Demonstrating the utility of influenza syndromic surveillance with high-volume medical claims in the United States

Background & AimsSeasonal respiratory viruses such as influenza cause substantial illness in the United States, over-whelming healthcare facilities and reducing economic productivity. Effective surveillance of these viruses is therefore critical for timely risk communication, strategic resource allocation, and coor-dinated public health responses that mitigate viral spread. Syndromic surveillance, which tracks patient symptoms rather than confirmed diagnostic results, plays an essential role in disease monitoring. While this form of surveillance aids in early trend detection, widespread adoption, particularly for unobserved disease burden estimation, has been hindered by insufficient validation against laboratory-confirmed cases and the lack of accurate syndromic profiles. In this study, we leverage a high-volume medical claims database to develop data-driven syndromic profiles for influenza based on symptom patterns from lab-confirmed cases. We then apply these syndromic profiles to estimate total symptomatic case dynamics and burden (both tested and untested) by geography and demography. Methods & ResultsWe analyzed a large medical claims database covering healthcare visits for over 40% of the United States population annually from 2016 to 2020. We used a regression modeling approach to develop syndromic profiles based on lab-confirmed cases of influenza. With these models, we estimated spatiotemporal dynamics at the county-week scale and season prevalence from time series data on symptom occurrence in healthcare settings. We validated our estimates by comparing them with traditional surveillance data. Symptom-inferred disease estimates aggregated to state and national-levels showed strong agreement with existing surveillance systems in both spatiotemporal trends and magnitude of disease activity. Across all seasons examined, influenza prevalence was spatially heterogeneous, with the southern United States experiencing the highest burden. ImplicationsOptimized syndromic surveillance has promise to serve as a representative, fine-scale, and admin-istratively efficient system for tracking infectious diseases. Public health priorities such as disease forecasting, transmission parameter estimation, and hospital bed allocation can benefit from high-resolution data and disease-specific syndromic profiles. Overall, disease-specific syndromic surveillance will provide more precise monitoring, strengthening public health preparedness and response capabilities.