FluSurv-NET Hospitalizations

Attribute Details
Source Name flusurv
Data Source CDC FluSurv-NET via CDC GRASP API
Geographic Levels state, msa, nation, misc
Temporal Granularity Week, ending Saturday
Reporting Cadence Weekly
Temporal Scope Start 2003-10-04 (2003w40)
Temporal Scope End Ongoing
Extra Key Columns None
License Public Domain US Government

This source reproduces the legacy V3 flusurv endpoint. Each demographic and virus-specific rate is published as an individual signal under standard V5 conventions. See Relationship to V3.

Table of contents

  1. Overview
  2. Indicators (Signals)
    1. Overall and Virus Type
    2. Age Strata
    3. Race, Ethnicity, and Sex
  3. Estimation
    1. Metric Definition
    2. Smoothing
    3. Uncertainty
    4. Temporal Handling
    5. Geographic Handling
  4. Relationship to V3
  5. Schema
    1. Columns
  6. Missingness & Privacy
  7. Limitations
  8. Lag & Backfill
  9. Source and Licensing

Overview

The Influenza Hospitalization Surveillance Network (FluSurv-NET), administered by the CDC, conducts population-based surveillance for laboratory-confirmed influenza-associated hospitalizations. FluSurv-NET covers select counties in Emerging Infections Program (EIP) states and additional Influenza Hospitalization Surveillance Project (IHSP) states, representing approximately 9 percent of the U.S. population.

Delphi fetches hospitalization data weekly from the CDC FluSurv-NET portal. The dataset reports hospitalization rates per 100,000 residents overall and stratified by age, race and ethnicity, sex, and influenza type (A and B).

Indicators (Signals)

All signals represent weekly hospitalization rates per 100,000 residents.

Overall and Virus Type

Indicator Name Pathogen or Disease Metric Type Description
rate_overall Influenza Rate per 100k Overall weekly influenza hospitalization rate.
rate_flu_a Influenza A Rate per 100k Weekly hospitalization rate for confirmed Influenza A infections.
rate_flu_b Influenza B Rate per 100k Weekly hospitalization rate for confirmed Influenza B infections.

Age Strata

Indicator Name Pathogen or Disease Metric Type Description
rate_age_0 Influenza Rate per 100k Weekly rate for ages 0–4 years.
rate_age_1 Influenza Rate per 100k Weekly rate for ages 5–17 years.
rate_age_2 Influenza Rate per 100k Weekly rate for ages 18–49 years.
rate_age_3 Influenza Rate per 100k Weekly rate for ages 50–64 years.
rate_age_4 Influenza Rate per 100k Weekly rate for ages 65 years and older.
rate_age_5 Influenza Rate per 100k Weekly rate for ages 65–74 years.
rate_age_6 Influenza Rate per 100k Weekly rate for ages 75–84 years.
rate_age_7 Influenza Rate per 100k Weekly rate for ages 85 years and older.
rate_age_0tlt1 Influenza Rate per 100k Weekly rate for infants under 1 year (<1 yr).
rate_age_1t4 Influenza Rate per 100k Weekly rate for ages 1–4 years.
rate_age_5t11 Influenza Rate per 100k Weekly rate for ages 5–11 years.
rate_age_12t17 Influenza Rate per 100k Weekly rate for ages 12–17 years.
rate_age_18t29 Influenza Rate per 100k Weekly rate for ages 18–29 years.
rate_age_30t39 Influenza Rate per 100k Weekly rate for ages 30–39 years.
rate_age_40t49 Influenza Rate per 100k Weekly rate for ages 40–49 years.
rate_age_gte75 Influenza Rate per 100k Weekly rate for ages 75 years and older.
rate_age_lt18 Influenza Rate per 100k Weekly rate for pediatric populations (<18 years).
rate_age_gte18 Influenza Rate per 100k Weekly rate for adult populations (18 years and older).

Race, Ethnicity, and Sex

Indicator Name Pathogen or Disease Metric Type Description
rate_race_white Influenza Rate per 100k Weekly rate for White individuals.
rate_race_black Influenza Rate per 100k Weekly rate for Black or African American individuals.
rate_race_hisp Influenza Rate per 100k Weekly rate for Hispanic or Latino individuals.
rate_race_asian Influenza Rate per 100k Weekly rate for Asian or Pacific Islander individuals.
rate_race_natamer Influenza Rate per 100k Weekly rate for American Indian or Alaska Native individuals.
rate_sex_male Influenza Rate per 100k Weekly rate for male individuals.
rate_sex_female Influenza Rate per 100k Weekly rate for female individuals.

Estimation

Metric Definition

Weekly hospitalization rates per 100,000 population are computed upstream by the CDC for each catchment area \(i\), surveillance week \(t\), and demographic stratum \(k\):

\[\text{Rate}_{it}^k = 100{,}000 \cdot \frac{Y_{it}^k}{N_{it}^k}\]

where \(Y_{it}^k\) is the count of residents hospitalized with laboratory-confirmed influenza, and \(N_{it}^k\) is the census population estimate for that stratum in the participating catchment counties.

Delphi ingests these rates directly and publishes each category as an individual signal.

Smoothing

All signals are unsmoothed weekly rate estimates.

Uncertainty

CDC releases do not report standard errors, confidence intervals, sample sizes, or underlying raw hospitalization counts.

Temporal Handling

Observations cover 7-day epidemiological weeks defined by the CDC’s Morbidity and Mortality Weekly Report (MMWR). An MMWR week runs from Sunday through Saturday, standardizing public health reporting across calendar years.

  • reference_time is the Saturday week-ending date of the surveillance week.
  • report_time is the release date when Delphi ingested the weekly CDC snapshot. Revisions update historical rates as hospitals complete retrospective case ascertainment.

