Datasets:

EPI-Eval
/

nhsn-hrd

pretty_name: CDC NHSN Hospital Respiratory Data (HRD)
license: other
size_categories:
  - 10K<n<100K
tags:
  - cadence-weekly
  - geo-us
  - surveillance-respiratory
  - pathogen-influenza
  - pathogen-sars-cov-2
  - pathogen-rsv
  - tier-1
  - availability-open
schema_version: '0.1'
source_id: nhsn-hrd
source_url: >-
  https://data.cdc.gov/Public-Health-Surveillance/Weekly-Hospital-Respiratory-Data-HRD-Metrics-by-Ju/ua7e-t2fy/about_data
manifest_section: §1.1
surveillance_category: respiratory
pathogens:
  - influenza
  - sars-cov-2
  - rsv
availability: open
availability_notes: null
access_type: socrata
tier: 1
cadence: weekly
geography_levels:
  - national
  - subnational-region
  - subnational-state
geography_countries:
  - US
gold_standard_for:
  - flusight-forecast-hub
  - covid19-forecast-hub
  - rsv-forecast-hub
vintaged_version_of: null
succeeds: hhs-protect
derived_from: []
value_columns:
  - name: totalconfflunewadm
    unit: admissions/week
    value_type: incident
    description: New laboratory-confirmed influenza hospital admissions (FluSight target)
    aggregation: sum
  - name: totalconfc19newadm
    unit: admissions/week
    value_type: incident
    description: New laboratory-confirmed COVID-19 hospital admissions
    aggregation: sum
  - name: totalconfrsvnewadm
    unit: admissions/week
    value_type: incident
    description: >-
      New laboratory-confirmed RSV hospital admissions (mandatory reporting from
      epiweek 2024-46)
    aggregation: sum
  - name: numinptbeds
    unit: beds
    value_type: stock
    description: Total staffed inpatient beds (capacity denominator) — point-in-time
    aggregation: mean
  - name: numinptbedsocc
    unit: beds
    value_type: stock
    description: Staffed inpatient beds occupied — point-in-time
    aggregation: mean
  - name: numicubeds
    unit: beds
    value_type: stock
    description: Total ICU beds — point-in-time
    aggregation: mean
  - name: numicubedsocc
    unit: beds
    value_type: stock
    description: ICU beds occupied — point-in-time
    aggregation: mean
  - name: totalconfc19icupats
    unit: patients
    value_type: stock
    description: Confirmed COVID-19 patients in ICU — point-in-time
    aggregation: mean
  - name: totalconffluicupats
    unit: patients
    value_type: stock
    description: Confirmed influenza patients in ICU — point-in-time
    aggregation: mean
  - name: totalconfflunewadmper100k
    unit: admissions/week per 100k population
    value_type: rate
    description: Weekly flu admission rate per 100k population
    aggregation: rate
  - name: totalconfc19newadmper100k
    unit: admissions/week per 100k population
    value_type: rate
    description: Weekly COVID-19 admission rate per 100k population
    aggregation: rate
  - name: totalconfflunewadmcumulativeseasonalsum
    unit: admissions
    value_type: cumulative
    description: Cumulative flu admissions since start of respiratory season
    aggregation: max
  - name: totalconfc19newadmcumulativeseasonalsum
    unit: admissions
    value_type: cumulative
    description: Cumulative COVID-19 admissions since start of respiratory season
    aggregation: max
  - name: totalconfrsvnewadmcumulativeseasonalsum
    unit: admissions
    value_type: cumulative
    description: Cumulative RSV admissions since start of respiratory season
    aggregation: max
notes:
  extra_columns:
    - column: location_name
      description: >-
        2-letter postal abbreviation as reported by the source. Kept for
        traceability; `location_id` is canonical.
    - column: respseason
      description: >-
        Respiratory season identifier (e.g. '2024-2025'). NHSN-specific; useful
        for grouping by season but doesn't appear in other sources.
  interpretation_caveats:
    - column: numinptbeds
      caveat: >-
        Counts only *staffed* beds (those with available staff to operate). HHS
        Protect's analogous capacity column had different inclusion criteria; do
        not naively splice across the 2024-11 boundary.
    - column: totalconfflunewadm
      caveat: >-
        Counts new admissions where flu was confirmed by lab on or within 14
        days of admission. Reporting jurisdictions varied this rule pre-2024-11;
        treat pre-2024-11 values as approximate.
    - column: totalconfrsvnewadm
      caveat: >-
        RSV mandatory reporting started epiweek 2024-46; pre-2024-46 values are
        NaN or sparsely populated by voluntary reporters.
    - column: '*cumulativeseasonalsum'
      caveat: >-
        Cumulative within a respiratory season (resets at season boundary). Use
        with `value_type: cumulative` discipline — never sum, diff to recover
        incident.
  general: >
    NHSN HRD is the FluSight scoring target from the 2024–25 season onward. The
    feed includes

    backfilled HHS Protect data going back to 2020-08, but the underlying
    reporting program

    changed in 2024-11 (voluntary → mandatory). For long-term modeling that
    crosses that boundary,

    treat pre-2024-11 and post-2024-11 as having different observational
    regimes.


