below, we surely see the most remarkable data of all this analysis.
It demonstrates the exceptional strength that the massive contagion of 29-F had.

A deflagration with long-term effects , which propels Catalonia in a critical dynamic for the coming weeks, until today .

An acceleration of 10 days


In the following graph, we have simply calculated the increase in 10 days, aligning the countries above 40 deaths, which corresponds to Catalonia after the jump (from 41 to 1070 † in 10 days) .
Catalonia and Barcelona (RegSan) rise faster than any other country / region.

And also, they are homogeneous, with 2500% rise.
We have also calculated the increase in Catalonia with the numbers of Catalonia from the Spanish government source (confirmed deaths from 33 to 827 †), and almost the same exceptional increase comes out, 2400%.

ALL of Catalonia climbs extremely fast, and not only Barcelona. Cat-sin-Igua rise 2810% (de 40 a 1165 †)
(At the end of the Summary chapter, we illustratively explain what happened.)

The explosion is so strong, that over the next 10 days, it goes up much more than any other country or region . And it comes out of this acceleration above 1000 deads.

Graphic 2



B3. Synchronized explosion throughout Catalonia

Until March 17, only less than one death a day had occurred in Barcelonès. (Barcelona area)
No other health region (RegSan), had 1 dead, apart from Igualada.
The Catalan health system was on alert since the Igualada confinement on March 12th. So we think that only a very few deaths could have gone under the radar in the hospital system, until March 17th.

Collection of deaths for each Health Region (RegSan)

We have managed to reconstruct for each RegSan in Catalonia, the number of daily deaths, analyzing the reliable data of 137 articles in the local press .
Although the Generalitat stopped giving the death rates by health regions, the local newspapers continued to give the death count in each hospital, detected cases, and sometimes the ICU numbers.

All the collected history fits perfectly with the global sum of deaths given each night by Gencat.
We have put in Open Source, the database that we have recreated, with all the links to each press article and the Gencat communications.
CatCovid_Data / Cat-RegSan / Units

All RegSan (except Barcelona) have their first death on March 17th or 19th

Although it seems incredible, the timing is perfect . All RegSan are victims at the same time of the massive contagion from the 29-F Demo, and the consequent first deaths 17 days later. And in Perpignan too! (where the demo ocurred in France)
We have not found such a sync between its different sub-regions in other European regions.
It can only come from a common massive infection in one day.

Note: It may be that one of these deaths occurred on March 16th and was confirmed on March 17th. In the same way that The Lancet study gives 17 days as the period for deaths to rise sharply after infection, but It does not prevent that some deaths can occur on the day 15th or 16th.
What is conclusive is that deaths rise precisely in this statistical 'opening' of a few days.


You can download the map in GeoJSON format: Cat_RegSan_geo.json

The local press announcing the first death on March 17th


Here are the links: Press / First_Dead

Osona.com from March 18th points out that the first death is from the 17th.
The Tarragona newspaper of March 17 points to the first death the day before in the afternoon. After analyzing the way to collect data from Catsalut, this death belongs to the 17th. It seems that Catsalut collects data from noon to noon in RegSan, and then gathers them to announce them around 9pm (in this period).
The Tarragona newspaper of March 19th points to the first death in the Ebre area.
L'Agence Régionale de Santé d'Occitanie announces the first death in Pyrénées-Orientales on March 19th.

Gencat stops giving details on the deceased

On March 17th, Gencat stop giving the location of the dead by Sanitary Region. (23 new deaths)
And it publishes a 11-line statement , giving details only on the Igualada outbreak (2 new deaths).
Gencat_Comunicat_17-March

Instead of reporting an explosion of deaths in the rest of Catalonia (from 6 to 27 deaths) , and the first deaths in 4 health regions

For 1 month, Gencat has not given any details of the location of the deaths, apart from an oral reply to the National newspaper on March 24 and an attached communication on 26.
As of today, no info on this time period.

Synchronized curves in each Health Region

We see how the curves of:
Barcelona, ​​Girona and Central-sin-Igua take off. Girona increases by 100% on the 23rd.
Lleida-Pireneu, Tarragona and Igualada go up very fast.
While Tarragona, Ebre and Perpignan appear less affected.




B4. Bus-like layout

On March 25, we saw the map of the ANC bus departure points, and it seemed to us that Barcelonés, Gironés, and Central Catalonia concentrated the vast majority of these exit points.

But the Govern had stopped giving the location of the deaths since March 17.
After compiling the data of deaths by RegSan, and verifying that the deaths occurred mainly in these regions, we needed to know the number of buses by RegSan, to find a correlation.

Official distribution of deaths, 6 days after the explosion

On March 24 at noon, responding to a question from the National, Counselor Vergés gives a breakdown by Health Region of the previous day. (e-Noticies - Datos por Comarcas)

This breakdown corresponds EXACTLY to our numbers compiled in the local press on March 23.
Barcelona 239, Girona 28, Central 59 (Igualada 36 and Central-sin-Igua 23), Lleida-Pir 8, Tarragona 4, Ebre 1.

6 days after the explosion of deceased, we can calculate the mortality rate in each RegSan, with official numbers. We have not represented Igualada, because its mortality (51.4) was out of scale.
We can see that the rate is quite high in Barcelona, ​​Girona and Central-sin-Igua .


Distribution similar to ANC buses

(This chapter is under hypothesis until we put the links to the many tweets in our database)
This distribution of 'first' deaths after the explosion, corresponds to the departure of the protesters by RegSan , if we rely on the ANC coach map.
Keep in mind that Central-sin-Igua (441707 inhabitants) has half the population of Girona (839960).

We consider that the distribution of buses in Catalonia is a very good indicator of the origin of the protesters.
It can be estimated that one-third to one-half of the protesters came by bus (29,000 = 580 buses x 50 people), within a demonstration of 75,000 people.

