NoCOV

As COVID19 is spreading worldwide, nowcasting, short and mid-term forecasts of the epidemic in the general population is more than ever critical to support decision making. As only a fraction of COVID19 cases can be confirmed by virological testing yet, clinical surveillance remains of utmost importance to monitor the disease. In France, the Sentinelles network has updated its surveillance protocol to monitor all Acute Respiratory Infections in this respect. Data collected by the network will provide an essential source of information. However, the analysis of these data will be complicated by the many epidemiological unknowns regarding the disease and the unprecedented epidemic situation - scale of the outbreak, containment measures and behavioral change. NoCOV will rely on sophisticated numerical approaches to extract meaningful information from surveillance data. We will provide estimates of COVID19 incidence and obtain short- and medium-term forecasts to inform on possible sustainable intervention measures in the case of a resurgence of the epidemic past summer 2020. The project will be articulated as follows. We will adapt an agent-based model (ABM) developed by the team to COVID19. This model is based on a realistic reconstruction of the French population, accounting for individual characteristics and daily activities in different settings - household, school/workplace and community. We will fit the model to Sentinelles data together with other data sources (drug sales, symptoms reported by Grippenet.fr) with cutting edge inference and statistical learning techniques to provide nowcasts and short-term projections. We will estimate key parameters: e.g. virus transmission potential; susceptibility, transmissibility and disease severity by age class; effect of seasonal factors. This will make it possible to understand the evolution of the population immunity as the outbreak evolves. We will provide reliable medium-term projections on the epidemic unfolding, quantifying the likelihood of epidemic rebound and large scale spread after summer 2020. Eventually, we will test in silico sustainable measures to contain such a wave: large-scale testing and quarantine of contacts, selective social distancing applied to targeted population groups, and increased self-awareness through mobile applications. Throughout these steps NoCOV aims at providing continuous monitoring of the epidemic situation and ground knowledge for public health decisions at different time horizons. Project results will be made immediately public through reports and shared within the REACTing working group and the public health authorities.