The week is once again “decisive”: confinement or not? As every week for three months, the executive will have to decide using the available health data.
But these are wrong! as detailed in Le Parisien, several weeks several indicators display overestimated data since the widespread deployment of antigenic tests in pharmacies. Or several weeks.
With a difference, variable according to the departments, which can reach 10% at the national level.
Which indicators affected?
The Directorate General of Health admits errors concerning the number of positive tests.
An error which therefore affects the number of new cases per day, but also the incidence rate (new cases over 7 days / 100,000 inhabitants) and the positivity rate.
Why?
Since the test sites have multiplied (pharmacies, laboratories, hospital, doctor …), the health authorities must “duplicate” the results.
Example: an antigen test comes back positive. To be confirmed or for variant screening reasons, it must then be confirmed by a second PCR test.
Of course, everything is planned so that these two tests count only once: the Vitale card, in principle used each time, makes it possible to avoid counting two positive tests of the same person twice.
But “entry” errors do occur, especially during antigen testing in pharmacies, where information is entered by hand into the computer system, before being anonymized.
If there is no entry error (name, first name, security number, date of birth, contact details …) and the data are the same as those of the Vitale card, no problem.
But the slightest deviation can lead to a “different” file number … and therefore to a duplicate.
Limited consequences
These errors would concern up to 10% or even 11% of positive tests.
But their number varies enormously according to the departments: it depends strongly on the number of tests carried out in pharmacy, the search for variants, etc.
On the data themselves, the 26,343 new contaminations recorded on Sunday could be “only about 24,000. And the incidence rate (238 new cases per 100,000 inhabitants) could be” only “215.
The same goes for the positivity rate, which counts the number of positive tests for 100 tests performed.
An incidence rate of 250 is now considered as an alert threshold. However, these errors only change the rate itself, not the dynamics of the epidemic: the error will have a consequence on the level, not on the increase or decrease in the circulation of the virus.
At the current rate, the threshold would be “really” exceeded at the national level 24 or 48 hours after its “statistical” exceedance.
Above all, these indicators are only a part of the data which make it possible to decide on health measures.
Hospital data is not affected by this problem. But they take longer to come up.
Santé Publique France intends to solve the problem, perhaps by systematizing the use of the Vitale card, regardless of the place of the test.
–