We managed to collect 430 of the 583 buses declared by the ANC, waiting for the ANC to publish a full breakdown .
We have analyzed dozens of tweets, facebook and articles in which local ANC pointed out the number of buses that sent to Perpignan. We will soon put the many links in our database.

• The blue dots correspond to the departure locations of one or more buses. Source: ANC Autocars - Perpinyà 29F
• Large numbers in red, number of buses by RegSan (72% reliable estimate).
• The blank numbers correspond to the number of global buses for each region and come from the information of the local ANC.
• The yellow numbers are those that we have calculated with counties where there were many blue points but with bus data from only a few points (example BCN: Gracia 8, Meridiana 5, Sarria 5, St Andreu 4, etc ...).


We can clearly see that the 3 RegSan most affected by the first deaths are the ones with the most buses .

It is worth noting the Vallés Occidental (70), Barcelona Extra-radio, Osona (34), Girona (32) and Olot, with many buses regarding the size of their population.
Today they have positive rates per 100,000 of (860 to 1170) higher than Barcelona center (737).

Tarragona with much fewer buses than Central-sin-Igua (with the same population) has a positive rate of 336 and Tortosa of 230.
Gencat - Cases Map

The buses, a cluster of contagion

We will see in the next article (the simulation), that the ANC buses, not only (hypothesis):
• they have brought contagious and returned thousands of new infected in every corner of Catalonia .
• but that on the buses, where 1 or 2 contagious people got up at the beginning of the day, the bus came back largely infected .
You can read the summary of a study here:
South China Morning Post - Hunan contagion in bus
And an article on Igualada:
El Confidential - Los Igualada buses

We think that is what happened in Igualada, but also in many parts of Barcelona, ​​Girona and Central.
Our simulation shows that a city like:
• Vic was able to go from having less than 10 (real) cases of Coronavirus before the demonstration, to having 200 the following day. 20 times more.
• Igualada was able to go from having 100 (real) cases of Coronavirus before the demonstration, to having 500 the following day. 5 times more




Impact assessment



C1. Mass contagion on 29-F

We have seen in the 4 preceding chapters how Catalonia jumps into a rampant and synchronized epidemic on March 17.

We have used epidemic simulation tools for days, and it is absolutely impossible to reproduce this jump and the following acceleration without giving the algorithm input, a very specific contagion of thousands of people.

The 2 record increase in deaths means that:
On March 17, Catalonia registered the effect of a previous massive and punctual contagion .

17-day period between being infected and dying

Scientific studies and the article Medium-Coronavirus gave a period of 17.3 days on average between being infected and dying , with links to a compilation of studies on the coronavirus.

We decided to check this 17-day period , and we made a compilation of data from numerous scientific publications. A study by The Lancet confirmed that 17 days after the contagion, deaths skyrocket, creating a peak effect in case of massive contagion in a single day.
The interesting thing about this study is that it set this rise to 17 days for a focused group of people from 63 to 76 years, similar to those that interest us among the protesters.

You can find it in our article:
CatCovidTransparencia / Periodos del Covid-19


The massive jump came from the Perpignan demonstration, 17 days before.


All calculations are detailed in chapters C2 and C3 .
Here, let's try to be as 'big audience' as possible, we round the numbers, for easy reading.

The numbers given are indicative, of course, but ... .

It is known that it is very difficult to forecast the future evolution of an epidemic, particularly a new one such as the Coronavirus.
But, it is much 'easier' to do an ex-post analysis. Because the numbers already exist. This is our case, and the case of many studies. We are not doing foresight!

Simplified calculation of deflagration of deaths

Here we work on Cat-sin-Igua . Then we reinstate the deceased from Igualada.

We calculate that on March 17, a specific contagion generated a jump to a higher epidemic, a jump of 17 deaths in Cat-sin-Igua. (Chapters C2)
This calculated jump is only the part of the epidemic that 'corresponds' to the manifestation, 70% of the total Cat-sin-Igua epidemic . The rest of the 'normal' epidemic represents 30%. We are talking about the ratio after the jump. We are not talking about the proportion in early May with 6000 deaths in hospitals.

In the previous graph (periods), we see that, in a single day, 100 people infected (and who are going to die), generate 4.3% of deaths on day 17 (and 95.7% the following days).
We can calculate, in total, that the first wave of this contagion created nearly 400 deaths . (4.3% = 17 † >> 100% = 395 †). Later in the simulation, we will adjust this first calculation.

We can see in the following graph the sequence of the deflagration. It looks better because it is a daily growing representation of deaths.


This first wave of 400 deaths enters with a very high peak on March 17 (350%, from 6 to 27 † Cat-sin-Igua), and is distributed at 50% over the first 6 days, and the rest over 15 days (Periods graph).
A second wave arrives 3.5 days later , generated by the new ones infected by the protesters. ( infection-to-infection , rise between 3 and 4 days)
This second wave is more 'diluted' in time , because new infections do not occur in 1 day like 29-F, but between March 3 and 17. It causes the second death spike on March 20-21, much less strong (69%, 98 † Cat-sin-Igua).
A third 'wave' is delivered 7 days later , even more 'diluted' and distributed. On March 24 in the graph (57%, 475 † Cat-sin-Igua).
It should be noted that, although replicas / waves 2 and 3 seem small, they represent higher increases than any other country at this level of its epidemic. It is seen in the Graphic 1 .

Deflagration is the sum of these 3 waves.
It has lasted 10 days, and has brought Cat-sin-Igua to over 1,000 deaths. (From 6 dead to 1013 †)
At this level, the Igualada part (57 †) is very small and Catalonia has a total of 1070 † (from 18 †).

There were waves, because the contagion was punctual and strong. In a 'normal' epidemic development, there are no such marked waves, the rise is progressive.
An analogy is the effect of a stone thrown into the water that generates a first wave and attenuated replicas. Or a tsunami, created by the punctual rupture of the earth's plate.
It is a similar mathematical model, consequence of a strong punctual event.

