5th wave of Covid-19: are the Institut Pasteur forecasts all correct?

The Omicron wave is on the decline in France. It’s time for the assessment of this fifth wave for epidemiologists. And precisely, on February 15, the Pasteur Institute published a report which compares the figures of the epidemic to its predictions. And it would seem that the French medical establishment was very little mistaken in its figures.

“High” transmissibility and “low” severity

As of December 27, the date of the first estimates by the Institut Pasteur, the impact of the Omicron variant had been defined as being “less severe but highly transmissible”. It is on the basis of this observation, but also on the following hypotheses that the Institute was able to make its forecasts, while specifying that “the spread of the virus was difficult to anticipate” and that this element would affect the veracity of the assumptions that will be made.

In addition to this variable, the establishment also took into account:

• Vaccination coverage;
• The behavior of the French, for whom the institute has counted on a drop in contacts of “10 to 20%” during the month of January”;
• Shorter hospital stays than with the Delta variant;
• Fewer visits to intensive care.

Fairly accurate projections

According to the figures published in its report of February 15, the researchers at the Pasteur Institute had been right. “The model anticipated 121,000 cumulative hospitalizations between December 1, 2021 and February 11, 2022, there were 118,000″. Another estimated figure that turned out to be very close to reality: hospitalizations.

Where the institute expected a peak on January 21 with 2,650 hospital admissions, it really took place on January 24, three days later, with 2,600 entries that day. Ditto for the peak of contamination, reached on January 21 when it had been estimated for January 16.

A few mistakes though

The epidemiologists, however, qualified some of their overly optimistic results. This concerns the length of stays in critical care and the number of hospital beds required. For hospital stays, they were initially judged to be half as long with Omicron as with Delta, i.e. 6 days instead of 12.”8 days instead of 6 days for Omicron patients would have better described the peak size” finally believe the researchers in their report.

And on the subject of hospitalizations, the Institute has counted on 18,000 necessary beds, when it finally took 19,000.

Different study modalities

“Errors” explained in the report by three elements: taking into account only “patients registered as newly hospitalized in conventional hospitalization or critical care”a work “on the date of the event (date of admission, date of entry into critical care, etc.) and not on the date of registration of the event”and “taking into account all patients hospitalized with a SARS-CoV-2 infection, regardless of the reason for hospitalization” detail the report.

An explanation that does not seem to convince epidemiologist Martin Blachier, who spoke about the report in a rather ironic tweet: “Pasteur simulations: are predictions with so much uncertainty really serious? What can we do with that? The advantage is that with an umbrella so wide open, there is no risk of getting wet…”.

Future predictions from the Institut Pasteur are optimistic, with a drop in hospital admissions and more hospitalizations related to the Delta variant. See you in a few months to see if they are confirmed…