移動データを使って病気の広がりを追跡する

タイトル: Human mobility patterns to inform sampling sites for early pathogen detection and routes of spread: a network modeling and validation study

概要: BackgroundDetecting and foreseeing pathogen dispersion is crucial in preventing widespread disease transmission. Human mobility is a critical issue in human transmission of infectious agents. Through a mobility data-driven approach, we determined municipalities in Brazil that could make up an advanced sentinel network, allowing for early detection of circulating pathogens and their associated transmission routes. MethodsWe compiled a comprehensive dataset on intercity mobility spanning air, road, and waterway transport, and constructed a graph-based representation of Brazils mobility network. The Ford-Fulkerson algorithm, coupled with centrality measures, were employed to rank cities according to their suitability as sentinel hubs. FindingsOur results disentangle the complex transportation network of Brazil, with flights alone transporting 79{middle dot}9 million (CI 58{middle dot}3 to 10{middle dot}1 million) passengers annually during 2017-22, seasonal peaks occurring in late spring and summer, and roadways with a maximum capacity of 78{middle dot}3 million passengers weekly. We ranked the 5,570 Brazilian cities to offer flexibility in prioritizing locations for early pathogen detection through clinical sample collection. Our findings are validated by epidemiological and genetic data independently collected during the SARS-CoV-2 pandemic period. The mobility-based spread model defined here was able to recapitulate the actual dissemination patterns observed during the pandemic. By providing essential clues for effective pathogen surveillance, our results have the potential to inform public health policy and improve future pandemic response efforts. InterpretationOur results unlock the potential of designing country-wide clinical sample collection networks using data-informed approaches, an innovative practice that can improve current surveillance systems. FundingRockefeller Foundation grant 2023-PPI-007 awarded to MB-N. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSWe searched PubMed on Jun 1, 2023, without language or date restrictions, for the following query: ("mobility network*" OR "transport* network*" OR "sentinel network*" OR "surveillance network*") AND "model*" AND "surveillance". The 469 search results were systematically evaluated, and we identified seven original research studies that applied modeling-based approaches to inform the placement, design, or layout of surveillance/sentinel networks. Of these seven studies, four aimed at optimizing the layout of networks for the monitoring of influenza-like illnesses (ILI), while the others aimed at detecting problems arising from the use of medicines based on pharmacy surveillance; detecting the reporting of common acute conditions through a sentinel network of general practitioners; and optimizing the surveillance strategy for plant pests (S. noctilio). Most studies employed maximum coverage algorithms that aim to maximize the protected population. Only a single study incorporated mobility patterns to inform the planning of site placement. Studies that involved ILI sentinel networks were geographically restricted to two United States states (Iowa and Texas), and only one study performed a comprehensive whole of United States modeling. Added value of this studyDespite the urgent need to improve the capacity and timeliness of clinical sample collection for public health surveillance, very few studies have tackled the design problem for optimal placement of these sampling sites, and even fewer have used large-scale mobility data to inform these design choices in an epidemiologically-relevant way. Our work contributes to this challenge by leveraging airline/roadway/fluvial mobility data for Brazil that, converted into a graph-based representation and using network metrics, allowed us to pinpoint an optimal layout strategy that could improve the current flu surveillance network of this country. Using data collected during the COVID-19 pandemic, we validated the transmission routes and pathways of SARS-CoV-2 spread, confirming that the mobility data-informed spread scenarios recapitulated the actual dissemination of the virus. Implications of all the available evidenceMobility data, coupled with network-centered approaches, can complement the identification of strategic locations for early pathogen detection and spread routes.