In mainland France between November 17, 2015, and March 11, 2016, the CNR identified 83 cases of Salmonella Dublin infection, making this Salmonella Dublin outbreak one of the largest in France in recent years.
Salmonella Dublin is a serotype of Salmonella that is particularly invasive in humans and more frequently responsible for severe illness and higher mortality rates compared to other Salmonella serotypes. Salmonella Dublin can be isolated in cattle, with raw milk and raw-milk cheeses serving as the typical transmission vectors in foodborne illnesses.
This investigation was innovative in at least two respects. From a microbiological standpoint, the CNR employed two methods. The first, Multilocus Variable-Number Tandem Repeat Analysis (MLVA), had already been used in previous foodborne illness investigations. The second, Whole Genome Sequencing (WGS), was used for the first time in this Salmonella Dublin outbreak and allowed for better differentiation of case clusters compared to MLVA. The use of WGS can facilitate the characterization of the link between human cases and potential sources of contamination and can also provide a better picture of the overall context of the outbreak by linking it to other concurrent or past outbreaks. WGS has been routinely used at the CNR since 2017. From an epidemiological perspective, for the first time in France, the recruitment of controls for an epidemiological investigation was conducted using a cohort of individuals participating in an online research project, the GrippeNet study.
This investigation underscores the importance of multidisciplinary collaboration to identify potential sources of contamination and implement appropriate control and management measures. We would like to take this opportunity to thank everyone who contributed to this investigation.

Indeed, the 83 cases were identified in 12 of the 13 regions of metropolitan France, with 19 cases (23%) coming from the Bourgogne-Franche-Comté region alone. The median age of the cases was 70 years. In total, 10 deaths (12%) were reported, with no information available on the cause of death. We were able to interview 63 cases (76% of cases); 68% had been hospitalized, and 66% of cases for which information was available reported chronic conditions (asthma, cancer, etc.).
Microbiological, epidemiological, and environmental investigations pointed to two raw-milk cheeses, Morbier and Mont d’Or, as the likely sources of contamination for this Salmonella Dublin outbreak. Retrospective use of whole-genome sequencing (WGS) also confirmed the occurrence of previous Salmonella Dublin outbreaks in 2012.
The results of this investigation pointed to several cheese producers in the same region as potential sources of the outbreak. It was difficult to precisely identify the batches consumed by the cases because some cheeses were sold by the slice and also due to the high frequency of consumption of these cheeses by the cases.
Following this investigation, the group of Morbier and Mont d’Or producers implemented an action plan, including systematic testing for Salmonella in batches of Morbier and Mont d’Or sold starting in February 2016, more regular farm visits by veterinarians, and improved detection and culling of infected livestock.

In early 2012, the GrippeNet research project was launched by the Sentinelles network (Inserm – Sorbonne University) and Santé publique France. Its goal is to collect epidemiological data on influenza directly from the public via the internet, on an anonymous basis. Each week, participants report any symptoms they may have experienced since their last login.
Recruiting our participants from the GrippeNet study for our investigation proved highly effective, offering two main advantages. The first was the speed with which the process was implemented. It took four weeks from the start of the SalmonellaDublin investigation to sending the questionnaire—which included questions about food consumption during early January 2016—to GrippeNet participants, in order to compare their responses with those of the cases. The second advantage was the high participation rate among “Grippenauts,” with 2,914 (47%) of the 6,200 participants in the GrippeNet study completing our questionnaire. The majority (92%) did so within two days of the questionnaire being sent. In total, the questionnaire collection period among “Grippenauts” lasted 12 days.
This method, unprecedented in France but already tested in the United Kingdom, also served as a pilot study for us to assess the relevance of this approach for future food poisoning investigations.