(Prospective!) We estimate that of these first 1000 deaths, 70% come from the demonstration, and the remaining 30% from the epidemic that would have occurred 'normally'. (Chapter C3)

After 10 days, Catalonia returns to a more classical contagion reproduction rate (and death growth as well).
It is the phase we call 'Normal' because it has a curve similar to the normal behavior of other highly affected regions. Although already, with many dead and then infected.
It is seen as the increase falls very low between day 33 to 35, just 18 to 20 days after March 15 . It is the effect on deaths of the 'Confinement' phase . It comes later than 17 days because the general population is not 65 to 75 years old ...

The particular conditions of 29-F.

Before calculating how many infected the 29-F, generated these 400 deaths from the first wave, we want to propose a hypothesis.
This hypothesis is that the very advanced age of many protesters 'caused' a significant number of deaths in the first wave (400 †), without having to come from a huge number (x50) of infected .

We are working for the simulation, on photos where it is clearly seen that more than half of the protesters are between 60 and 75 years old. With older, higher mortality rate.
But, the studies that we consulted also point out that, as more advanced the age, the sooner the death occurs, but also, these deaths are more grouped in time.

We think that only a significant concentration of contagion, allied to an elderly population , could cause this acute and synchronized explosion of deaths.
A similar contagion in a much younger manifestation, such as 8-M, would have had a much more diluted effect on the time of death and a 5-10-fold lower mortality rate. (Chapter C2)

Our hypothesis is that, on March 17, 2 increases in deaths were added to reinforce this explosion of deaths:

• Protesters from 65 to 75 years old infected on February 29: rise in deaths after 17 days. With a 5% mortality rate.
• Protesters from 55 to 70 years old who infect their relatives / parents in nursing homes between March 3 and 7.
• These relatives in nursing homes, over 80 years old die 12 to 14 days later (faster due to their age). Then they also begin to die in an important way around March 17. With a very high mortality rate of 15% .
Covid-19_Fatality


We note that the pink curve is very similar to the jump in the logarithmic curve (graph 3) of those who died in Catalonia.

Once launched, this explosion of contagion continues to spread in Catalonia and geriatrics, and 17 days later deaths will rise very rapidly.

Protesters under the age of 65 had little chance of dying due to their age, but they were decisive in the spread of the coronavirus throughout Catalonia.

Simplified calculation of contagion

We have to separate the 400 dead from the first wave into 2 groups (protesters and nursing homes).
Our first simulations give a half / half working model. ('hits' the decease rise of the next 10 days)
We assume that 200 † come from protesters from 65 to 75 years old, and 200 † come from nursing homes.

We have the problem that in March, those who died in hospitals were counted and those in nursing homes were not.
At the moment, we are working with the deceased in hospitals, we are going to continue with this distribution and we will then extrapolate to the total number of deaths.
In this phase, we just want to get an estimate and see if it's feasible.

• 200 deaths in nursing homes with 15% mortality gives: 1,300 infected in nursing homes .
• 200 dead of the group protesters 65-75 with 5% mortality gives: 4000 infected from 65 to 75 years .
We must add the other protesters under 65 years. (estimate 40% of the peanut). It gives us 2,600 infected from 20 to 65 years old .

We obtained a total of 6,600 infected people, on February 29 at the demonstration and the buses.
That they infected 1,300 people in nursing homes from March 3 to 7.

It means that 1 out of every 5 infected protesters infected a relative in a nursing home in the following days. It doesn't seem impossible. Visiting his parents, he was able to infect other residents as well.
6600 infected at the demonstration, at first glance, seems very high, but ...

In the chapter D4 , we have 3 outbreaks evaluated .

From an German study on contagion / super-spreading at the Gangelt carnival an estimate emerged that 15% (1875 people) of a population of 12,500 inhabitants were infected . It is very high.

At a religious gathering of 2,500 people in Mulhouse on February 24.
In this phase, France had, like Catalonia, a few cases detected in its territory . We now know that the 'real' cases were so much more.
The French government intervened in the 8 days (unlike the Govern) and quickly calculated that at least 1,000 people infected with 2,500 participants came out. 15 days later, I check their numbers, and evaluated in 2000 the infected (out of 2,500 participants!)

In the same way, the Korean government came out with surprisingly high estimates of contagion in a religious sect, with 2,400 people infected over 9,300 members (25%!)

Well, in the end it may be that 6,600 infected people are not so much, in a demonstration of 100,000 people and 583 buses, acting as a mouthpiece for contagion.

Note: 40 buses with 1 contagious that gets on the bus in the morning can give 1600 infected on the return, only on these buses. (40 buses x 40 infected). Only in Igualada, it is said that 5 buses returned, largely infected. (see ANC-Buses)

This evaluation comes from the fact that we chose the least favorable calculation for our hypothesis, with a jump of 17 deaths in chapter C2 and C3. It is the low estimate. The average would be in a jump of 19 to 20 dead. (15% more)
As we said in the summary, we prefer to choose the least favorable calculation, so as not to be accused of 'loading' the demonstration.

So, the number of infected would be closer to 7500.

Qualitative assessment of impact

It is impossible to calculate precisely how many infected there were in Catalonia before the demonstration. Lots of young people with near-zero death rates.
But, yes, we can approximate how many people were infected in the demonstration. Due to its age, its high mortality rate and a very quantifiable and precise jump of deaths over time.

Catalonia was about 8 days late over Madrid on February 28. (Comparing the death curves 17 days later)
The confinement of March 15, would have worked well, without the demonstration.
It would not have had time to reach a strong contagion to many sectors of society, such as in Valencia, Andalusia and many other regions. There would be the contagion of Igualada although much less strong (without manifestation) and a growing contagion in Barcelona and other urban areas.

