Sci Rep
. 2025 May 5;15(1):15638.
doi: 10.1038/s41598-025-99078-6. Revisiting the link between COVID-19 incidence and infection fatality rate during the first pandemic wave
Benjamin Glemain 1 , Charles Assaad 2 , Walid Ghosn 3 , Paul Moulaire 2 , Xavier de Lamballerie 4 , Marie Zins 5 6 , Gianluca Severi 7 8 , Mathilde Touvier 9 , Jean-François Deleuze 10 ; SAPRIS-SERO study group; Nathanaël Lapidus 11 , Fabrice Carrat # 11 , Pierre-Yves Ancel # 12 , Marie-Aline Charles 12 , Gianluca Severi 7 8 , Mathilde Touvier 9 , Marie Zins 5 6 , Sofiane Kab 6 , Adeline Renuy 6 , Stephane Le-Got 6 , Celine Ribet 6 , Mireille Pellicer 6 , Emmanuel Wiernik 6 , Marcel Goldberg 6 , Fanny Artaud 7 , Pascale Gerbouin-Rérolle 7 , Mélody Enguix 7 , Camille Laplanche 7 , Roselyn Gomes-Rima 7 , Lyan Hoang 7 , Emmanuelle Correia 7 , Alpha Amadou Barry 7 , Nadège Senina 7 , Julien Allegre 9 , Fabien Szabo de Edelenyi 9 , Nathalie Druesne-Pecollo 9 , Younes Esseddik 9 , Serge Hercberg 9 , Mélanie Deschasaux 9 , Marie-Aline Charles 12 , Valérie Benhammou 13 , Anass Ritmi 14 , Laetitia Marchand 14 , Cecile Zaros 14 , Elodie Lordmi 14 , Adriana Candea 14 , Sophie de Visme 14 , Thierry Simeon 14 , Xavier Thierry 14 , Bertrand Geay 14 , Marie-Noelle Dufourg 14 , Karen Milcent 14 , Delphine Rahib 15 , Nathalie Lydie 15 , Clovis Lusivika-Nzinga 11 , Gregory Pannetier 11 , Nathanael Lapidus 11 , Isabelle Goderel 11 , Céline Dorival 11 , Jérôme Nicol 11 , Olivier Robineau 11 , Cindy Lai 16 , Liza Belhadji 16 , Hélène Esperou 16 , Sandrine Couffin-Cadiergues 16 , Jean-Marie Gagliolo 17 , Hélène Blanché 10 , Jean-Marc Sébaoun 10 , Jean-Christophe Beaudoin 10 , Laetitia Gressin 10 , Valérie Morel 10 , Ouissam Ouili 10 , Jean-François Deleuze 10 , Laetitia Ninove 4 , Stéphane Priet 4 , Paola Mariela Saba Villarroel 4 , Toscane Fourié 4 , Souand Mohamed Ali 4 , Abdenour Amroun 4 , Morgan Seston 4 , Nazli Ayhan 4 , Boris Pastorino 4 , Xavier de Lamballerie 4
Affiliations
Several studies found an association between COVID-19 incidence, cumulated over the first pandemic wave, and the risk of death for infected individuals. They attributed this association to hospital overload. We studied this association across the French departments using 82,467 serological samples and a hierarchical Bayesian model with spatial smoothing. In high-incidence areas, we hypothesized that hospital overload would increase infection fatality rate (IFR) without increasing infection hospitalization rate (IHR). The analyses were adjusted for intensive care beds per capita, age of the population, and diabetes prevalence (as a surrogate for obesity). We found that increasing departmental incidence from 3 to 9% rose IFR from 0.42 to 1.14% (difference 0.72%, 95% CI 0.49-1.01%), and IHR from 1.66 to 3.61% (difference 1.94%, 95% CI 1.18-2.80%). An increase in incidence from 6 to 12% in people under 60 was associated with an increased proportion of people over 60 among those infected, from 11.6 to 17.4% (difference 5.8%, 95% CI 2.9-8.8%). Higher incidence increased the risk of death for infected individuals and their risk of hospitalization by the same magnitude. These findings could be explained by a higher age among infected individuals in high-incidence areas, rather by than hospital overload.
Keywords: Bayesian statistics; COVID-19; Causal graph; Hierarchical modeling; Hospital overload; Spatial modeling.
