Abstract
Globally, millions of lives are impacted every year by infectious diseases outbreaks. Comprehensive and innovative surveillance strategies aiming at early alert and timely containment of emerging and reemerging pathogens are a pressing priority. Shortcomings and delays in current pathogen surveillance practices further disturbed informing responses, interventions, and mitigation of recent pandemics, including H1N1 influenza and SARS-CoV-2. We present the design principles of the architecture for an early-alert surveillance system that leverages the vast available data landscape, including syndromic data from primary health care, drug sales, and rumors from the lay media and social media to identify areas with an increased number of cases of respiratory disease. In these potentially affected areas, an intensive and fast sample collection and advanced high-throughput genome sequencing analyses would inform on circulating known or novel pathogens by metagenomics-enabled pathogen characterization. Concurrently, the integration of bioclimatic and socioeconomic data, as well as transportation and mobility network data, into a data analytics platform, coupled with advanced mathematical modeling using artificial intelligence or machine learning, will enable more accurate estimation of outbreak spread risk. Such an approach aims to readily identify and characterize regions in the early stages of an outbreak development, as well as model risk and patterns of spread, informing targeted mitigation and control measures. A fully operational system must integrate diverse and robust data streams to translate data into actionable intelligence and actions, ultimately paving the way toward constructing next-generation surveillance systems.