But Catalonia multiplied by many its infected in 1 day on February 29, and jumped from a low epidemic to a high and rampant epidemic.
Explained in another way, Catalonia jumped between 4 and 7 days ahead in the epidemic. (Chapter C2)

It should not be forgotten that on March 17, Barcelona went from a 'quiet' week with less than 1 death a day, to ... dozens of deaths per day. Caused by the occasional explosion of contagion in the demonstration.

But another really important aspect is that, the contagion in the demonstration, spread the epidemic where it would never have gotten so strong. In non-central areas and in a very old population.

Catalonia 'lost' in large part the main objective of confinement (protect the most fragile and least central sectors)
Because the explosion and the next broadcast was already done, in the previous 15 days.
The demonstration propelled Catalonia in a late timing, similar to a 'Madrid'.

For now we leave it here. More development in the next article and the simulation.

We wanted to show that there is a huge deflagration of the deceased (from 18 to 1013 † by Cat-sin-Igua), which could only have been created by a very punctual contagion in the demonstration on 29-F.

Who infected the protesters?

To infect many people in a demonstration, you need people in contagious phase.
Various outbreak investigations indicate that 1 person can infect dozens of people in favorable conditions and in just one day.
In 30 minutes on a bus, with air conditioning, it can infect 9 people in 30 minutes. (see next chapter)

What we can assert from these investigations is that:
When contagious people got on a bus in the morning, the bus came back largely infected.
• When contagious people did NOT get on a bus in the morning, the bus returned with infected people at the demonstration.
• And that the people infected at the demonstration did NOT infect people on the bus (minimum incubation of 2 days).

Igualada, Barcelona and Girona and Central areas, appear, according to the analysis of the curves, to be the areas where many buses were contagious.
They were the areas where some cases had been tested, and that is where the most buses came from.
We also saw that they are the areas where the curve of deceased shot out, indicating many infected. (They are needed to generate many 'dead' from the start, 17th of march).

It's visible in this: Figure 8

We think that in Lleida, Tarragona and Ebre, mainly people infected by the demonstration returned, in much smaller numbers in the case of Ebre.
Although a few deaths of the first wave appeared, it took the second wave (+3.5 days) of contagion by protesters to 'launch' the rise in Lleida (strong rise on the 21st), and the third wave (+7 days) in Tarragona (strong rise on the 24th).

Little note of optimism ...
From the analysis of tweets of the ANC of the Ebre, it seems that they were very lucky.
They arrived very late in Perpignan (coming from further away), and complained that half of their few buses never made it to the demonstration and had to turn around. The French police had closed the access due to overflow. Molta sort ...



C2. Curve modeling

A little simplified math.

We only point out in this article the mathematical reasoning and the final data, so as not to overwhelm many people. You can find all the data in the Github repository.

Almost all countries / regions have a continuous curve

3635/5000 We compare Catalonia with similar regions, to facilitate reading. We are not going to detail the curves for 50 countries.
But as we saw in the graph 3 , Cat-sin-Igua has the largest and most exceptional curve break compared to 50 countries and regions that have a continuous curve.

We calculate one or more sections of continuous curves ( - - - - - ), which is closer to each phase of each country.
They are sections of quadratic curves, exponent of the base e .
f (x) = e ^ (ax ^ 2 + bx + c)

That the data follow a continuous curve, means that there is no break in the evolution of the epidemic.
The curve may go up more quickly, or flatten, or change direction (change of section), but it remains continuous.


Catalonia has 4 phases

We note that evolution is continuous after the jump on March 17. Very similar to Madrid and Belgium. This continuity contrasts with the part before the jump.

We have found 4 phases (sections) of the progression of deceased, which 'stick' with quadratic curves.
We do not 'glue' the sections as we want, these sections are generated by algorithms that search for sections that are closest to the points.
These sections are seen in the graph in the next chapter .

• Phase 1: 'Foot' (green) . The epidemic before the explosion.

After the break is broken, a continuous series of 3 sections begins :
• Phase 2: 'Deflagration' (yellow) . When the curve progresses faster than any other country. Lasts 10 days. We notice 2 waves at a distance of 3 to 4 days.
• Phase 3: 'Normal' (white) . When the curve progresses in a 'classic' way. Very similar to Madrid or Belgium
• Phase 4: 'Confinement' (pink) . When the curve flattens out. The effect 18 days later (see period: contagion> death).

The phases in the following image of the daily increase are much better seen.
In the graph 1 , it is noted that during the 'deflagration' phase (from March 17 to 26), Catalonia was keeps on top of the other countries. It is the record of increase in 10 days.


Mathematical verification of the jump and sections

These 4 sections are not an arbitrary decision, but the result of a mathematical process.
The jump and the regularity of the sections of Catalonia are evaluated with the coefficient of determination (R square) .
This coefficient tells us how, statistically, some values ​​are close to a line or curve.

All its sections are above a determination coefficient of 0.997
Quite simply, they are 0.003 from being perfectly adjusted, they are very 'stuck'. (A coefficient of 1 means that all the points fall perfectly in the section.)
But, if we try to create a continuous section that covers the jump of March 17 in Catalonia, the coefficient of the jump drops to 0.964 , or 0.036 if it is perfectly adjusted.

Thus the jump mismatch is 12x times stronger than any other variation in the other sections. There is a very important break.

All of this is visually evident, but it is important to prove it mathematically.
And also, it helps us to quantify the jump in the next chapter, and thus evaluate the part of the epidemic that comes from the demonstration.


First impact evaluation with the sections:
On day 17, in the green section, Cat-sin-Igua would have had 8 †, and in the yellow section it has 25 †.
It gives us a jump of 17 † and that the demonstration participated in 68% of the epidemic just after the jump.

It is the tragedy of Perpignan, the explosion of contagion was devastating for Catalonia, because it happened at the beginning of the epidemic.

Mathematical evaluation of the Lombardia jump

There is another region where there was an explosion and a significant jump in contagion.
Lombardia, with the soccer match against Valencia on February 19, and the party in the streets at night of 40,000 people.
This party was declared Party Zero, a dramatic accelerator of the epidemic in Lombardy .
We see how, on March 7, 17 days after the game, the deceased curve jumps . The jump is less visible than that of Catalonia, because it comes later in the epidemic.

