RÉSUMÉ
 
Un nouveau variant du SARS-CoV-2 (VOC-202012/01, ici appelée VOC) est apparue fin 2020 au Royaume-Uni. Des estimations préliminaires indiquent que 1,4% des cas de COVID-19 nouvellement diagnostiqués en w01 dans le pays sont dus au VOC. En utilisant le modèle développé par l'INSERM pour répondre à la pandémie de COVID-19, nous fournissons des projections des trajectoires épidémiques dans les semaines à venir en raison de l'augmentation de la circulation du variant britannique dans le pays. Nous estimons que le variant deviendrait dominant en France entre fin février et mi-mars, en fonction de l'évolution épidémique et de l'augmentation estimée de la transmissibilité du VOC. Les nouvelles hospitalisations hebdomadaires devraient atteindre le niveau du pic de la première vague (environ 25,000 hospitalisations) entre mi-février et début avril, en l’absence d’interventions. Ces résultats montrent la nécessité de renforcer les mesures de distanciation sociale et d’accélérer la campagne de vaccination pour faire face à la menace du variant VOC.

Figure 1. Predicted weekly number of hospitalizations due to non-VOC and VOC strain of SARS-CoV-2. The orange curve indicates hospitalizations due to non-VOC strain. Black dots correspond to weekly hospitalization data (up to week 1, 2021). Projections are obtained assuming Reff(non-VOC) = 1.0 (left), Reff(non-VOC) = 1.1 (center), Reff(non-VOC) = 1.2 (right). Top panels: The dashed green curve represents weekly hospitalization due to VOC strain, assuming a 50% increasing in transmissibility with respect to the wild strain, corresponding to Reff(VOC) = 1.5, 1.65, 1.8 respectively (from left to right).The solid green curve indicates the total (non-VOC and VOC) number of hospitalized cases. Bottom panels: As in the top panels, assuming a 70% increase in transmissibility for VOC strain with respect to the non-VOC strain, corresponding to Reff(VOC) = 1.7, 1.9, 2 respectively (from left to right). In all panels: curves represent median values; shaded areas indicate 95% probability ranges obtained from 500 stochastic runs.
 

Figure 2. Estimated week of VOC dominance. Estimated time at which the weekly number of new VOC cases accounts for 50% of all weekly new cases. Dots indicate median values. Error bars represent 95% probability ranges obtained from 500 stochastic runs.  
 

Table 1. Estimated week in which the number of hospitalizations exceeds the peak reached during the first and the second lockdown. We use as threshold values 25,000 weekly hospitalizations for the peak of the first wave, and 20,000 weekly hospitalizations for the peak of the second lockdown.
 
 
KEY FINDINGS
 

• Given a 70% increase in transmissibility, we expect that the VOC strain will become dominant by the end of February / beginning of March, even assuming a constant non-VOC epidemic activity. A 50% increase in transmissibility would delay the date of dominance by approximately two weeks (Figures 1 and 2). An important rise of cases, however, is expected already in the weeks preceding the dominance of VOC, with weekly hospitalizations approaching the values reached in the second wave.

• In the best-case scenario (stable epidemic activity due to the non-VOC virus) and assuming a 50% increase in VOC transmissibility, we would reach the second wave’s peak of hospitalizations by the end of March (week 13), if no interventions are put in place (Table 1). The level of hospitalizations of the first wave’s peak would be reached approximately one week later. If the increase in VOC transmissibility is higher, these same levels would be reached three weeks before.  

• In the worst-case scenario, with Reff(non-VOC) = 1.2 and a 70% increase in transmissibility, we expect that the number of weekly hospitalizations will exceed 20,000 by mid-February (week 7), if no interventions are put in place.

• These results support the need for strengthened social distancing measures to face the threat of the VOC variant. Accelerating the vaccination campaign will help mitigate the curve in the upcoming monthsand reduce the hospitalizations.
  

