Correlation of National Influenza Surveillance Data to the Local Experience

Background:  The utility of surveillance data, and specifically syndromic surveillance data, is determined by its level of accuracy in estimating the disease experience at the local level.  This project reviewed three data sources: BioSense, U.S. Influenza Sentinel Provider Surveillance System data, and Florida Influenza Surveillance System data with a focus on county influenza activity codes in order to evaluate the ability of each system to approximate influenza activity.
Data Sources:  BioSense data available in FL contains DoD/VA outpatient visit data and lab test data through LabCorp.  Interpretations based on BioSense data alone may not accurately approximate the influenza activity within a geographic area as small as a county unless it contains a military base.  However, Florida does have wide geographic representation within the U.S. Influenza Sentinel Provider Surveillance System.  For these data, 63% of Florida’s counties have at least one sentinel provider.  Even with the high percentage of coverage, the extrapolations of this data to geographic areas within Florida may not be appropriate.  Florida is a diverse state in terms of geography, latitude, and population which makes local data essential.  Every week, each county reports an influenza activity code to the Florida Influenza Surveillance System that is reflective of the local influenza activity level.
Methods:  Data from BioSense were compared to Florida sentinel physician data reported to the U.S. Influenza Sentinel Provider Surveillance System to determine the level of correlation between the two systems.  Those systems were then compared to the county activity code as reported through the Florida Influenza Surveillance System.
Results:  The correlation between the three systems was measured.
Conclusions:  National level syndromic surveillance data should be used cautiously in Florida.  The lack of regional and local representation does not lend itself to extrapolations to disparate areas of the state.