This surveillance system draws on various data sources, some of which predate the pandemic and others established specifically to collect data on COVID-19-related mortality. Each source helps track deaths occurring within the country, with each providing specific information (see Box – Mortality Surveillance).
As of May 31, the number of deaths associated with the COVID-19 epidemic ranged from 25,000 (estimated based on INSEE civil registry data) to 30,000 deaths (estimated based on hospital and nursing home data). Among the deaths associated with COVID-19, nearly 18,900 occurred in hospitals and 10,320 were reported by nursing homes.

Anne Fouillet, head of mortality surveillance at the Data Support, Processing, and Analysis Directorate (DATA) of Santé publique France, presents this report and the public health questions it raises.

The report published today indicates excess mortality of between 25,000 and 30,000 deaths. Which populations are most affected? Is this an exceptional mortality situation? How does this compare to other European countries?

Of the 25,030 excess deaths from all causes, 23,400 of the deceased were aged 65 or older (representing over 93% of excess deaths, an increase of 18.2% compared to expected mortality in this age group), and 1,510 were aged 15–64 (an increase of 7% compared to expected mortality in this age group). Within this latter age group, it was people aged 45–64 who were most affected. Conversely, mortality among children under 15 decreased by 14% over the period at the national level.
Compared to expected mortality, the proportion of excess deaths among men was higher than that among women.
At the peak in week 14 (March 30 to April 5), mortality nationwide reached a level approaching that recorded during the two weeks of the August 2003 heatwave. The estimated number of deaths during this period also exceeded the cumulative total of deaths estimated during the five previous winter epidemics.
According to the ECDC (European Centre for Disease Prevention and Control) for the European Union and the United Kingdom, France ranks sixth among countries reporting the highest number of COVID-19-related deaths per 100,000 inhabitants since the start of the epidemic, following Belgium, the United Kingdom, Spain, Italy, and Sweden.
Across the 24 European countries/regions participating in the European consortium for all-cause mortality surveillance (EuroMOMO), an excess of 173,350 deaths was estimated for the period from March 2 to May 31, 2020.

You observe significant disparities in mortality across regions. Which regions are the hardest hit? How can this situation be explained?

The estimated excess deaths at the national level do indeed mask significant regional disparities. Île-de-France is the region that has been most impacted (+64% compared to expected mortality), followed by Grand-Est (+37%) and in particular the departments of Bas-Rhin, Haut-Rhin, and Moselle. The Auvergne-Rhône-Alpes, Bourgogne-Franche-Comté, and Hauts-de-France regions also experienced a significant excess of deaths (between +15% and +20% compared to expected mortality).
In mainland France, Île-de-France was the only region in which excess deaths were observed among people aged 15–64 of all genders. In Île-de-France, both people aged 15–44 and those aged 45–64 were affected by excess deaths.

The geographic heterogeneity of excess mortality primarily reflects the dynamics of the epidemic across the country, which hit the Grand-Est and Île-de-France regions particularly hard, and to a lesser extent the Hauts-de-France, Auvergne-Rhône-Alpes, and Bourgogne-Franche-Comté regions. Conversely, regions less affected by the epidemic show moderate excess mortality, or even lower-than-normal mortality in two regions.

Geographical disparities in mortality at finer scales, particularly at the municipal or inter-municipal level in the Île-de-France region, have also been highlighted by complementary studies.

It is too early to explain the heterogeneity of the impact across the region; the explanatory factors are multifactorial and intertwined, including the sociodemographic and economic characteristics of the populations residing in these areas, as well as medical and health geography. In-depth research drawing on data from multiple sources (socioeconomic, health, geographic, SAPRIS and EPIVOC cohorts) will be necessary to better understand this heterogeneity.

The medical section of death certificates is the only source providing medical causes of death. To what extent has the limited adoption of electronic certification hindered current knowledge regarding COVID-19-related mortality?

Electronic certification is a source that is both responsive and precise in terms of the information available for deceased individuals. It provides individual information (age, sex, geographic area of death, type of place of death) and the medical causes of death, making it possible to track both COVID-19-associated mortality and non-associated mortality and to characterize the health status of the deceased. This can help to promptly identify comorbidities and risk factors that contribute to the development of severe forms of COVID-19.

At the time of the epidemic, the insufficient and uneven rollout of electronic death certification across the country did not allow for a robust estimate of the direct and overall impact of the epidemic. In particular, its uneven deployment depending on the type of place of death—more frequently used in public hospitals and very rarely used to certify deaths at home—prevents the generalization of the observed results to the entire country.

In-depth analyses will be necessary to provide a final estimate of excess mortality directly and indirectly associated with COVID-19, once all paper certificates have been received and processed by Inserm-CépiDc. This analysis is scheduled for late 2020 or early 2021.

It is now essential to roll out the use of electronic certification as soon as possible throughout the country, whether in healthcare facilities or at home, to ensure comprehensive and responsive surveillance of mortality by cause of death on a routine basis and particularly in the event of a threat to public health.