RÉSUMÉ
 
Compte tenu de nouveaux variants et de l’incidence élevée enregistrée en France, les mesures de distanciation sociale ont été progressivement renforcéesen janvier 2021. En utilisant un modèle mathématique à 2 souches calibré sur les estimations de la première enquête Flash pour la circulation du variant britannique (B.1.1.7), nous avons estimé que ces mesures ont conduit à la stabilisation des hospitalisations, résultante d’un équilibre entre le recul de la souche historique (celle en circulation avant l’arrivé du variant; taux de reproduction effectif estimé <1) et la croissance rapide du variant britannique. Ces estimations sont validées par les résultats préliminaires de la deuxième enquête Flash du 27 janvier. Le variant britannique est attendu devenir majoritaire fin février-début mars en France, avec une grande hétérogénéité géographique (mi-février en Île-de-France). Dans l'absence de mesures de contrôle renforcée, une croissance rapide des cas est attendue dans les semaines à venir.

https://i.imgur.com/8gElNx0.png
Figure 1. Projected weekly hospitalizations due to SARS-CoV-2 historical strain and B.1.1.7 variant in France and Île-de-France. From left to right: scenario assuming a strengthening of social distancing (SD) measures from w06, compatible with the situation experienced in the last weeks of the second lockdown; scenario based on estimates up to w05 and assuming no additional changes; scenario assuming a relaxation of social distancing measures from w06, compatible with the situation in early January. Dots correspond to weekly hospitalization data used for fitting the model. The solid black curve refers to the overall trajectory, due to the concurrent circulation of the historical strain (dashed green curve) and B.1.1.7 variant (solid green curve), assuming 50% increase in transmissibility (40% and 60% increases are shown for sensitivity in the SI). The shaded area around the curve corresponds to the 95% confidence interval; it is not shown for the single strains, for the sake of visualization. The second wave is shown for reference, together with indications of the timing of social distancing measures.
 
https://i.imgur.com/67yh2lz.png
Figure 2. B.1.1.7 projected prevalence over time and estimated dominance week in France and and Île-de-France. Left and center: estimated percentage of B.1.1.7 cases over time, assuming 50% (range 40-60%) increase in transmissibility for the variant. Points represent the consolidated estimates from Flash1 survey; point-lines represent the preliminary estimates from the Flash2 survey. We considered S-gene target failure percentages of positive tests for each territory4 and estimated the proportion of B.1.1.7 between 70% (Flash1 survey’s proportion) and 100% (indicated with the point). Right: estimated week of B.1.1.7 dominance, assuming 50% (range 40-60%) increase in transmissibility for the variant, and considering the estimated scenario (middle point) and the scenarios with strengthening (lighter color, left point) and relaxation (darker color, right) of social distancing measures.  
 
KEY FINDINGS
 

• Based on the estimated B.1.1.7 prevalence in early January, the model predicted that the observed plateau results from the counterbalance between two opposing dynamics: a decreasing circulation of the historical strain (with effective reproductive numbers Re(France)=0.95 (0.94-0.96), Re(IDF)=0.95 (0.93-0.97)) opposed to the exponential increase of the variant. Control measures were not enough to lead to a decline in the variant spread due to its more efficient transmission.

• The rapid increase of B.1.1.7 prevalence over time estimated by the model is in agreement with the preliminary estimates resulting from the Flash2 survey. The model predicts that the variant would become dominant in week 9 (8-9) in France, and 7 in Île-de-France. Changes in social distancing would negligibly affect the estimated dominance date, as the measures would have a concurrent effect on both strains. Estimated variations are larger for changes in the assumed increase in variant transmissibility.

• Strengthening social distancing measures show clear advantages in gaining time to extend the plateau before the expected resurgence of cases. The scenario proposed here, with a 10% reduction in current Re, is less efficient than the early phase of the November mild lockdown, with schools open. It corresponds to the situation estimated in the later phase, when relaxation emerged, mobility increased and telework progressively reduced. The additional measures taken on January 31, a possible slowdown induced by winter school holidays, as well as a strengthening of the test-trace-isolate strategy5 may contribute to further stretch the current plateau.

• In absence of more rigorous and intensified social distancing, B.1.1.7 will drive a rapid growth in the next weeks, with an earlier timing in those regions reporting a large presence of the variant (e.g. Île-de-France). Increasing vaccination overage is key10, but its mitigating impact is expected to become important starting April, according to estimated rollout plans and assuming efficacy only against symptomatic forms of infection (see Appendix).
  

RÉSUMÉ
 
Compte tenu de nouveaux variants et de l’incidence élevée enregistrée en France, les mesures de distanciation sociale ont été progressivement renforcéesen janvier 2021. En utilisant un modèle mathématique à 2 souches calibré sur les estimations de la première enquête Flash pour la circulation du variant britannique (B.1.1.7), nous avons estimé que ces mesures ont conduit à la stabilisation des hospitalisations, résultante d’un équilibre entre le recul de la souche historique (celle en circulation avant l’arrivé du variant; taux de reproduction effectif estimé <1) et la croissance rapide du variant britannique. Ces estimations sont validées par les résultats préliminaires de la deuxième enquête Flash du 27 janvier. Le variant britannique est attendu devenir majoritaire fin février-début mars en France, avec une grande hétérogénéité géographique (mi-février en Île-de-France). Dans l'absence de mesures de contrôle renforcée, une croissance rapide des cas est attendue dans les semaines à venir.

