County-Level Influenza Risk Mapping: Assessing the Impact of PM2.5 and Socioeconomic Factors
SHRABANI TRIPATHY, Joseph V. Puthussery, Taveen Kapoor, Rajan K. Chakrabarty, Washington University in St. Louis
Abstract Number: 670
Working Group: Aerosol Exposure
Abstract
Seasonal influenza outbreaks occur annually across the USA, resulting in hospitalizations, fatalities, and substantial economic repercussions. Centers for Disease Control and Prevention (CDC), USA estimates that flu has resulted in 9.3 – 41 million illnesses and 4,900 – 51,000 deaths annually between 2010 and 2023. While influenza can affect individuals of any age group, certain populations are more susceptible to infection and severe disease. Recent research has linked air pollution, particularly PM2.5, with increased rates of influenza infection, establishing it as a potential determinant of influenza transmission. There has also been a link between socioeconomic factors and influenza, causing the disproportionate burden and suffering of the poor due to unequal conditions. However, the relative impacts of these parameters on infection outbreaks remain unclear. To address this gap, we employ a machine learning-based approach to assess the relative influence of various socio-economic parameters and annually averaged PM2.5 levels on influenza outbreaks. We quantify the extent of interrelation between these factors and influenza risk by analyzing over 150 different indicators—including population demographics, race, household information, insurance coverage, income, etc. The method assigns each indicator's relative degree of influence with the influenza infection rate at each county and helps identify the most important factors.
The resulting data is used to generate a county-level risk index for the contiguous United States (CONUS). These risk maps identify regions with elevated risk and highlight the contributing factors to higher risk in specific counties. Risk quantification helps identify major contributors to the total risk associated with a particular county, enabling the adoption of specific mitigation strategies to minimize risk. The easily interpretable results provided by these maps are invaluable for stakeholders and policymakers, highlighting disparities in health outcomes and aiding in targeted intervention and resource allocation to mitigate the impact of seasonal influenza.