![[ Visit Client Website ]](images/banner.jpg)
BACKGROUND: The ability of the National Syndromic Surveillance Program (NSSP), formerly known as BioSense, to conduct syndromic surveillance for nationwide all-hazard situational awareness and enhanced response to hazardous events and outbreaks is in part based on the representativeness of its surveillance system, i.e., its ability to accurately describe health-related events over time and in the population. An objective of NSSP is to receive in its Cloud-based health information system (platform) sufficient near real-time electronic data from U.S. nonfederal hospital emergency department (ED) visits in all 50 states and the District of Columbia to support regional and national syndromic surveillance. However, limited information is available on representativeness of these ED visits.
METHODS: We examined representativeness of NSSP hospital ED visits using data from the 2012 American Hospital Association (AHA) Annual Survey Database, the most recent data available, and linked databases from the Dartmouth Atlas of Healthcare, the U.S. Census, and the Health Resource and Services Administration, along with information from the participating jurisdictions on which AHA-listed hospitals in their jurisdictions were submitting data to the NSSP platform in October, 2014. Using those data, we compared national ED visits with ED visits from hospitals reported to be submitting data to the NSSP Platform.
RESULTS: We found that approximately 60.4 million out of 135.6 million (~45%) ED visits identified by AHA were reported to have been submitted to NSSP. However, representativeness varied substantially by region, with a much smaller percentage of ED visits included from the West and New England regions and a higher percentage from the Midwest. Coverage by state ranged from 0% to approximately 100%. ED visits in the NSSP platform were similar to national ED visits in terms of many characteristics of the hospitals (e.g., children’s v. teaching hospitals), and hospital ED catchment areas (e.g., percentage of residents in the catchment area below the poverty level). However, ED visits in hospitals with low annual ED visits, hospitals with rural trauma centers, and hospitals serving populations with high percentages of Hispanics and Asians were under-represented.
CONCLUSIONS: This novel assessment of representativeness of NSSP ED visits, using publicly available health and population data sets, found that NSSP ED visit data are largely representative in some locations and by some ED hospital and catchment area population characteristics. However, the NSSP could improve representativeness by increasing participation by jurisdictions in specific geographic areas and among hospitals with certain characteristics serving specific under-represented populations.