A little simplified math.
The 2 curves have the coefficient of determination: 0.995 and 0.997. they are very 'stuck' to the points
But as with the Cat-without-Igua jump, if we try to create a curve that covers the jump, we get a coefficient of 0.991. The jump is only 2x times stronger than other variations, less than Cat-sin-Igua.

This day Lombardia goes from 154 to 267 dead. A significant increase (73.4%) that is reflected in the green peak to the right of the first chart . It is notable that this peak is above the other countries, and it is very high, being so advanced in its epidemic (12 days after 6 †).

Simplified impact assessment with the sections:
On the day of the jump, in the green section, Lombardia would have had 196 †, and in the yellow section it has 283 †.
It gives us that the party participated in 30% of the subsequent epidemic. Yes, although Lombardia has a jump between sections of 87 † (283-196) and Cat-sin-Igua of only 17 † (25-8), the participation in its subsequent epidemic is much lower.

Where is the 8-M in Madrid?

The Madrid curve is very regular, from start to finish . And very vertical in this regularity, one of the most vertical in the world, which explains the impressive number of deaths.

We have searched for a jump around March 25 (17 days after 8-M), and have not found it.
We attribute it to two factors:
Even more than the Lombardy party, the 8-M deaths arrives very late in the Madrid epidemic, already with 1825 †. So its effect is relatively low on the already important daily death number (250).
Another factor to consider is the age of the protesters. On 8-M, the population is much younger than in Perpignan. The vast majority of those who are going to catch it (there was a significant contagion, for sure) are not going to die.
And the repercussions on future infections and deaths are going to dilute a lot over time.

Impact evaluation :
• 70% of the subsequent epidemic in Perpignan, because it happened very early in its epidemic.
• 30% of the subsequent epidemic in Lombardia, because it happened in the middle of its epidemic (+12 days than Perpignan).
The 8-M arrives in the Madrid epidemic much later than Lombardia (+9 days).
We can consider that the impact of 8-M may represent less than 10% of the subsequent epidemic. And then it is very difficult to detect.



C3. Estimation of deaths related with the 29-F

This entire chapter is under hypothesis and uses prospective interpolations and calculations.
The estimates 'results' have to be taken very, very carefully.
Also, for a material that is, deaths of thousands of people.


But in this phase we do not have the objective of calculating an accurate evaluation, but rather to 'give' our simulation a starting point, which the algorithms will gradually adjust .
It is a round trip process between:
1st estimate> 1st simulation> 2nd estimate adjustment> 2nd simulation, etc ...
Any contribution from specialists in epidemiology, analysis, statistics is welcome.

In order to adjust the parameters of this 1st simulation, we need, in a first estimation, to separate:
• the part of the death curve that comes from the impact of the 29-F.
• from the part of the death curve that comes from the development of a 'normal' epidemic in Catalonia, which we are going to estimate.

We think that (hypothesis) the Government of the Generalitat could have intervened since March 9, to reduce the effects after the contagion in the demonstration.
So we decided to also calculate a second estimate of the curve in case of:
Control / intervention of Gencat.


We work with death rates in Cat-sin-Igua hospitals .
Compared to the size of the deflagration, very quickly, the previous Igualada outbreak is of little relevance (57 † vs 1013 †).

As we calculated in the previous chapter, we have a relatively clear starting point on the day of the jump , on March 17.
8 † of the normal epidemic (Foot section)
17 † of the manifestation epidemic (Deflagration section)
• total of 25 †

Two cases (hypothesis)

This first graph is on a logarithmic scale to see the curves well.
The actual curve is shown in orange, and the cases / estimates in green and red.
And also the real curves of deaths of Andalusia and the Basque Country to have a reference (aligned with the Foot of Cat-without-Igua).
On April 2 (day 33 on the chart), Gencat stops giving numbers for Igualada and then Cat-sin-Igua.
This very small jump in the orange curve from Cat-sin-Igua to Catalonia are those of 107 † from Igualada that are added. at this point, it represents less than 5% of deaths.


Case.1 No-Mani . (in green), No Manifestation

The demonstration did not take place, and Catalonia develops a 'normal' epidemic .
• There is neither jump nor deflagration (record acceleration).
• The virus does not enter the nursing homes so soon, at the beginning of March.

Many studies point out that each country / region has its own curve behavior, depending on population density factors, social behavior (Latin society), tests and preventive individual isolation, etc ...

We have 'interpolated' the trajectory of Cat-sin-Igua as follows:
• extended the quadratic function of the 'Pie' section to about 27 †. It falls on March 21.
• 'added' the behavior (its speed) of the actual curve from this day. Thus we add the end of the deflagration, the 'normal' section and we also keep the inflection of the confinement on the 33rd day (April 2).
Mathematically, this 'transfer' from the real curve to the No-Mani curve, 'works' because we are working on a logarithmic scale, which preserves the proportions .

Surprisingly, this first 'simplified' interpolation gives us something similar to the shape of the curves in Andalusia and the Basque Country.
With almost twice the number of deaths (unfavorable to Catalonia), this seems to us to be a sufficient estimate as a base hypothesis for our simulation.

Case.2 Control . (in red) Gencat intervention

(hypothesis) The Generalitat of Catalonia and its health organizations realize, between March 9 and 16, that there was a problem with a significant contagion in the demonstration and they intervene (see Chapter D2).

We are working on this scenario.
At the moment the estimate is 'just' somewhere in between Non-Mani and the real curve.
We think it would have to be a collective job.

Estimates


This second graph, with exactly the same data on a normal scale, shows the differences in final deaths (in hospital).
We have extended the real curve for Catalonia until May 1 (day 62), so the 2 case / estimate curves.



