The unvaccinated are still, overwhelmingly, more hospitalized than the vaccinated. But as the vaccination campaign continues to progress, the vaccinated could become the majority, making this indicator obsolete, even misleading, for the general public, about the effectiveness of the vaccine.
Images of a barbecue with friends, with a lively debate on the interest, or not, of vaccination. Then a voiceover that cuts the discussion: “We can debate everything except numbers. Today in France, 8 out of 10 people hospitalized because of Covid are not vaccinated. ” This little film is the last government clip to encourage the population to be vaccinated against Covid-19.
The figures presented are correct: they come from the Department of Research, Studies, Evaluation and Statistics (DREES). Thanks to the pairing of the Si-Vic (hospital data), Si-Dep (test results) and Si-Vac (vaccine data) databases, the statistics department of the Ministry of Health has been publishing regularly, since mid-summer , data on people hospitalized, deceased or tested positive for Covid-19, according to their vaccination status.
According to the latest publication on August 27, among those who entered the hospital with Covid-19 between August 9 and 15, 76% were unvaccinated (80% for critical care), compared to only 17 % completely vaccinated (13% in critical care). For deaths, “These ratios amount to 73% for unvaccinated people and 23% for fully vaccinated people”, adds the statistical agency.
For hospitalizations, taking into account only the non-vaccinated and the fully vaccinated (therefore not the partially vaccinated), the percentage of non-vaccinated in the hospital even drops from 76% to 82%. That is a 4.5 times higher risk of ending up hospitalized for the unvaccinated than for the vaccinated. But as accurate as they are, these numbers are not without their problems.
Two groups of the same size
First concern: these percentages, gross, do not really enlighten us. Because the share of each of the groups in the population is, at the start, not identical: over the same period, there were, in France, 35% of unvaccinated and 51% of vaccinated (the others being partially vaccinated) .For groups of equal populations, the proportion of unvaccinated in hospitals should be greater. Another distortion: the non-weighting according to age groups, while the populations most likely to be hospitalized – the oldest – are more vaccinated than the average.
To avoid some of these pitfalls, the Drees studies, in the same document, the vaccination status of two groups of the same size (1 million vaccinated and 1 million unvaccinated). From this it emerges that, from August 9 to 15, there were 94 entries of Covid-19 patients to the hospital in the non-vaccinated group, against 14 in the vaccinated group. In other words, for a comparable population size, there are 6.5 times more hospitalizations among the unvaccinated than among the fully vaccinated. And not 4.5, as one can deduce from the figures quoted in the government campaign.
The other major concern with these figures as a percentage of hospitalized according to vaccination status is that they may quickly become obsolete. They are in fact, as already mentioned, closely linked to the proportion of vaccinated and unvaccinated in the population. However, as the proportion of vaccinated people will increase in the population, the proportion of these in new hospitalizations will also increase.
If, for example, the population coverage reaches 90% of fully vaccinated, and the vaccine is 85% effective against severe forms, the vaccinated will become the majority in hospitals. Let us take a population of 100 people, made up of 90 vaccinated and 10 non-vaccinated, contaminated and susceptible to serious forms. As the vaccine is 15% fallible, about 14 people from the group of 90 vaccinated will end up in the hospital. In the second group, the unvaccinated, all will go to the hospital, but they will be only 10, so small their number has become. Ultimately, there will be 60% of the vaccinated (14 people) and 40% of the unvaccinated (10 people) in the hospital. Without this calling into question the efficacy of 85% of the vaccine.
In a few weeks, the government could well be retorted that the vaccinated represent a majority of hospitalized, against 20% today. It is then up to the executive to explain that his indicator, apart from a comedy, was not necessarily the most relevant.