Since I shared an article yesterday about the ‘miracle cure’ for Covid-19 that was published by Didier Raoult my eye fell on another story today. A series of tweets this time, by Katja Baerenfaller (@jaggakatja) from the University of Zurich, about the latest corona virus publication from Raoult’s group. Below I copied the texts.
Here is the article she is talking about.
The claim of this study is that the mortality rate for common corona viruses is not significantly different to that for SARS-CoV-2.
The data are ‘from 1 January 2013 to 31 December 2019, 21662 samples were tested by the IHU Méditerranée Infection diagnostic laboratory. Among these, 770 samples were positive for corona virus, with eight deaths (mortality rate 1%).’
Viruses from the corona virus family are known for a while and mostly cause respiratory infections in the common cold. During the testing period, none of the dangerous virus strains SARS-CoV-2, SARS-CoV and MERS-CoV) were circulating in that region. So what was done here was to test patients that were in hospital treatment for the presence of viruses that cause the common cold.
For those that were tested positive, it was assessed how many had died. From this number of hospitalized patients who have died and were positive for corona virus infection, the mortality rate was assessed. This assumes that these patients were dying from the corona virus infection and no information is provided why these patients were in hospital treatment!
Further on: ‘Systematic testing … for SARS-CoV-2 was performed from 1 January 2020 to 2 March 2020. In total, 7059 samples from patients presenting with infectious symptoms were tested by the IHU Méditerranée Infection diagnostic laboratory. Among them, 543 samples were positive for coronaviruses, with two deaths (mortality rate 0.36%): 277 samples were HKU1, 146 samples were NL63, 77 samples were OC43 and 43 samples were 229E. No cases of SARS-CoV-2 were identified among these samples.’
So also for the 2020 data none of the dangerous corona virus strains was detected, but the deaths are attributed to corona virus infection assuming causality. There is no information available on the number of deaths in the patients that were tested negative for coronavirus (if any of these has died, what would have been the cause following this logic??).
In summary, without the number of deaths in corona virus-negative hospital patients, no background distribution and hence no over-representation can be calculated, and assuming causality of infection with corona virus strains causing the common cold for patients dying in hospitals is not legitimate.
Note: Author Philippe Colson is on the journal’s editorial board.
By the way: Max de Foret wrote a more elaborate story about this article in French.