Infectious disease epidemics remain one of the great threats to humanity, with increased disease emergence and the threat of a global pandemic, there are renewed efforts to forecast and predict these events. We propose to develop tools to measure the effectiveness and impact of epidemic nowcasting, forecasting and intervention. In order to do this, we will compare outcomes across interventions to determine the cost-effectiveness of interventions and optimize resource allocation. The goal of improving the precision of outbreak predictions is to increase the timeliness and accuracy of the response and, ultimately, to reduce costs in terms of mortality, morbidity and economies. We can measure the accuracy of epidemic predictions post-hoc by comparing predicted timing, extent, region and affected populations to actual epidemic parameters as they occur. We propose this as one method of validating and informing predictive models and tools. New tools for epidemic forecasting are under continuous development, facilitated by cutting-edge digital technologies including AI and machine learning. We will compare the accuracy of these innovative approaches against current gold standards in epidemic forecasting, where they exist, e.g., the Swiss influenza Sentinella network. We will also assess the uptake of forecasts by Ministries of Health (e.g., policy reform, resource allocation, emergency declarations), hospital management (e.g., extra staffing, vaccine stockpiles, PPE) lay press (e.g., televised coverage of warnings, newspaper articles, public information campaigns), pharmaceutical companies (e.g., vaccine sales, stockpiling) and public response (e.g., vaccine uptake, precautionary measures, sentiment analysis on social media).
Dr Danny Sheath, University of Geneva
Prof. Antoine Flahault, University of Geneva
Prof. Ramanan Laxminarayan, University of Princeton