[LUM#13] In the Mood for CoV
Mood aims to improve global health monitoring and speed up decision-making by providing health agencies with the right epidemic intelligence tools. This European project, launched in January, could not have come at a better time.
While science moves at a different pace than current events, sometimes the two collide in a rather brutal way. Conceived in a world where Covid-19 did not yet exist, the Mood project (Monitoring outbreaks for disease surveillance in a data science context), which brings together 25 partners in 12 European countries and the US, aims to assess and prioritize the needs of health agencies in order to co-develop new epidemic intelligence tools. The goal is to improve the early detection of emerging pathogens and enable health managers to respond as quickly as possible.
Launched last January, the project has become—by necessity—a full-scale experiment. "With the arrival of Covid, the European Commission very quickly asked us to redeploy resources to meet the immediate needs of the agencies," explains Renaud Lancelot, researcher at the Astre* laboratory and scientific coordinator of the project. "The European Centre for Disease Prevention and Control (ECDC), based in Stockholm, was our main point of contact during this period."
A matter of time
Because in times of crisis, time is of the essence, the team coordinated by Renaud Lancelot carried out the research, data collection, and analysis that the overwhelmed health agencies were unable to do. "We were asked to work on human mobility, to provide agencies with maps and time series representing this mobility so that they could be integrated into models of the spread of the epidemic" and thus to test different scenarios for controlling and monitoring the pandemic, foremost among which was lockdown.
Another focus for the Mood consortium is the analysis and evaluation of data produced by these same agencies.The ECDC has compiled a time series of mortality cases in Europe and worldwide. This work has been unanimously praised, but paradoxically, the agencies have been unable to analyze it due to a lack of time and resources." We have summarized this series in the form of a dashboard so that the data collected by the ECDC can be used by the agency and so that decision-making can be based on the most recent information possible."
Detecting weak signals
Mood also proposes to improve the detection of weak signals of an emerging epidemic. "These signals must correspond to a proven potential danger," says Renaud Lancelot. "In the case of COVID-19, there have already been several outbreaks of coronavirus, so we know that this can lead to a major pandemic."
In this hunt for weak signals, anything goes. "We use all available means to detect health risks among all accessible information , " explains the epidemiologist. If we want to draw a parallel, epidemic intelligence is a bit like espionage. " While researchers rely on official reports produced by agencies, they do not neglect unofficial information: analysis of social networks (see box), rumors, non-specific signs... And here again, it's all a question of tools.
Generating secure data collection flows, creating dynamic maps and tables to track developments, developing text mining software and smartphone applications, etc. The range of tools available for estimating the probability of a virus outbreak is varied, but as Renaud Lancelot points out, only their adoption by health agencies and the establishment of a short information circuit can have a real impact on public health. "That's why it's so important to invest in fundamental approaches and to have the tools, methods, and networks already in place, because when a crisis hits, there's no time for improvisation."
Social media, the weapon of mass prevention
Anticipating the level of local transmission through social media analysis is the aim of the DigEpi project, which stands for "digital epidemiology," funded bythe French National Research Agency and the Occitanie region. The principle is simple: analyze what people are saying on social networks, particularly Twitter, by feeding keywords such as Covid-19, hydroxychloroquine, and lockdown into algorithms. "We then observe how these themes fluctuate and compare these curves with mathematical models of actual infections in order to anticipate a possible peak in infections," explains Benjamin Roche, director of research atIRD.
These analyses also provide anthropologists working on this project with valuable insights into the dynamics of the feelings expressed by the population. They also make it possible to characterize the behavior and reactions of the public to the measures taken. "Will people isolate themselves, go out, wear masks, spread rumors, or seek information about treatments? All these elements give us a better understanding of the public's perception of the various measures," explains the epidemiologist. Launched in April in four major French cities, including Montpellier, this study will last 18 months before possibly being extended to all French departments and to Mexico.
Find UM podcasts now available on your favorite platform (Spotify, Deezer, Apple Podcasts, Amazon Music, etc.).
*UMR ASTRE Animals – Health – Territories – Risks – Ecosystems (CIRAD – INRAE)