France : COVID19 – Cases in Hospitals 31/03/2020

https://wp.me/p7ciWq-kx

I – Increase in hospital bed requirements linked to COVID18, from 03/25 to 03/31

This analysis confirms the rough evaluation from the cumulative data (see part II of this post). Density appears not to be significant at this momentum. It has nevertheless been kept in the model.

This is an analysis based on public data, and subject to revisions or errors including the processing.

Data sources: Géodes – données en Santé Publique, INSEE.

Analysis

Multiple Regression – COVID19 in hospitals 200331-COVID19 In hospitals 200325
Dependent variable: COVID19 in hospitals 200331-COVID19 In hospitals 200325
Independent variables:
Density
Inhabitants per km²

.                                                                         Standard            T
Parameter                                 Estimate           Error       Statistic    P-Value
CONSTANT                                 73.014            10.4148       7.01062   0.0000
Density                                    -0.168036          0.425494   -0.394921   0.6938
Inhabitants per km²              0.226678          0.425421      0.532832 0.5954

Analysis of Variance
Source                                  Sum of Squares      Df      Mean Square    F-Ratio     P-Value
Model                                     1.96388E6               2            981941.             95.05     0.0000
Residual                                 1.01239E6            98             10330.5
Total (Corr.)                           2.97627E6          100

R-squared = 65.9846 percent
R-squared (adjusted for d.f.) = 65.2904 percent
Standard Error of Est. = 101.639
Mean absolute error = 70.5319
Durbin-Watson statistic = 1.69812 (P=0.0650)
Lag 1 residual autocorrelation = 0.143963

Chart : increase in bed requirements from March 25 to March 31.

Dep200331increasepredicted

II – Analysis on cumulative data to 03/31

Both population and density of population are significant factors in explaining the COVID-29 outbreak in France. With an adjusted R squared of 76.7% , to be compared to 74.4% two days ago the regression continues to gain in signification (see also F and T statistics).  Indeed, the role of the Est Région continues to be reduced in the model, meaning that the entire French territory is following a trajectory closer to that of the Est.

The slopes of the least squares adjusted lines continue rising, especially concerning population density, from 0.076 on the 29th of March (cumulative data) to 0.084 on the 31st. In other words the concentration of population is the major influence behind the expected increases. From this it is easy to understand that the main needs in hospitalisations will be concentrated in the largest cities : Paris, Marseilles, Lyon, Nice…  The needs for Paris (and Ile-de-France) for example will exceed 7500 beds with ventilators in the coming days.

This is an analysis based on public data, and subject to revisions or errors including the processing.

Data sources: Géodes – données en Santé Publique, INSEE.

Analysis

Multiple Regression – COVID19 in hospitals 200331
Dependent variable: COVID19 in hospitals 200331
Independent variables:
Inhabitants per km²
Population

.                                                                        Standard                  T
Parameter                             Estimate             Error                Statistic          P-Value
CONSTANT                            -39.4282            30.1763              -1.30659           0.1944
Inhabitants per km²             0.0898564        0.00839416      10.7046             0.0000
Population                              0.000325191    0.0000396694   8.19752           0.0000

Analysis of Variance
Source                                 Sum of Squares         Df        Mean Square          F-Ratio    P-Value
Model                                   1.05145E7                  2             5.25727E6              165.32     0.0000
Residual                              3.11647E6                 98            31800.7
Total (Corr.) 1.3631E7                                       100

R-squared = 77.1369 percent
R-squared (adjusted for d.f.) = 76.6703 percent
Standard Error of Est. = 178.328
Mean absolute error = 112.192
Durbin-Watson statistic = 1.36107 (P=0.0005)
Lag 1 residual autocorrelation = 0.315876

Charts

Dep200331predicted

Dep200331densityDep200331populationDep200331residuals

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s