Did the use of masks lead to a significant reduction of cases in Germany?
Summary: Masks were not significantly effective at the population level.
This is a translation of the German thread written by Sten Rüdiger (Builder of the Covid contact analysis used by Robert-Koch-Institut).
Did the use of masks lead to a significant reduction in #SARSCoV2 case numbers in Germany? The calculation of the transmissibility T, which should be reduced by masks, from the reproduction numbers of the last three years gives a clear answer. We were able to calculate the value of T relative to the wild-type from the R-values considering the contact counts.
As well known, R results from the product of the number of contacts and transmissibility. The calculated curve of T clearly shows a first constant phase of the epidemic in 2020 (wild type) and from 2021 three variant-driven and immunity-dampened waves. During 2020 and since spring 2022, however, the T values have remained more or less constant.
After the introduction of the masks (from April 2020), contrary to expectations, no decrease in the probability of transmission can be seen in the T-curve. Likewise, the removal of the mask requirement does not appear to be associated with an increase in T.
This leads to the conclusion that masks were not significantly effective at the population level. Therefore, I see no reason to reintroduce an expanded mask requirement in the fall.
Aug 17 2022
How will contacts develop in the next months and how heavy will the #SARSCoV2 wave be this fall?
We examined our mobile data from company NET CHECK, and in particular, compared with historical data from the pre-Corona period. The so called contact index, developped by us together with @ChariteBerlin and @HPI_DE, was further modified and applied to earlier historical data scaled differently for technical reasons -> CX2.0. Values of 200-500 were achieved in pre-Corona fall 2019.
Since values below 300 dominated in spring 2020, before Corona, we target values of around 250 for fall 2022. Deviations downwards (due to a still cautious part of the population) as well as upwards could be possible.
With this we investigate the evoution of the reproductive number when returning to normal contact behavior in autumn. This investigation becomes very clear with a simple splitting model. R consists of 2 factors:
1. contact number
2. transmission probability T (per contact)
Using CX and the eff. R from the @rki_en, we can calculate T. The result, smoothed over 2 months, is a progression of T in which we can clearly see a first constant phase of the pandemic (wild type) and the following variant waves (alpha, delta, BA1/2).
Multiplication of CX and T then allows to determine an adjusted R-projection corrected for reporting deviations (orange). It contains waves resulting from new variants as well as waves triggered by higher contact numbers (incl. summer 2022)
In the last weeks, the development of transmissibility has been relatively constant. Assuming that there will be no new variants with significantly higher T in the coming months, a forecast for the fall can be derived from T and CX. For this, we assume three scenarios for CX:
1) continued constant contact behavior as in summer (unlikely with falling temperatures).
2) Our expected increase in CX to more than 200 by November.
3) An extreme scenario of 350 by November.
The dashed curves correspond to the three scenarios. For Scen 1, the R value remains consistently below the critical value of 1 (no fall wave). For Scen 2, the result is an autumn wave that will produce high case counts in advance starting in November. There are risks of a higher wave in autumn if there are even higher numbers of contacts (scene 3) and new variants.
Thanks!