Case.1 Non-Mani . Conclusion

The objective is not to calculate the number of precise deaths without manifestation, but to have a 'base' curve for the simulation. Thus, we can then simulate the shape of the curve for Perpignan-related deaths.

Although 'very indicative', day 62 (May 1) gives us that with a 'normal' epidemic, Catalonia would have 2,087 † instead of 5,987 †.
It would have reached 35% of current deaths, plus or minus a third. (2087/5987)
And then, the demonstration participated in 65% of the current deaths.
To have an important margin, we write in the summary that represents more than 50%.
In any case, it remains an evaluation. Only a much more complete study could claim to give a more precise number. It is not our purpose.
We will use it as 'ground' for the simulation.

Although the evaluation of this scenario may seem very low, we consider that it is not so low.
As of May 1, Catalonia has 5,987 deaths (hospitals), with a death rate per million of: 780
In this Non-Mani estimate, Catalonia would have 2,087 deaths, with a death rate per million of: 272

In Graph 9 ( EU • Regions • DeathsPerMilion ) we see that the vast majority of regions in Europe have a rate below 195 .
As you can see, without demonstration and a rate of 272, Catalonia would still be in the middle of the green group , the group of highly touched regions in Europe.
With a third of its current deaths, Catalonia would have a higher death rate than Valencia, equivalent to Belgium and above the vast majority of regions in Europe.

You have to understand that Catalonia has not done it badly , it has done it EXTREMELY BADLY.

... and that comes from the acceleration record from March 17 to 27, that launched Catalonia in an exceptional trajectory.
... and that this record has a cause, the massive contagion of the 29-F Demo.




C4. Summary simulation of contagion

We are developing a complete simulation of contagion in Catalonia, before, during and after the event. We will present it in a new article.
We will write down in this chapter the main results of this simulation.

Any contribution from specialists in epidemiology, analysis, statistics is welcome.
You can use our database. Do not hesitate to contact us.

Meanwhile, you can find our first indicative calculations in the chapter:
C1. Mass contagion on 29-F




Health and political action



D1. Easy confirmation with an investigation of 104 deaths

There is a way to quickly and effectively assess the impact of the demonstration on the epidemic in Catalonia.

104 official deaths were registered from March 17th to 20th inclusive. (from 18 to 122 †).

Simply, the investigation have to ask the relatives of the deceased if:
• the deceased went to the demonstration, in case of not living in a nursing home.
• a relative of the deceased went to the demonstration, in case he lived in a nursing home.

Simply ask this, nothing more. It can be totally anonymous through Catsalut.

With only 12 'confirmations' among these 104 deceased, the effect of the demonstration would be demonstrated.
As more confirmations, more important the effect of the demonstration.

Explanation

From Covid-19, die almost exclusively people over 65 years.
In Catalonia there are about 1,360,000 people over 65, 18.9% of the population. Idescat_Grups d'edat

There were a maximum of 75,000 people over 65 in the demonstration. (half, in a crowd of 150,000 people). And surely much less.

In other words, the demonstration had less than 5.5% of people of this age in Catalonia.
With relatives of protesters in nursing homes we must be below 8%.
So protesters or their relatives should represent 8 deceased out of 104. We give a margin of error up to 12.
We think that there would be much more than 12 deaths DIRECTLY related to 29-F.
And it would allow to measure more precisely the impact of contagion this day.

Deaths after March 20 may come from a person infected by a protester, and then the relationship with the demonstration will be much more difficult to investigate.

A small, simple and non-invasive investigation allows to check the contagion in the manifestation. There is no argument against.

We appeal to people who have information about an acquaintance who died between March 17 and 20, and who had a relationship with the demonstration.



D2. Chronology of the possible non-action of the Catalan Government

We have decided to record the dates in which, in our opinion, the government could have detected a massive contagion on 29-F and its communication and reaction those days.
We are aware that all governments experienced an exceptional situation that makes it very difficult at the time to 'read' the data.
But, the 29-f demonstration has a very close relation with the catalan government, like the 8-M demo for the government of Spain.
Readers can make their own opinion if the government eventually gave priority to "not questioning" the demonstration, and that it supplanted the priority of health actions that involved unveiling a massive contagion on 29-F demonstration.

Before the demonstration


February 15-28

It is not as if the risk of major contagion in the 29-F demonstration is a posterior conclusion.
15 days before, the cancellation of the MWC had been decided , after a week of tense discussions.
It is evident that the Government and Health advisers spoke about the risk of contagion in a similar event of 100,000 people.

The important thing is that the Govern and the Health organisations knew that there was a risk.
A possible contagion could be watched, at least they had it in mind.


Note: We consider errors authorizing large events in these early epidemic periods, and not only in retrospect but before they occurred.
But the Government of Catalonia did like the others.
It was even worse from the Government of Spain to authorize the 8-M demonstration at an even more advanced stage of the epidemic.
And we cannot understand at all the decision of the French government to hold the first round of the municipal elections on ... March 15.

We are not questioning the responsibility of the Generalitat to authorize 29-F.
But there will always be the doubt that the political agenda supplanted the precautionary principle


February 25

The French government authorized the demonstration, to support Pujol, mayor of Perpignan (Macron's party), who was in difficulty facing the Front National candidate for the next municipal elections.
( Confidential - Aló Macron? )
The French government spoke of a contagion risk , and announced to its population that it would take drastic measures.

The French government imposed an almost quarantine of the demonstration. With very strong control by the police, and a strict limitation on the parking of the Parc des Expositions on the outskirts of Perpignan.

February 29

Apparently, no contagion control measures were taken by the organisers.
It was such a success that the event was delayed, people spent more time than expected on the buses, at the demonstration.
The entire Junts per Cat (political party) part of the Govern participated, and many personalities of the independence movement.
We think that the presence of many like-minded journalists, on a personal basis, 'prevents' them from evoking a massive contagion this day.