https://i.imgur.com/8gElNx0.png
Figure 1. Projected weekly hospitalizations due to SARS-CoV-2 historical strain and B.1.1.7 variant in France and Île-de-France. From left to right: scenario assuming a strengthening of social distancing (SD) measures from w06, compatible with the situation experienced in the last weeks of the second lockdown; scenario based on estimates up to w05 and assuming no additional changes; scenario assuming a relaxation of social distancing measures from w06, compatible with the situation in early January. Dots correspond to weekly hospitalization data used for fitting the model. The solid black curve refers to the overall trajectory, due to the concurrent circulation of the historical strain (dashed green curve) and B.1.1.7 variant (solid green curve), assuming 50% increase in transmissibility (40% and 60% increases are shown for sensitivity in the SI). The shaded area around the curve corresponds to the 95% confidence interval; it is not shown for the single strains, for the sake of visualization. The second wave is shown for reference, together with indications of the timing of social distancing measures.
 
https://i.imgur.com/67yh2lz.png
Figure 2. B.1.1.7 projected prevalence over time and estimated dominance week in France and and Île-de-France. Left and center: estimated percentage of B.1.1.7 cases over time, assuming 50% (range 40-60%) increase in transmissibility for the variant. Points represent the consolidated estimates from Flash1 survey; point-lines represent the preliminary estimates from the Flash2 survey. We considered S-gene target failure percentages of positive tests for each territory4 and estimated the proportion of B.1.1.7 between 70% (Flash1 survey’s proportion) and 100% (indicated with the point). Right: estimated week of B.1.1.7 dominance, assuming 50% (range 40-60%) increase in transmissibility for the variant, and considering the estimated scenario (middle point) and the scenarios with strengthening (lighter color, left point) and relaxation (darker color, right) of social distancing measures.  
 
KEY FINDINGS
 

• Based on the estimated B.1.1.7 prevalence in early January, the model predicted that the observed plateau results from the counterbalance between two opposing dynamics: a decreasing circulation of the historical strain (with effective reproductive numbers Re(France)=0.95 (0.94-0.96), Re(IDF)=0.95 (0.93-0.97)) opposed to the exponential increase of the variant. Control measures were not enough to lead to a decline in the variant spread due to its more efficient transmission.

• The rapid increase of B.1.1.7 prevalence over time estimated by the model is in agreement with the preliminary estimates resulting from the Flash2 survey. The model predicts that the variant would become dominant in week 9 (8-9) in France, and 7 in Île-de-France. Changes in social distancing would negligibly affect the estimated dominance date, as the measures would have a concurrent effect on both strains. Estimated variations are larger for changes in the assumed increase in variant transmissibility.

• Strengthening social distancing measures show clear advantages in gaining time to extend the plateau before the expected resurgence of cases. The scenario proposed here, with a 10% reduction in current Re, is less efficient than the early phase of the November mild lockdown, with schools open. It corresponds to the situation estimated in the later phase, when relaxation emerged, mobility increased and telework progressively reduced. The additional measures taken on January 31, a possible slowdown induced by winter school holidays, as well as a strengthening of the test-trace-isolate strategy5 may contribute to further stretch the current plateau.

• In absence of more rigorous and intensified social distancing, B.1.1.7 will drive a rapid growth in the next weeks, with an earlier timing in those regions reporting a large presence of the variant (e.g. Île-de-France). Increasing vaccination overage is key10, but its mitigating impact is expected to become important starting April, according to estimated rollout plans and assuming efficacy only against symptomatic forms of infection (see Appendix).
  

Contactée par Libération, 1'équipe de l'institut insiste sur le fait que les chiffres n'ont jamais été présentés comme des prédictions, et qu'il a surtout toujours été précisé qu'il S'agissait de scénarios évolutifs. Mais concède malgré tout un défaut dans les premières modélisations: en s'appuyant sur des données trop éloignées dans le temps, l'institut n'a pas pris en compte le tassement de l'épidémie observé sur la seconde partie de septembre, et qui serait due aux mesures prises par les autorités ou à un changement général de comportement des Français. «Ce sont des scénarios qu'on fait en partant du principe que les hospitalisations  vont se poursuivre au même rythme, explique Simon Cauchemez, épidémiologiste à l'institut Pasteur.