Organization of Surveillance

Surveillance partners

In 2024, the biologists at the laboratories conducting bacteriological testing for healthcare facilities and participating in the Epibac network are as follows:

• List of correspondents

The network of laboratories

The network of microbiology laboratories was established in 1982 and has been collecting computerized and individual data since 1991. Participation is voluntary, and only laboratories that have provided data throughout the twelve months of the year are considered participants in the Epibac network. The continuous updating of the Epibac database takes into account the grouping of laboratories between facilities or within the same legal entity. Since 2010, the Epibac network has been extended to the overseas departments and regions (DROM).

The establishment of the national database

In mainland France, this surveillance network consists of facilities capable of treating patients with conditions monitored by Epibac. These are public and private hospitals with an annual volume of at least 5,000 admissions for short-stay medical care or that provide pediatric services.

In the DROMs, the database was established, for each department, to include all healthcare facilities treating patients with conditions monitored by Epibac, regardless of their status or volume of activity.

The case definition for Epibac surveillance

Invasive infections were defined by bacterial isolation and/or a positive PCR* in the blood for bacteremia and/or in cerebrospinal fluid (CSF) for meningitis. The bacteria studied are: Haemophilus influenzae, Neisseria meningitidis*, Listeria monocytogenes, Streptococcus pneumoniae, Streptococcus pyogenes (group A streptococcus), and Streptococcus agalactiae (group B streptococcus).

^{* The PCR detection method has been included in the case definition since 2009.}

Data Collection

From 1982 to 1990, the data collected covered all bacteria isolated from blood or CSF and were recorded on two forms: one for positive blood cultures and one for positive CSF cultures.

Starting in 1991, data were recorded on a single form, which made it possible to distinguish isolated bacteremias (infection with isolation in the blood but no associated isolation in the CSF).

A semi-annual or annual report is compiled by the laboratory and submitted to Santé publique France. For each sample, the following information is collected: the site and date of collection, the method used to detect the bacterium, the patient’s date of birth and sex, the serogroup if applicable, and the name and address of the microbiology laboratory providing the data. This may be the laboratory that collected the sample or the one that performed the analyses, depending on their organizational structure. Duplicate cases are identified based on the following variables: the site and date of the sample collection, the patient’s date of birth and sex, the identified bacterium, and the laboratory’s department.

Method for Estimating National Incidence

An annual incidence estimate is calculated to account for underreporting of cases by participating laboratories and the network’s partial coverage.

Network completeness

Network completeness corresponds to the number of cases reported by participating laboratories relative to the total estimated cases. It was assessed through several capture-recapture studies that measured underreporting of cases within this network for mainland France.

The completeness rate of data transmitted by participating laboratories has been estimated since 2002 and varied between 74% and 89% depending on the year and the bacteria studied. In mainland France, the applied value of this annual rate is 80%. In the overseas departments and regions (DROM), this rate is also used without any specific measurement having been conducted.

Network coverage

Until 2021, in the Epibac bulletins, network coverage was calculated annually by dividing the number of medical hospital stays in facilities where microbiology services were provided by Epibac laboratories by the total number of medical hospital stays nationwide. These medical hospitalization data are provided by the Annual Survey of Public and Private Healthcare Facilities (SAE) conducted by Drees. The coverage rate used for extrapolation was not specific to age, bacterial etiology, or clinical presentation (bacteremia/meningitis) of the cases.

Starting in 2023, the method for calculating network coverage was adapted to account for certain sociodemographic characteristics of the cases as well as the activity and region of the healthcare facilities covered by the network. It is based on data from the Medical Information Systems Program (PMSI). In the PMSI, MCO (Internal Medicine, Surgery, Obstetrics) hospital stays for a condition meeting the Epibac case definition were selected based on the reported diagnostic codes (ICD-10 codes or combinations of ICD-10 codes identifying each of the invasive infections monitored by this network). Since 2013, in the PMSI database, healthcare facilities have been recorded by geographic location rather than solely by legal entity. The network’s coverage rates have therefore been calculated annually starting in 2013. For previous years (2000 to 2012), an average coverage rate equivalent to that of the years from 2013 to 2019 was calculated.

At the national level, annual network coverage is calculated by dividing the number of hospital stays meeting the Epibac case definition at healthcare facilities where microbiological diagnoses are performed by Epibac laboratories by the total number of such stays in France, by age group and by bacterial etiology.

Estimation of national incidence

The incidence of invasive bacterial infections is estimated after two statistical adjustments:

• adjustment using the completeness rate of the number of cases reported by laboratories participating in the network. This process corrects for underreporting bias;

• extrapolation based on the coverage rate of the number of cases estimated from hospital activity at healthcare facilities. This method allows for the estimation of the total number of cases in France.

Incidence rates by year and age group are calculated by dividing the estimated number of cases per disease by the number of people residing in France for the year in question (source: INSEE**).

^{**Starting in 2006, the INSEE population projections used previously were replaced by localized population estimates (ELP). The comparative tables for previous years since 1991 have been corrected.}

Bulletin of the Surveillance Network for Invasive Bacterial Infections

For bulletins published since 2023, the indicators have been estimated using the new method for calculating network coverage, which has been in use since 2003.