Skip to main content

Estimating the burden of SARS-CoV-2 in France

Our take —

This study estimated the extent of SARS-CoV-2 infection and rates of hospitalization, ICU admission, and deaths in the French population. The study estimated that the lockdown reduced the reproductive number from 2.90 (95% CrI: 2.80–2.99) to 0.67 (95% CrI: 0.65–0.68). It estimated 2.8 million (4.4%) people in France were infected up to May 11, 2020, but was unable to account for closed communities such as retirement homes, and may have underestimated the number of people infected.

Study design


Study population and setting

Authors used passive hospital surveillance data from the French national public health agency and Poisson Likelihood models to estimate the number of hospitalizations, ICU admissions, and deaths from COVID-19 in 8 age groups per sex. Authors used active surveillance data from the Diamond Princess cruise ship and Poisson Likelihood models to estimate the number of deaths aboard the ship. Both estimates were extrapolated to describe the dynamics of SARS-CoV-2 infection and COVID-19 disease burden in France through May 11, 2020. Finally, authors assessed how SARS-CoV-2 transmission was impacted by country-wide lockdowns.

Summary of Main Findings

Overall, results estimated 2.8 million people were infected with SARS-CoV-2 in France, approximately 4.4% of the population. An estimated 3.6% (95% Credible Interval: 2.1–5.6%) of infected persons were eventually hospitalized for COVID-19, and of those hospitalized, 19.0% (95% CrI: 18.7–19.4%) were admitted to the ICU. Median time from hospitalization to ICU admission was 1.5 days and increased with age. Regardless of ICU admission, 18.1% (95% CrI: 17.8–18.4%) of those hospitalized died . Among those infected (regardless of hospitalization), 0.001% of people <20 years old died, compared to 10.1% (95% CrI: 6.0–15.6%) of those >80 years old. The R0 (basic reproductive number) prior to lockdown was estimated at 2.90 (95% CrI: 2.80–2.99). After the lockdown, R (effective reproductive number) dropped to 0.67 (95% CrI: 0.65 – 0.68), representing a 77% (95% CI: 76–78%) reduction in transmission.

Study Strengths

The study integrated both passive and active surveillance data, which may make estimates more robust as compared to hospital-only based reporting. Authors used the date of disease occurrence, rather than date of report, which corrected for delays in reporting. Authors validated the modeling framework using known data.


The study did not include non-hospitalized COVID-19-related deaths and excluded institutionalized populations (i.e., persons living in retirement communities), which may have resulted in the number of infections being underestimated within the population and limits generalizability.

Value added

This study estimated country-wide infection and case fatality rates, as well as the effect lockdown efforts had on the reproductive rate in France. These findings have important policy implications and results may be applicable to other high-income countries with similar timelines regarding implementation and easing of mobility restrictions.

This review was posted on: 21 July 2020