It is "the new study that nobody talks about", according to followers of conspiracy theory and "anti-containment".
An article widely relayed in recent weeks, published in the
European Journal of Clinical
Investigation
by several researchers, would demonstrate the ineffectiveness of containment measures to fight against the Covid-19 epidemic.
Supported by Didier Raoult, this thesis is however dismantled by several scientists.
What does this study say?
She is interested in “non-pharmaceutical interventions” (INP), implemented in several countries during the health crisis.
Behind this somewhat vague name hide more concretely the containment measures: obligation to stay at home and closure of businesses.
The authors of the study looked in 10 countries, the effect of these more or less restrictive INP.
If “advantages” cannot be “excluded”, according to the authors, they do not find “significant impacts on the case curve” in the countries with the heaviest constraints.
They claim that "similar reductions in the growth of cases may be achievable with less restrictive interventions."
Who wrote this post?
This study is the work of four authors: Eran Bendavid, Jay Bhattacharya, Christopher Oh and John PA Ioannidis, all researchers at Stanford University in the United States.
John Ioannidis, better known than his peers, is an American epidemiologist, "famous for his rigorous evaluations - and his frequent debunking - of treatments for diseases," noted the Washington Post last December.
Before being very widely criticized for his words and some of his work on Covid-19.
In particular, he estimated, based on death figures from the epidemic of the Diamond Princess cruise ship, that the virus could only kill 10,000 in the United States, reports the American newspaper.
As a reminder, the epidemic has killed 443,275 people to date in the country.
What do we blame for this latest study?
In an article published on PubPeer, a site that allows the review of scientific articles, Australian epidemiologist Gideon Meyerowitz-Katz and Loni Besançon, researcher at Monash University, Australia identify a series of biases.
The first of these, the size of the sample studied.
The authors of the study only focused on 10 countries.
As a result, comparisons are found between Sweden and Iran "without taking into account the plethora of cultural, social and political differences between countries which could have an impact on the number of their cases during a pandemic".
The second problem is the choice to use Sweden and South Korea as a control group, that is to say as examples of countries which have not implemented restrictions.
Contrary to popular belief, limitations have indeed been put in place in Sweden, as we recently explained.
Separately, the country “recorded fewer cases at the start of the pandemic simply because it had one of the lowest numbers of tests of any country [considered in the study].
Comparing the number of cases in Sweden and Italy does not make much sense if we do not take into account the disparities in the number of tests carried out ”, specifies Gideon Meyerowitz-Katz to CheckNews, who notes that the authors admit a limit to their study, namely the fact of basing their calculations on the number of confirmed Covid cases in each country.
South Korea, for its part, has experienced one of the “longest periods of home learning in the world,” the two researchers point out.
Thus, the definitions of “restrictions” chosen by Ioannidis and his colleagues are too subjective according to Gideon Meyerowitz-Katz and Loni Besançon.
Morning essentials newsletter
A tour of the news to start the day
Subscribe to the newsletterAll newsletters
Another aspect is also overlooked, the correlation between the implementation of restrictive measures and the number of cases.
As this German economist points out, NPIs tend to harden as the growth of cases worsens.
“But the growth of cases may already be taking off when you just tightened your NPIs, resulting in exponential growth of cases before IPNs do anything.
This results in a positive correlation between INP and the growth of cases ”.
NPIs tend to get tougher the worse the case growth gets.
But case growth might already be taking off when you've just tightened your NPIs, making cases further grow exponentially before the NPIs do anything.
This results in a positive correlation between NPIs and case growth.
- Andreas Backhaus (@AndreasShrugged) January 13, 2021
It is therefore not the measures that increase the cases, but the health situation which is worsening, points out Andreas Backhaus.
What do the authors say?
Contacted by Liberation, they sent a response through the voice of one of the authors, Eran Bendavid.
He assures, among other things, that this study continues the work of another researcher, Solomon Hsiang, published in June: “The concerns raised by our study regarding the selection of countries, the selection of INP, the delays [between the implementation policy work and its effects], are all legitimate, and have their origin in Hsiang's article.
We did not want to change these aspects to facilitate comparisons and avoid concerns about the selective use of data.
So if a reader's conclusion is that these concerns are important enough to invalidate our results […] then the same conclusion applies to Hsiang's article, ”he says.
Hsiang had come to the opposite conclusions.
"Their article seems to reproduce our analysis, and applies our approach to other countries," Solomon Hsiang confirms to Liberation.
But according to him "the analyzed data in fact conclude that very restrictive INP have important and significant beneficial effects".
Have there been any other studies on the subject?
The effectiveness of the containment measures, if it is difficult to really grasp, has been the subject of other analyzes.
Last June, a study carried out by Imperial College London, based on the main measures taken such as the ban on public events, the restriction of movement or the closure of shops and schools had, for example, made it possible to establish that containment had saved 3.1 million people in 11 European countries, including France.
Another from the University of Berkeley, also published in June, assures that the containment would have avoided 530 million cases of Covid-19 in six countries: China, South Korea, United States, France, Iran and Italy.