. 2025 May 5;15(1):15638.
doi: 10.1038/s41598-025-99078-6. Revisiting the link between COVID-19 incidence and infection fatality rate during the first pandemic wave
Benjamin Glemain 1 , Charles Assaad 2 , Walid Ghosn 3 , Paul Moulaire 2 , Xavier de Lamballerie 4 , Marie Zins 5 6 , Gianluca Severi 7 8 , Mathilde Touvier 9 , Jean-François Deleuze 10 ; SAPRIS-SERO study group; Nathanaël Lapidus 11 , Fabrice Carrat # 11 , Pierre-Yves Ancel # 12 , Marie-Aline Charles 12 , Gianluca Severi 7 8 , Mathilde Touvier 9 , Marie Zins 5 6 , Sofiane Kab 6 , Adeline Renuy 6 , Stephane Le-Got 6 , Celine Ribet 6 , Mireille Pellicer 6 , Emmanuel Wiernik 6 , Marcel Goldberg 6 , Fanny Artaud 7 , Pascale Gerbouin-Rérolle 7 , Mélody Enguix 7 , Camille Laplanche 7 , Roselyn Gomes-Rima 7 , Lyan Hoang 7 , Emmanuelle Correia 7 , Alpha Amadou Barry 7 , Nadège Senina 7 , Julien Allegre 9 , Fabien Szabo de Edelenyi 9 , Nathalie Druesne-Pecollo 9 , Younes Esseddik 9 , Serge Hercberg 9 , Mélanie Deschasaux 9 , Marie-Aline Charles 12 , Valérie Benhammou 13 , Anass Ritmi 14 , Laetitia Marchand 14 , Cecile Zaros 14 , Elodie Lordmi 14 , Adriana Candea 14 , Sophie de Visme 14 , Thierry Simeon 14 , Xavier Thierry 14 , Bertrand Geay 14 , Marie-Noelle Dufourg 14 , Karen Milcent 14 , Delphine Rahib 15 , Nathalie Lydie 15 , Clovis Lusivika-Nzinga 11 , Gregory Pannetier 11 , Nathanael Lapidus 11 , Isabelle Goderel 11 , Céline Dorival 11 , Jérôme Nicol 11 , Olivier Robineau 11 , Cindy Lai 16 , Liza Belhadji 16 , Hélène Esperou 16 , Sandrine Couffin-Cadiergues 16 , Jean-Marie Gagliolo 17 , Hélène Blanché 10 , Jean-Marc Sébaoun 10 , Jean-Christophe Beaudoin 10 , Laetitia Gressin 10 , Valérie Morel 10 , Ouissam Ouili 10 , Jean-François Deleuze 10 , Laetitia Ninove 4 , Stéphane Priet 4 , Paola Mariela Saba Villarroel 4 , Toscane Fourié 4 , Souand Mohamed Ali 4 , Abdenour Amroun 4 , Morgan Seston 4 , Nazli Ayhan 4 , Boris Pastorino 4 , Xavier de Lamballerie 4
Affiliations
- PMID: 40325150
- PMCID: PMC12053613
- DOI: 10.1038/s41598-025-99078-6
Several studies found an association between COVID-19 incidence, cumulated over the first pandemic wave, and the risk of death for infected individuals. They attributed this association to hospital overload. We studied this association across the French departments using 82,467 serological samples and a hierarchical Bayesian model with spatial smoothing. In high-incidence areas, we hypothesized that hospital overload would increase infection fatality rate (IFR) without increasing infection hospitalization rate (IHR). The analyses were adjusted for intensive care beds per capita, age of the population, and diabetes prevalence (as a surrogate for obesity). We found that increasing departmental incidence from 3 to 9% rose IFR from 0.42 to 1.14% (difference 0.72%, 95% CI 0.49-1.01%), and IHR from 1.66 to 3.61% (difference 1.94%, 95% CI 1.18-2.80%). An increase in incidence from 6 to 12% in people under 60 was associated with an increased proportion of people over 60 among those infected, from 11.6 to 17.4% (difference 5.8%, 95% CI 2.9-8.8%). Higher incidence increased the risk of death for infected individuals and their risk of hospitalization by the same magnitude. These findings could be explained by a higher age among infected individuals in high-incidence areas, rather by than hospital overload.
Keywords: Bayesian statistics; COVID-19; Causal graph; Hierarchical modeling; Hospital overload; Spatial modeling.