A clear majority of French people would be in favor of using an application that records their social interactions and warns them if they have been in contact with a sick person from Covid-19, or warns those they have encountered if they are themselves infected. This is the lesson of a survey published Tuesday, March 31, conducted with a representative sample of more than 1,000 French people having a mobile phone on March 26 and 27. This study was commissioned by a research team from the British University of Oxford, which is working on this type of application to fight the pandemic.
These researchers mathematically modeled the effect of a tracking application that immediately identifies people at risk of becoming infected before they even have symptoms of Covid-19 and estimated that such an application was likely to “Control the epidemic without the need for prolonged and very costly general containment measures”. Their research has just been published in the prestigious journal Science.
An application using Bluetooth
The researchers presented respondents with the device they imagined: an application, installed on a smartphone and using Bluetooth wireless technology, capable of detecting whether another mobile phone equipped with this same application is in the immediate vicinity.
The envisaged application does not access anything other than Bluetooth (no access to the directory, messages, etc.) and does not allow geolocation: it simply registers devices equipped with the same application having been in his immediate environment for at least 15 minutes, a situation presenting a risk of infection with the new coronavirus.
In the system presented to respondents, when the owner of such an application is diagnosed with Covid-19, those whom the patient has encountered are immediately notified and they are asked by the health authorities to put themselves in strict quarantine. Those alerted do not know who put them at risk of infection or where. The more than 1,000 French respondents whose answers have been taken into account first had to answer several questions correctly to make sure they understood the system devised by the researchers.
In this context, they would be almost 48% of the people questioned to install it ” without a doubt “ and 31% to do so ” probably “, a percentage that hardly changes with age. Eight out of ten people therefore directly plan to install such an application. This probability increases in the event that an infection appears in its environment: they would then be almost two thirds to install it ” without a doubt “. Ironically, more than 93% of those surveyed would follow the quarantine guidelines received from the app – more than people who would install the app in the first place. A share that increases if, at this quarantine, is matched the possibility of being tested with Covid-19.
Figures to put into perspective, according to the authors of the study. “We were only able to discuss the operating mode and installation of the application in very general terms, while the precise details of implementation could greatly affect installation decisions”, they write, while concluding that “Even if 80% of our respondents expressed a desire to install such an application if it was available, the installation rate could actually be much lower”.
Respondents favorable to an “automatic” installation
Almost two out of three respondents would favor the installation of this application automatically by telephone operators – a mode of operation which does not seem to be technically feasible, however. A similar percentage would keep ” probably “ or ” without a doubt “ the app on their phone if it was installed there automatically.
The study identifies three main obstacles to the widespread adoption of this application, a sine qua non for its effectiveness: the risks of hacking the phone on which the application is installed, as well as the possibility that this surveillance could be extended even after the pandemic (a quarter of respondents respectively) and the risk of an increase in anxiety linked to the use of this application (one fifth of those questioned). The technical difficulties of installation or activation remain marginal.
In addition, more than half of those surveyed would be in favor of the data collected by the application being made available to researchers after the end of the epidemic. Finally, the researchers note that in Germany, the United Kingdom and Italy, where the researchers also conducted a survey, the results are “Very similar” to those observed in France. “As the application is only useful if there is a sufficient number of users, we judge our results to be hopeful about the viability of such an approach”, welcome the researchers.
“We are not currently working on instruments that would make this tracking mandatory” – Edouard Philippe, April 1
This survey does not only support the scientific approach and the project of the Oxford research team. She intervenes as applications projects supposed to identify individuals at risk of being infected for having known a patient are multiplying in the world. China and Singapore – the latter country in a fashion similar to that imagined by researchers – have already deployed such applications. In France, “The advisability of setting up a digital strategy for identifying people who have been in contact with infected people” is part of the roadmap of the second scientific committee on the Covid-19 set up by the Elysée.
Interviewed in the National Assembly Wednesday 1er In April, Prime Minister Edouard Philippe hinted that a follow-up mechanism could be envisaged if it were based on voluntary work. “Today we have no legal instrument and we are not currently working on instruments that would make this tracking mandatory. The question can be asked whether it could be voluntary and produce effects in this case. This is a question which, at this stage, remains open “, said the Prime Minister to MPs.
Do not draw general observations
However, it would be unwise to draw general lessons from this survey and its results in no way indicate the acceptability of an application that is more demanding in terms of personal data, for example tracking the geolocation of its users. The rigor of this survey serves its representativeness: the respondents were only questioned on a specific project, formulated by a respectable university and known worldwide.
The acceptability of such a project would necessarily vary depending on the precise terms of an application that may be deployed in France, in particular with regard to the protection of personal data. This study is not without interest, however, since the approach of applications registering the proximity of other users thanks to Bluetooth connectivity, like that imagined at Oxford, has been adopted by a large number of projects of this type.
Finally, the survey did not submit to respondents some important variables, which could vary the acceptability of such an application. This is for example the case of the security of the data collected, which is not raised, or that of who exactly has this data and implements this application: only an indefinite “Health authority” is mentioned. In addition, the survey takes for granted that the application is perfectly functional, bug-free, and effectively contributes to the fight against the pandemic. An idea with which it is de facto difficult to disagree on the merits.
Our selection of articles on the coronavirus