    Hospital reporting became voluntary 2024-05-01 to 2024-11-09; mandatory
    reporting resumed

    under NHSN HRD starting epiweek 2024-46. The voluntary period has reduced
    coverage.
extra:
  case_definition_url: https://www.cdc.gov/nhsn/psc/respiratory-data.html
  source_columns_omitted: >-
    199 of 213 source columns omitted (mostly adult/ped breakdowns,
    percentage-change variants, and per-hospital-reporter normalizations).
    Available on the source endpoint if needed.
  flusight_target_column: totalconfflunewadm
computed:
  last_ingested: '2026-04-25T22:38:34Z'
  row_count: 19966
  time_coverage:
    - start: '2020-08-08'
      end: '2026-04-18'
  geography_unit_count: 67
  observed_cadence_days: 7
  missing_gaps: []
  data_hash: 5babe542cebfb216

CDC NHSN Hospital Respiratory Data (HRD)

NHSN HRD is the FluSight scoring target from the 2024–25 season onward. The feed includes backfilled HHS Protect data going back to 2020-08, but the underlying reporting program changed in 2024-11 (voluntary → mandatory). For long-term modeling that crosses that boundary, treat pre-2024-11 and post-2024-11 as having different observational regimes.

Hospital reporting became voluntary 2024-05-01 to 2024-11-09; mandatory reporting resumed under NHSN HRD starting epiweek 2024-46. The voluntary period has reduced coverage.

Source: https://data.cdc.gov/Public-Health-Surveillance/Weekly-Hospital-Respiratory-Data-HRD-Metrics-by-Ju/ua7e-t2fy/about_data

Coverage

Time: 2020-08-08 → 2026-04-18
Cadence: weekly (observed median spacing: 7 days)
Geography levels: national, subnational-region, subnational-state — 67 unique location IDs
Countries: US
Pathogens: influenza, sars-cov-2, rsv
Surveillance category: respiratory
Rows: 19,966

Columns

Column	Unit	value_type	Aggregation	Description
`totalconfflunewadm`	admissions/week	`incident`	`sum`	New laboratory-confirmed influenza hospital admissions (FluSight target)
`totalconfc19newadm`	admissions/week	`incident`	`sum`	New laboratory-confirmed COVID-19 hospital admissions
`totalconfrsvnewadm`	admissions/week	`incident`	`sum`	New laboratory-confirmed RSV hospital admissions (mandatory reporting from epiweek 2024-46)
`numinptbeds`	beds	`stock`	`mean`	Total staffed inpatient beds (capacity denominator) — point-in-time
`numinptbedsocc`	beds	`stock`	`mean`	Staffed inpatient beds occupied — point-in-time
`numicubeds`	beds	`stock`	`mean`	Total ICU beds — point-in-time
`numicubedsocc`	beds	`stock`	`mean`	ICU beds occupied — point-in-time
`totalconfc19icupats`	patients	`stock`	`mean`	Confirmed COVID-19 patients in ICU — point-in-time
`totalconffluicupats`	patients	`stock`	`mean`	Confirmed influenza patients in ICU — point-in-time
`totalconfflunewadmper100k`	admissions/week per 100k population	`rate`	`rate`	Weekly flu admission rate per 100k population
`totalconfc19newadmper100k`	admissions/week per 100k population	`rate`	`rate`	Weekly COVID-19 admission rate per 100k population
`totalconfflunewadmcumulativeseasonalsum`	admissions	`cumulative`	`max`	Cumulative flu admissions since start of respiratory season
`totalconfc19newadmcumulativeseasonalsum`	admissions	`cumulative`	`max`	Cumulative COVID-19 admissions since start of respiratory season
`totalconfrsvnewadmcumulativeseasonalsum`	admissions	`cumulative`	`max`	Cumulative RSV admissions since start of respiratory season

Additional data columns

location_name — 2-letter postal abbreviation as reported by the source. Kept for traceability; location_id is canonical.
respseason — Respiratory season identifier (e.g. '2024-2025'). NHSN-specific; useful for grouping by season but doesn't appear in other sources.

Interpretation caveats

Things that may differ from how other sources define a similar measure. If you're combining this dataset with another, read these first.

numinptbeds — Counts only staffed beds (those with available staff to operate). HHS Protect's analogous capacity column had different inclusion criteria; do not naively splice across the 2024-11 boundary.
totalconfflunewadm — Counts new admissions where flu was confirmed by lab on or within 14 days of admission. Reporting jurisdictions varied this rule pre-2024-11; treat pre-2024-11 values as approximate.
totalconfrsvnewadm — RSV mandatory reporting started epiweek 2024-46; pre-2024-46 values are NaN or sparsely populated by voluntary reporters.
*cumulativeseasonalsum — Cumulative within a respiratory season (resets at season boundary). Use with value_type: cumulative discipline — never sum, diff to recover incident.

Related datasets in EPI-Eval

Gold standard for: flusight-forecast-hub, covid19-forecast-hub, rsv-forecast-hub
Succeeds: hhs-protect (different reporting program; do not splice rows directly without normalization)

Access

Availability: open
Access type: socrata
License: other
Tier: 1

Schema version 0.1 · Last ingested 2026-04-25T22:38:34Z · source_id: nhsn-hrd · Manifest section §1.1