Geographic Handling

All four geographic levels (nation, state, msa, and misc) are computed upstream by the CDC. Since Delphi ingests these rates directly from the CDC without spatial interpolation, regional aggregation, or imputation, fill_method is always set to source.

FluSurv-NET monitors hospitalized cases across designated sentinel hospital catchments rather than whole state or national populations.

Delphi maps CDC catchments to standard geographic levels where applicable:

Geographic Level (geo_type) Location (geo_value) Description
nation us Whole-network aggregate across all participating FluSurv-NET sites
state Two-letter postal codes Participating sentinel states
msa 10580, 40380 Albany (10580) and Rochester (40380) metropolitan statistical areas in New York
misc network_eip, network_ihsp CDC partial-network rollups for Emerging Infections Program and IHSP sites

Rates reflect designated hospital catchments within each area rather than full statewide or nationwide populations. Delphi serves these upstream figures directly without regional rollups or spatial imputation.

Relationship to V3

This source reproduces the legacy V3 flusurv endpoint (https://api.delphi.cmu.edu/epidata/flusurv/).

In V3, queries returned wide records containing all demographic rate fields in a single response object. In V5, each stratum is queried as an individual signal conforming to the standard Delphi schema.

V3 Field V5 Signal Notes
rate_overall rate_overall Exact match
rate_age_0 rate_age_0 Exact match
rate_age_1 rate_age_1 Exact match
rate_age_2 rate_age_2 Exact match
rate_age_3 rate_age_3 Exact match
rate_age_4 rate_age_4 Exact match
rate_age_5 rate_age_5 Exact match
rate_age_6 rate_age_6 Exact match
rate_age_7 rate_age_7 Exact match
rate_age_0tlt1 rate_age_0tlt1 Exact match
rate_age_1t4 rate_age_1t4 Exact match
rate_age_5t11 rate_age_5t11 Exact match
rate_age_12t17 rate_age_12t17 Exact match
rate_age_18t29 rate_age_18t29 Exact match
rate_age_30t39 rate_age_30t39 Exact match
rate_age_40t49 rate_age_40t49 Exact match
rate_age_gte75 rate_age_gte75 Exact match
rate_age_lt18 rate_age_lt18 Exact match
rate_age_gte18 rate_age_gte18 Exact match
rate_race_white rate_race_white Exact match
rate_race_black rate_race_black Exact match
rate_race_hisp rate_race_hisp Exact match
rate_race_asian rate_race_asian Exact match
rate_race_natamer rate_race_natamer Exact match
rate_sex_male rate_sex_male Exact match
rate_sex_female rate_sex_female Exact match
rate_flu_a rate_flu_a Exact match
rate_flu_b rate_flu_b Exact match

What changed in V5:

  • Signal Standardization. Demographic rate categories are queried as individual signals using the standard signal parameter rather than separate columns in a single record.
  • Geographic Distinction. Sentinel states use geo_type = 'state', the two New York sub-state sites map to standard MSAs (10580, 40380), the full-network rollup maps to nation:us, and partial-network aggregates map to geo_type = 'misc'.
  • Revision History. Past releases and revisions are queried using snapshot_date or the /archive/ endpoint instead of the legacy issues and lag parameters.

Schema

Columns

Column Key Type Data Type Description
signal Primary Key string The name of the indicator (e.g., rate_overall, rate_age_0).
report_time Primary Key date The CDC update date on which Delphi fetched the data (YYYY-MM-DD).
geo_type Primary Key string Geographic category (nation, state, msa, misc).
geo_value Primary Key string Location identifier (us, two-letter state postal code, 5-digit CBSA code, or network identifier).
fill_method Primary Key string Treatment of missing data or aggregation. Always source because rates are ingested directly without imputation.
reference_time Primary Key date Saturday week-ending date of the surveillance week (YYYY-MM-DD).
value Value Column float Laboratory-confirmed influenza hospitalization rate per 100,000 population.

Missingness & Privacy

  • No cell suppression or differential privacy masking is applied because data is published upstream exclusively as aggregate population rates.
  • Rates are published only for surveillance weeks and strata where active surveillance occurred. Periods outside active surveillance are omitted rather than reported as zero.

Limitations

  • FluSurv-NET comprises more than 90 counties across 14 states, covering approximately 10 percent of the U.S. population (34 million people). Rates reflect designated catchment counties in participating states and do not represent entire statewide or nationwide populations, although the covered population is demographically similar to the national population.
  • Surveillance was established for children younger than 18 during the 2003–04 season and expanded to include adults during the 2005–06 season. Earlier seasons do not include adult data.
  • Network composition changes over time as states and hospitals join or exit surveillance. For instance, North Carolina joined during the 2023–24 season, and Ohio ceased participation after 2024–25.
  • FluSurv-NET tracks patients hospitalized with laboratory-confirmed influenza. Variations in clinical testing practices across facilities and seasons can influence case detection.

Lag & Backfill

  • Updated data is published by the CDC on Fridays.
  • Data is preliminary and subject to backfill as hospitals complete retrospective case ascertainment and medical record reviews. Historical revisions can be accessed via the /archive/ endpoint or by querying with snapshot_date.

Source and Licensing

FluSurv-NET is a collaborative surveillance network administered by the CDC. Data is in the public domain. When utilizing this dataset, citation of FluSurv-NET is recommended:

Chaves, S. S., Lynfield, R., Lindegren, M. L., Bresee, J., & Finelli, L. (2015). The US Influenza Hospitalization Surveillance Network. Emerging Infectious Diseases, 21(9), 1543–1550.