why has it become so difficult to know where the epidemic is?

It’s the little music of the start of the school year: a positive friend here, a masked colleague there, a minister who postpones a trip to self-isolate… Covid-19, after being forgotten by many, is back in the daily lives of some of us. “We have an epidemic that is resuming, slowly and at low intensity”without requiring new measures, explained the Minister of Health, Aurélien Rousseau, last week on RTL. Friday September 15, faced with the increase in the circulation of the virus, he finally decided to change gear, bringing forward the next vaccination campaign by two weeks, which will therefore begin on October 2.

It is up to everyone to decide the level of precautions they want to adopt in their daily life. But if you want to rediscover the old reflex of observing the curve of positive cases or that of hospitalizations, and get your own idea of ​​the situation, it’s too late: they no longer exist. Since this summer, the collection of most data related to the virus has ended. Indicators still exist, but the image they give of the epidemic has become blurred.

The end of statistics on screening tests

Since the morning of July 1, the Si-Dep website has displayed an error message accompanied by a cross on a red background. Little known to the general public, this acronym designated the database fed by the results of Sars-CoV-2 screening tests, the content of which was easily accessible online and fed into the numerous dashboards of the health situation (this is also where the certificates appeared after the tests). It is no longer updated.

Its fate had been sealed since the summer of 2022 and the adoption of a law putting an end to the exceptional measures taken to fight the epidemic, which had set the expiration date of this file at June 30, 2023. Its usefulness as a thermometer of the circulation of the virus had already been undermined due to“less incentive for screening”, notes epidemiologist Mircea Sofonea. Especially since antigen and PCR tests have no longer been fully reimbursed since March 1, except for a few at-risk or very exposed categories. New software, Néo-Sidep, makes it possible to report the results of laboratory tests, but no longer those from pharmacies. And you will not see the color of these figures, which are not posted online.

At the hospital, only the reasons for going to the emergency room are collected

We also no longer have hospital data, whether it concerns the number of hospitalizations of positive people, entries into intensive care or deaths linked to the virus in establishments. Although no law requires it, their publication also ended on June 30 – hospital staff no longer had the obligation to report cases since the spring. Renouncing this very narrow mesh is “normal”, recognizes Mircea Sofonea, whose statistical work nevertheless depends on data. The intensity of screening and the increase in cases “required a lot of money and energy from the teams”, particularly at the public hospital, already under pressure. He “was necessary to evolve towards a system that resembles the surveillance of other respiratory viruses” like the flu, believes Yves Buisson, epidemiologist responsible for monitoring Covid-19 at the Academy of Medicine.

On what basis can we then affirm today, like the Minister of Health, that the epidemic “resumes slowly and at low intensity” ? The indicators come from data that was already collected before the pandemic. At the hospital, these are the reasons for going to the emergency room, collected by the Oscour network, and which are the subject of a weekly bulletin from Public Health France. In the bulletin published on September 12, it is thus specified that the passages “for suspicion of Covid-19” recorded an increase of 30% the week of September 4 to 10, compared to the previous week. That is, 3,651 people who consulted hospital emergencies for this reason, a level which had not been reached since the very end of 2022, but much lower than the epidemic peaks.

“The rise [des informations] is not systematic”warns Mircea Sofonea, who recalls the context “difficult” many emergency departments, where careful data collection is not the priority. Above all, these figures do not provide information on the evolution of patients once in the emergency room. We know what percentage of them are ultimately hospitalized (33%), but not the seriousness of their situation. “We are missing the story downstream, and the story upstream”regrets the statistician.

Three sources and three different images of the epidemic

For a broader picture of virus transmission, sensors should be kept outside the hospital. Every week, Public Health France publishes data related to the use of SOS Médecins, the network of home medical visits. Between September 4 and 10, 4,067 passages were linked to Covid-19, or 17% more than the previous week, according to the bulletin posted online on Monday. At the same time, the Sentinels network is also monitoring Covid-19 as part of its monitoring of numerous viral infections. Based on the observations of 560 general practitioners and pediatricians, statisticians estimate an incidence rate in the entire population. Between September 4 and 10, it was estimated at 82 cases per 100,000 inhabitants, which would represent nearly 55,000 new cases throughout the country (+55%), a level worthy of the most important peaks of 2022.

“We have to understand that we have a difference in quality” between this data, explains Mircea Sofonea. The SOS Médecins network thus covers 48 departments in France as well as Martinique, and disproportionately serves large cities. Sentinels, for its part, reflects the observations of a small portion of the territory’s doctors, even if they are then statistically corrected. “It is undoubtedly the source which is closest to the actual circulation of the virus in the general population”estimates the researcher, who consults both Sentinelles, SOS Médecins and Oscour to get an idea of ​​the situation.

“The problem is when we have disconnects between these indicators, for example when all three rise, but not at all at the same speed”, he underlines, a scenario which applies very well to the situation at the beginning of September. To understand these discrepancies, researchers are reduced to hypotheses. When data fluctuates, is it a sign of a new wave or just a “greater attention” carried by the virus? If the increase is greater in town than in hospitals, “does this mean that a wave mainly affects young people”less severely affected? Without widespread screening, pieces of the puzzle are missing that would allow us to understand it. “It creates fog”summarizes Mircea Sofonea who, like other colleagues, observes whether the trends are confirmed “among our neighbors who are still monitoring the epidemic more rigorously”like the United Kingdom or Switzerland.

Important vigilance to preserve the healthcare system

For the researcher at the University of Montpellier, France is poorly prepared for the possibility of a significant new wave: “Surveillance has deteriorated and we have not replaced it with anything else, nor with a road map that would allow resources to be redeployed in the event of an epidemic resumption.” However, there are ways to maintain surveillance at lower cost, such as“random sampling from the population”, by regularly testing groups of French people representative of the general population to assess the situation. He also mentions the monitoring of wastewater, for which the Academy of Medicine was also mobilized, recalls Yves Buisson: “Today, this monitoring is carried out in 300 wastewater treatment plants, the network should be multiplied by 10”, he believes. These samples would allow very early detection of a resurgence of the epidemic. “We also monitor the wastewater, to see”assured Aurélien Rousseau on September 8 on the air of Sud Radio. But no data has yet been made public.

Before investing or not in these new tools, however, a central question must be answered: is it still necessary to monitor Covid-19 so carefully? Yves Buisson puts the seriousness of this poorly controlled circulation into perspective, because the legacy of vaccination and previous contaminations “helps limit admissions to intensive care and deaths”. This less critical context “does not require rapid, day-to-day reactions”, he judges. Especially since, despite less precise tools, “we detected the rise in incidence in July”, and its link with the arrival in France of the variant nicknamed Eris. The health authorities have in fact maintained genomic surveillance of Sars-CoV-2, by sequencing the virus from samples of the patients tested. Enough to better target vaccination and sound the alarm if a particularly worrying variant begins to establish itself in the population: it is often this phenomenon which leads to a new large wave.

“Covid-19 is no longer, on its own, a threat to the health system”, also recognizes Mircea Sofonea. But the flu and RSV, a respiratory virus dangerous for infants, returned to their usual level with the end of health restrictions, and were already weighing heavily on the hospital before the pandemic. The persistence of Covid-19 adds an additional cause of hospitalizations and “contributes to the risk of tension”. This is the reason why the health authorities now want to combine the anti-flu and anti-Covid vaccination campaigns. It will then be necessary to convince, without curves clearly showing the danger, that some still need a booster dose.