The V201 model

Here is the latest Admissions and Deaths forecast summary from the spreadsheet: 

Forecast using spreadsheet v 201 – dated 15/4/20

Country

Admissions

Deaths

Peak date

Clear date

Comment

Wuhan

82,870

3,719

22/2

13/2

Includes China numbers. probably.

UK

84,279

32,280

13/4

12/4

 

UK London

18,472

11,277

7/4

3/4

 

Italy

156,363

55,072

25/3

26/3

The continuing lock-down achieves only delay

Spain

166,019

41,976

11/4

6/4

 

Brazil

22,169

7,278

17/4

10/4

Lock down takes 10 days to work here

Spain Madrid

40,469

13,606

7/4

5/4

 

France

95,403

84,966

8/4

1/4

Lock-down not working (??)

USA nation

557,571

119,814

16/4

12/4

Very low admissions – unclear whether lock-down is operational

USA New York

107,260

9,910

12/4

8/4

 

Germany

117,658

10,965

17/4

5/4

 

Sweden

10,483

1,950

11/4

4/4

Very low illness rates

India Nation

154,658,118

7,783,972

4/8

31/7

Assuming no effective lock-down in place

Australia

6,322

134

8/4

2/4

Very low illness rates



 
 
 

This model has numerous improvements over previous:
  • A non-contagious initial period
  • A contagious period that more reflects other published data (8 to 14 days)
  • Multiple virus outcome groups: this is because international experience is widely different.
  • Improved process flow (more work needed on this)
  • Added switch to off/on future lock-down. Base case is to retain lock-down indefinitely.



 
 
 

These charts show that the virus can be predicted - BUT the variables are many , and like any statistics, they can be manipulated (fairly easily).

The more experience to test the model against, the more accurate in its main assumptions it is likely to become. 

In this case, there are multiple ways that the virus could be achieving the current outcomes - the main ones being:
  • very fast transmission between people
  • the interactions between people being variable
  • slower transmission when people are unaware of their contagious state
  • the environment (e.g. weather, sunshine, etc)
  • general health of the infected (eg, age, underlying symptoms)




 
 
 
 
 
 
Useful Points to Note V201

The base data is known to be somewhat false. 
I have tried to make the base spreadsheet follow Chinese, Italian, Spanish and UK data.  There are a lot of variables to which outcomes are sensitive. Key ones are: 
  • % of population at risk (age dependant - for <60 10%, over 80 40%)
  • No of people individuals are in contact with per day (5)
  • Ease of contagion assumed (50%) 
  • No of days contagious unknowingly (10)
  • Time spent ill before either Hospital or recovery (2)
  • % dying at home (less than 1% normally)
  • % to hospital (age dependant, under 60 10%, over 80 at risk 90%)
  • % into ventilator (under 60 5%, at risk over 80 80%)
  • % dying after Ventilate (<60 less than 15%, over 80 at risk 60%)
These are not accurate - but they are indicative. I have used estimates from available specialists (noted in the file).