Data

What is National Syndromic Surveillance Program (NSSP) data?

National Syndromic Surveillance Program (NSSP) data are collected by the U.S. Centers for Disease Control and Prevention, in collaboration with local and state health departments and private sector partners. NSSP data include surveillance data tracking the percentage of emergency department (ED) visits due to influenza-like illness, including influenza, COVID-19, and RSV. NSSP data reporting of ED visits due to influenza-like illness began in late 2022. Data are available on a weekly time scale, with data from the prior week added Friday mornings.

What sub-state-level data are available?

At the sub-state level, reporting facilities can be aggregated to the level of health service area (HSA). HSAs are single- or multi-county clusters that correspond with healthcare-seeking patterns. All publicly-available county-level NSSP data are really HSA values, such that each county in the HSA will have the same value as the parent HSA.

A list of HSAs can be downloaded here.

Missingness

There is substantial missingness at the county level, especially for rural and lower-population locations.

Missingness has varied a bit since reporting began in late 2022. As of January 2025, NSSP does not report any data at the county level for the following states: Alaska, Alabama, Arkansas, Arizona, California, Florida, Missouri, North Dakota, New Hampshire, Ohio, Pennsylvania, Rhode Island, South Dakota, and Washington.

Of these, no county-level data is available since reporting began for Alaska, Alabama, Arkansas, Arizona, California, Florida, Missouri, North Dakota, New Hampshire, Ohio, South Dakota, and Washington.

We are open to working with public health partners from jurisdictions with no reported NSSP county-level data if they are able to facilitate access to an alternative source of sub-state-level data.

Accessing NSSP data

NSSP data can be accessed through the epidata API using epidatr (recommended) and covidcast R packages.

A vignette for using epidatr is available on CRAN.

library(epidatr)
library(covidcast)

Although the recommended way to pull data is with epidatr, we start by illustrating county-level data using the covidcast package’s plotting functionality. The following plot shows reporting counties as of Jan. 12, 2025. Recall all counties in an HSA will have the same value as the parent HSA.

nssp_county <- covidcast::covidcast_signal("nssp", "pct_ed_visits_influenza",
                                geo_type = "county",
                                time_type = "week")

plot(nssp_county, time_value = "2025-01-12")

The following code uses the epidatr package to return county-level data, where geo_value is the FIPS code of the county. Again, the values returned for an individual county are really those of the parent HSA.

epidatr::pub_covidcast(source = "nssp", signals = "pct_ed_visits_influenza", geo_type = "county", geo_values = "25015", time_type = "week")

Why use NSSP HSA data?

Metro-level disease trends and forecasts may differ markedly from state and national trends. Building the capacity to forecast at the metro-level may be crucial in times of outbreak planning and response.

NSSP HSA-level data are publicly available, making it quicker and easier to expand forecasting efforts to additional jurisdictions that report these data.

A more detailed visual breakdown of HSAs for each state can be found here. This document also provides line plots for each state illustrating the difference between the percentage of ED visits for the state (grey line) vs. each individual HSA in that state (orange line).