Discussion
There is increasing empirical evidence that our current view on
contagiousness and hazard potential of COVID 19 is heavily biased by
test strategy and selective sampling. Just to name a few:
- Based on retrospective blood analyses the date of infection for
patient zero in France was estimated to mid-December. With a serial
interval of four days as suggested by the RKI virus profile and a
conservative estimate for the basic reproduction factor of only 2,
this would mean that by March 12, when the French government started
to get serious on lockdown measures, the country would already have
had some eight million inhabitants infected. Likewise, with the virus
already present in France and Italy in November / December, it is a
very likely scenario that it also had reached Germany much earlier
than suspected, which has not been systematically investigated so far.
- Only some countries, like Germany or Italy, show a rapid increase of
infections and only during the first two to three weeks. Countries,
that are internationally criticized for their management of the
pandemic like Brazil or the Sweden record high infection rates, but
the increase is far from the projected exponential catastrophe
assuming a basic reproduction rate of 2 to 3.
- Despite lack of resources and limited possibilities of infection
control in many countries we do not see many reports of an actual
crisis or collapse of national health systems so far. There have been
local or regional shortages, of course. On the other hand, the region
in Northern Italy as well as the city of New York, which suffered
considerably from a lack of capacities, had difficult situations with
respiratory diseases of other origin, which indicates that there might
be some local vulnerabilities in those areas .
The “true” basic reproduction rate of SARS-COV-2 probably never was
anywhere close to the dimension of 2.5 to 3, as it is still reported in
official information . However, model simulation studies have been
typically fitted exactly to these biased numbers. For SIR (Susceptible
– Infected – Recovered) models, due to lack of representative data we
do not have any valid basis for estimating the number of infected nor
the number of recovered patients, nor the susceptible population. There
is emerging evidence on cross-immunity with harmless Corona viruses that
may already be in the region of 30 to 40 percent or even 80 percent
according to some very recent results on T-cell activity . This would
for example explain findings like couples sharing the same cabin on a
cruise ship or living with a symptomatic COVID patient in the same
household but not getting infected. It would also fit to the fact that
we see the infection rate in Sweden subside, although they should not
have had much more than 15 percent of their population contracting the
virus so far, which would not be sufficient for herd immunity.
In summary, public surveillance data as published by the RKI do not
confirm a strong impact of public health measures on SARS COV 2 spread
in the German population. Specifically, public mobility as an indicator
of compliance with social distancing rules seems to be influenced by
published infection rates rather than being a moderating factor for
infection. Data from the initial phase of the pandemic could be heavily
biased and thus provide a severely distorted picture of contagiousness
and risk of severe course of disease. Model simulations building on
these data will likewise be biased to exaggerated outcomes. There is an
ongoing need for solid and large-scale epidemiological studies that not
only address currently active infections but also consider resent
findings on t-cell based immunity and cross-immunity. The demand for
more empirical data has been raised early e.g. by the German society for
evidence-based medicine, and it is still valid. Rather than focusing on
infection containment in the broad population it might be more
reasonable to focus on correctly identifying and protecting the
high-risk population, which is probably much smaller than currently
perceived.
Figures