Using data from civil registry offices provided by INSEE and data from electronic death certificates—the sources for the SurSaUD® system—Public Health France conducts real-time mortality surveillance throughout the year.

Death reporting procedure and data transmission channels

A death is certified by a physician and documented in a death certificate. This certificate consists of two parts:

• an administrative section, which contains the administrative data of the deceased. It is intended for INSEE to update the National Directory of Natural Persons (RNIPP), which serves as the basis for social sector registries. This section is transmitted to INSEE by municipal civil registry offices;

• a medical and anonymous section, which contains the medical causes of death:

  • when the certificate is issued in paper format, it is sent to the Regional Health Agency (ARS), then forwarded to the Center for Epidemiology on Medical Causes of Death (CépiDc) at Inserm (Figure 1);

  • when the physician uses electronic death certification, the medical section is transmitted directly to the CépiDc and made available to the ARS (Figure 2).

By transmitting the data collected in these two sections, INSEE and Inserm contribute to the non-specific mortality surveillance carried out by Santé publique France.

Reactive mortality surveillance

All-cause mortality surveillance

Technical Organization

Mortality surveillance relies on daily electronic data transmission from civil registry offices to INSEE, which in turn sends all national data to Santé publique France after encryption (Figure 3).

For each death, the following information—derived solely from administrative records—is sent: date of birth, sex, date and municipality of death, and date of transmission of the information to Santé publique France. Due to the data transmission delay (90% of deaths recorded on a given day in the network’s municipalities are received within the following week), a weekly time interval is preferred for routine data analysis. A daily analysis can be conducted in the event of an occurrence likely to rapidly influence mortality, such as a heat wave, taking into account the limitations related to the transmission delay, which may be extended in the event of civil registry office closures (public holidays, absenteeism due to seasonal epidemics, lockdowns/curfews, etc.).

Rollout Across the Country

Real-time mortality surveillance based on data transmitted by INSEE began in 2004 with the 147 largest computerized municipalities in France, accounting for one-third of national mortality. Following this pilot, in late 2005, data transmission was expanded to all computerized municipalities, totaling 1,042 municipalities (mainland France and overseas territories). These municipalities then accounted for two-thirds of all deaths, or approximately 1,000 deaths per day.

This sample of municipalities has gradually expanded, and by 2024, Santé publique France will receive data from more than 16,000 municipalities. Routine analysis of weekly fluctuations in all-cause mortality requires data covering a historical period of at least five years. Thus, this analysis is based on a sample of 5,000 municipalities that have been submitting their data since 2011. It allows for monitoring of 85% of national mortality. This coverage includes the largest municipalities; mortality in rural or semi-rural areas is less well covered.

Mortality Monitoring by Cause

Technical Organization

Since 2007, physicians have been able to certify deaths electronically via a secure application. As soon as the physician electronically validates the certificate, the medical section’s information is immediately transmitted and becomes available within minutes to CépiDc-Inserm (Inserm’s Center for Epidemiology on Medical Causes of Death) and Santé publique France. Data from electronic death certificates thus provide timely access to individual information on the deceased (date of death, age, sex, type of place of death, municipality of death and residence), as well as the medical causes of death. These medical causes, as stated by the physician on the certificate, are available in free-text format.

Rollout Across the Country

The rollout of electronic death certificates has made little progress since its launch. During the H1N1 flu pandemic in 2009, an initial acceleration in adoption was observed. However, once the flu pandemic ended, adoption slowed in 2010, and only a slight increase was observed in the following years. For this reason, in 2013 and again in 2016, two ministerial directives to regional health agencies set a deployment target of 40% to be achieved by July 2018 in each region. Although no region met this target in 2018, some regions took action that led to real progress in electronic certification. In 2020, with the onset of the COVID-19 pandemic, an increase in deployment was observed, leading to 30% coverage of national mortality by early 2021 (Figure 4).

Since June 2022, electronic death certification has become mandatory in hospitals and long-term care facilities. In 2024, 47% of national mortality is recorded electronically. This coverage is higher for hospital deaths (70%). However, the system’s coverage remains limited for deaths occurring at home (10%) or in public places and in medical-social facilities (25%).

The progress of the rollout and the status of the submission of paper certificates to Inserm’s CépiDc can be tracked on the website https://opendata.idf.inserm.fr/cepidc/covid-19/.

Analysis Based on Mortality Data

Mortality exhibits a marked seasonal pattern, with higher numbers observed during the winter months, while the summer months show lower numbers (Figure 5).

The routine analysis of mortality fluctuations is based on comparing the recorded number of deaths from all causes with the expected number, estimated using a regression model developed and used by 29 countries/regions belonging to the EuroMomo consortium (European Mortality Monitoring, www.euromomo.eu). This model accounts for long-term trends and typical seasonal variations in mortality. The expected number of deaths therefore corresponds to the mortality that would be observed in the absence of any event likely to influence mortality, either upward or downward. This model is applied at the national, regional, and departmental levels, by age group.

National Weekly Surveillance Bulletin

Since 2014, a weekly bulletin monitoring all-cause mortality has been produced and posted on the Santé publique France website every week. When necessary, specific epidemiological reports are produced in addition to monitor particular health situations. All-cause mortality surveillance has been integrated into the system established by Santé publique France for monitoring and assessing the impact of the COVID-19 pandemic in France.

• View mortality bulletins

European Weekly Surveillance Bulletin

French data also contribute to a weekly European bulletin, which presents an analysis of aggregated data from the 29 countries/regions participating in EuroMomo and is published on www.euromomo.eu. This bulletin is transmitted to the ECDC, particularly to monitor the impact of seasonal winter epidemics and the COVID-19 epidemic.