After the demonstration


March 9th

As soon as March 9 , a major UCI jump occurs throughout Catalonia (from 5 to 18). It is a clear indication that a significant contagion had occurred on 29-F, since it corresponds to the period between contagion and ICU of The Lancet study. .
This day was the only day that the Govern did not communicate the UCI number.
Gencat_Comunicat_09- March

March 11

The Government notes a rise in cases in Igualada and prepares a plan for the confinement of Igualada (Conca d'Odena).
There was surely an express investigation to determine what was happening in Igualada.

March 12th to 16th

There is a very rapid rise in ICUs throughout Catalonia (from 23 to 66 ICUs), when in Igualada ICUs rise only from 3 to 8 ICUs.
For us, this rise is the wave of deaths that arrives throughout Catalonia, largely the result of a massive contagion on 29-F.
We think that the Govern could see this rapid increase these days.
That would explain the total change of comunication of the Government between day 11th and day 13th.

March 16th

Conclusion of these 7 days (from 9 to 16), we think that the contagion of 29-F could have been detected during this period.

March 17th

This day a synchronized explosion of the first deaths occurs in all of Catalonia (jump from 18 to 41 †, with 23 new deaths ). With the world's largest increase in all data recorded so far.
The death jump is so strong, that the ICU go down.
The Govern publishes an 11-line statement , without commenting the sharp rise and deaths in each health region. It only communicates about the Igualada outbreak (2 new deaths).
Gencat_Comunicat_17-March

The Govern stopped giving details on the place of the deaths for 1 month . As of today, it does not give the data for this period.

March 17th to 27th

In the following 10 days, Catalonia experienced one of the largest worldwide acceleration of deceased ever recorded
The Govern never reported or commented on an exceptional acceleration situation.
They did not give any number of deaths in Girona, Osona, Bagés, Barcelona, ​​apart from an oral response to the National newspaper on March 24 and an attached communication on 26.


March 20th

3 days after the explosion, the Government announces that it has investigated the Igualada outbreak.
He claims the contagion comes from an 80-person dinner on February 28, when the next day 600 protesters of the 29-F demo from Igualada (12 buses) spent hours together, in what studies describe as a mouthpiece of contagion.

April 16

1 month later, the govern starts again to give the numbers of deaths by health regions.



D3. Action of other governments against outbreaks

It should be noted that some governments, faced with major outbreaks, recognized and acted quickly to measure and control these outbreaks.
We are talking about governments that considered "major" outbreaks, meetings of 2,500 people.

France . Mulhouse

From February 17 to 24, a major outbreak occurred among the 2,500 participants of a religious congregation in Mulhouse, Grand-Est. A third of the attendees came from other parts of France and spread the virus.

8 days later, the government acted with tests and monitoring .
15 days later, this outbreak was estimated to have spread to more than 1,000 people .
1 month later, the French president declared a state of war against the coronavirus from Mulhouse, considered the zero zone of France. Today, Grand-Est is the French region with the highest death rate. (67% of that of Catalonia)
Press: France Bleu
Press: the bombe atomique du rassemblement évangélique de Mulhouse

Germany . Gangelt

An estimate came out of a German study of contagion / super-spreading at the Gangelt carnival that 15% of a population of 12,500 inhabitants were infected.
CatCovid_Data / Gangelt

Korea

In mid-February, a major outbreak occurred among members of a religious sect in Korea.
Soon, the government was acting with tests and follow-ups.

9300 members were isolated and tested .
On March 2, it is estimated that 60% of the 4,000 cases of coronavirus in the country are members of the sect .
In Spanish: BBC - Sect of South Korea

Lombardy

On February 19, the soccer game with 40,000 Bergamo fans , and the party in the streets at night.
This party was declared as the Zero Party, a dramatic accelerator of the epidemic in Lombardy.

The government did not act, the spread of COVID-19 was already uncontrollable. Unlike Catalonia, Lombardia already had 267 deaths and was very advanced.
But government health officials assume the dramatic significance of the blast.
Press: Vanguardia - El Atalanta - Valencia, an accelerator




Appendices



E1. People contacted in Catalonia

Persons in responsibility or with privileged access to data
in the period from March 9 to 27


Given the difficulty in confirming that those responsible for the Govern and the Salut organizations receive our analysis.
We ask some public questions, and we'll update this section as they respond.

Do you think that the data by health regions of this period should be publicly published?
Having access to the data, have you detected a possible contagion on 29-F? In case you did, when?
Do you think that a possible contagion on 29-F should be investigated?
Are you ready to allow independent experts on this investigation?

Public responses

It seems that the e-Notícies article and our analysis generate some first public questions.
Although, as was foreseeable, the Government focuses its response on saying that it was not possible to 'foresee' an eventual contagion in the manifestation of 29-F
Well obviously, it is not what we consider important. We ask for an official investigation

e-Notícies . May 9th . Budó se enfada por una pregunta sobre Perpiñán
Ara . May 10th . Entrevista al president de la Generalitat

People without responsibilities in the period from March 9 to 27

In the end, this analysis seems to have had sufficient diffusion. So we consider those 'actors' informed. They can contact us if they wish.

Newspapers

Politicians

Civil associations

Health

Scientists and Academics




E2. International Contacts

This analysis is translated into English, and soon Catalan, French and German.

We are contacting international epidemiologists, to obtain their assessment, given that no catalan epidemiologist contacted us.

We think that the contagion of 29-F demonstration is interesting to study, because it is the 'almost' perfect outbreak :
• happened at the beginning of the epidemic. Its effect / jump is very visible and calculable.
• massive, from 75,000 to 150,000 people.
• lasted almost 6 hours during the gathering.
• It extended throughout the day (departure of buses 7h, return 22h).
• fairly closed, with little interaction with the outside.
• with a very particular sociology (50% over 60 years).
• with a strong impact in non-central areas that had very few cases of coronavirus before the event.
• with a precise database by Health Regions.



