After a technical update from Public Health France, the indicators for monitoring the Covid-19 epidemic were revised downwards this Thursday, May 20, 2021. Indeed, the incidence rate had been overstated by 12% for several months in France due to the pseudonymization system which allowed for duplicates in the calculations. Passeport Santé explains.
To what extent has the incidence rate been overestimated for several months in France?
In France, to count the cases of Covid-19, the results of PCR, antigen, saliva and screening tests are entered in a database called SI-DEP. Each person tested is thus assigned a pseudonym (called pseudonymization) which guarantees the protection of their personal data. All the data are then used by several organizations: regional health agencies and Health Insurance for monitoring contact cases and Public Health France and the Ministry of Health for epidemic statistics.
However, people have been tested several times in a short period of time and in several different places. Likewise, if a person’s information was not entered in the same way (misspellings, different accents, upper or lower case, etc.), the pseudonymization system could assign two different pseudonyms. In addition, this has the consequence of creating duplicates.
A new calculation which brings down the incidence rate in France by 12%
It is in this context that the anonymization procedure has therefore been reviewed and corrected as announced by Public Health France on Thursday, May 20, 2021, declaring: “ As of this day, the indicators will be corrected in order to eliminate the duplicate effect, when several screening tests are carried out by the same person. “. The algorithm used has therefore been updated for the last three months and the new calculation rules show a drop in the incidence rate of 12% in France as well as an 8% drop in the positivity rate and 6 % of confirmed cases.
If the indicators were revised downwards, there would be no consequence on the dynamics of the epidemic. And for good reason, Public Health France explains that “ the new method of pseudonymization shows close curves and similar trends without consequence on the dynamics of the epidemic, its monitoring and its interpretation “.
