Extended Data Fig. 6: Model predictions versus serological estimates. | Nature

Extended Data Fig. 6: Model predictions versus serological estimates.

From: Underdetection of cases of COVID-19 in France threatens epidemic control

Extended Data Fig. 6

a, Serological estimates per region from three seroprevalence studies versus corresponding model projections. Error bars correspond to 95% confidence intervals. bm, For each region, the panel shows the predicted percentage of infected population over time (red curves and shaded areas for median and 95% confidence interval) and serological estimates of EpiCov26 (red dots), Santé publique France (SpF)25 (green dots) and ref. 7 (black dots in b, c, m). Medians and 95% confidence intervals for model projections are obtained from n = 500 independent stochastic runs. Plots are reported for all 12 regions of mainland France : Île-de-France (b), Grand Est (c), Provence–Alpes–Cote d’Azur (d), Hauts-de-France (e), Centre-Val de Loire (f), Auvergne–Rhône–Alpes (g), Bourgogne–Franche– Comté (h), Occitanie (i), Normandy (j), Pays de la Loire (k), Brittany (l) and Nouvelle Aquitaine (m).

Back to article page