E3. Contact and references

Any partial or total mention of elements or data on this website must be agreed with CatCovidTransparencia and refer to our website as:

CatCovidTransparencia
with link to:
https://catcovidtransparencia.blogspot.com/

You can contact us:
CatCovidTransparencia@gmail.com

Sign the petition at:
Change.org/p/cataluña-covid-transparencia-de-la-gencat




E4. Notes and context

It is not an official report, it is an analysis of data


We are not an official investigation and we never pretended to be.
We invite you to participate and we will point out any contradiction (with arguments, obviously) in this analysis.

We consider ourselves as a form of alert launcher, on data that seems exceptional to us.
Our scientific training has allowed us to carry out a first analysis of these data.
We present an analysis based strictly on official data (Gencat, Governments, Universities), for the two exceptional data, the jump of March 17 and the rise of the following 10 days.
The second part, the synchronization of the whole of Catalonia, comes from the data that we have collected strictly from articles in the local press.
The third part, bus-like distribution, comes from the data we have collected from messages from the local ANC.

About us


We are not a hidden office or a political group.
We are simply European citizens, settled for many years in Catalonia.
Before everything, we are democrats, with a strong exigency for transparency.
We think that the world is complex, but that the population can understand the difficulties for an administration to manage a crisis like this one.
As residents, we have every right to ask for transparency. We live in confinement with our families, too.

Although we launched an alert, we do not want to and do not have to be on the 'front' line.
We have always pointed out that it is the responsibility of Catalan society to demand this transparency from 'their' authorities, through its 'actors' or its inhabitants.

We have decided not to make linguistic correction of our analysis. Thus it is clear that we are neither Spanish nor Catalan. But we love both sides of our everyday environment.
We will soon present a Catalan version, thanks to some readers.

How we got to this analysis


Context:
Because of professional background and scientific interest, we participate in discussion groups in Europe, with epidemiologists, data analysts, artificial intelligence, economists, etc ...
Since the end of January, we have been very up to date with studies on the evolution of the epidemic.
From the evaluations of these expert discussions, at the beginning of March, it seemed clear to us that Catalonia did not have a few cases of contagion, but hundreds. And that a phase similar to Italy was coming.

Until the manifesto: (March 31)
With the announcement of the confinement, we decided to use epidemic simulation tools to see how the situation in Catalonia could evolve. We started to put in tables the Gencat communication datas to process them.
On March 24, we concluded, after comparing with Madrid and Lombardia that had been 1 to 2 weeks ahead, that the jump of 17th of march and the acceleration that Catalonia was experiencing was clearly 'abnormal'.
We have tried to 'square' this jump with simulation tools for 3 days and it was impossible to simulate this jump. These tools have the possibility of applying a specific event (confinement type) on an epidemic curve, which has the effect of greatly lowering the reproduction coefficient (Ro).
We realized that the jump and acceleration on March 17 was like an anti-confinement event, a reverse confinement. A punctual explosion.
we had in mind, some scientific studies on Covid-19 that pointed to a period of 2 to 3 weeks between contagion and death.
Click. The specific jump of the deceased came from a significant contagion jump, the demonstration in Perpignan.
We decided to verify this period and indeed, although the average period to die is 23 days, deaths suddenly start to rise very strongly at 17 days. (see our article)
We did other checks, compared the uniqueness of the March 17 jump to the world's most widely used Covid-19 (JHU) database, and decided it was enough to publish the manifesto. We also speculated about an early possible contagion of the nursing homes due to the advanced age of the protesters.

At this time, we attribute the fact that the Govern did not report the explosion and stopped giving the location of the deaths to a desire not to 'scare' the population.
But as a connoisseur of the political situation in Catalonia, we were in doubt that the Government did not want to highlight an explosion that could be related to 'its' demonstration.

April:
Our first idea was that if there was massive contagion in the demonstration, then, the deaths after March 17 had to appear throughout Catalonia that participated in the demonstration, and not only in Barcelona, ​​as a great urban center, where normally the contagion rise.
But Gencat this day 17th had stopped giving the location of the deceased.

So we started looking at the articles in the local newspapers for March 17th.
Incredible, the first deaths in Girona, Osona, Lleida and Tarragona were from the 17th of march.
As we compiled the articles, which gave an accurate count of each hospital, these synchronized and very fast curves for each health regions came out.
Then a lot of work, and a willingness to rely on reliable data.
The rest is in the analysis.

Impact evaluation of 29-F


We always make it clear that they are evaluations.
Only an official investigation could determine more precisely what is the volume of contagion that occurred on 29-F, and its 'part' in the Catalan epidemic.
But, as it seems that at the moment, nobody is willing to do it, we have decided to make a first estimate.
We think that with the data and the fact that the contagion occurred very early in the epidemic, an official investigation could obtain relatively accurate results.
We do not think that this is the case for the 8-M and the party in Lombardia that occurred much later in the epidemic and that it had a much more diffuse effects.
We think that although the number of infected was surely much more important on 8-M or in the Milan futbol game, its proportion and then impact is surely much lower than 29-F.
We explain it in chapter C2.

Authorize 29-F


We consider errors authorizing large events in these early epidemic periods, and not just in retrospect but before they occurred.
But the Government of Catalonia did like the others.
It was worse for the Government of Spain to authorize 8-M at an even more advanced stage of the epidemic.
And we cannot understand at all the decision of the French government to hold the first round of the municipal elections on ... March 15.
We are not questioning the responsibility of the Generalitat to authorize 29-F.
But there will always be the doubt that the political agenda supplanted the precautionary principle.

Possible non-action of the Govern


All governments experienced an exceptional situation that makes it very difficult at the moment to 'read' the data.
In ALL of this analysis, we make it clear that this is a hypothesis.

But, the 29-f has a close link with the Govern of Catalonia, like 8-M for the government of Spain.
Readers can make their own opinion if the Government eventually gave priority to 'not questioning' the demonstration, and that it supplanted the priority of some sanitary actions that involved revealing a contagion on 29-F demonstration.