FluView ILINet Outpatient Illness

Attribute Details
Source Name fluview_ilinet
Data Source U.S. Outpatient Influenza-like Illness Surveillance Network (ILINet) via CDC FluView
Geographic Levels state, census_division, hhs, nation
Temporal Granularity Week, ending Saturday
Reporting Cadence Weekly (Fridays)
Temporal Scope Start 1997-10-04 (1997w40); state-level data begins 2010-10-09 (2010w40)
Temporal Scope End Ongoing
Extra Key Columns age_group
License Public Domain US Government

This source reproduces the legacy V3 fluview endpoint. Age breakdowns are now queried using the age_group column, and signal names follow standard V5 conventions. See Relationship to V3.

Table of contents

  1. Overview
  2. Indicators (Signals)
  3. Estimation
    1. Metric Definition
    2. Smoothing
    3. Uncertainty
    4. Temporal Handling
    5. Geographic Handling
  4. Relationship to V3
  5. Schema
    1. Columns
    2. Extra Keys
  6. Missingness & Privacy
  7. Limitations
  8. Lag & Backfill
  9. Source and Licensing

Overview

The U.S. Outpatient Influenza-like Illness Surveillance Network (ILINet), administered by the CDC, monitors outpatient visits for influenza-like illness across the United States. Approximately 3,000 healthcare providers (sentinel outpatient clinics, emergency departments, and family medicine practices) voluntarily report weekly summaries of total patient encounters and the number of patients presenting with symptoms meeting the standard ILI case definition.

The CDC defines influenza-like illness as a fever of 100°F (37.8°C) or higher accompanied by a cough or sore throat, without a known cause other than influenza. Delphi fetches ILINet data weekly from the CDC FluView Interactive portal. This dataset provides unweighted and population-weighted ILI visit percentages, ILI encounter counts by age group, and healthcare provider participation numbers.

For companion laboratory surveillance data, see FluView Clinical Labs and FluView Public Health Labs.

Indicators (Signals)

Indicator Name Pathogen or Disease Metric Type Description
wili Influenza-like Illness Percentage Population-weighted percentage of outpatient visits with ILI. Available for nation, hhs, and census_division; suppressed at state.
ili Influenza-like Illness Percentage Unweighted percentage of outpatient visits with ILI ($100 \cdot \frac{\text{num_ili}}{\text{num_patients}}$).
num_ili Influenza-like Illness Count Weekly count of outpatient visits presenting with ILI. Stratified by age_group.
num_providers — Count Number of healthcare providers reporting to ILINet during the week.
num_patients — Count Total outpatient visits across all conditions recorded by reporting providers.

Estimation

Metric Definition

Let \(Y_{it}\) be the count of ILI encounters (ilitotal) in location \(i\) during week \(t\), and let \(N_{it}\) be the total patient visits (total_patients).

  • Unweighted ILI (ili) is the percentage of outpatient visits meeting the ILI case definition:

    \[\text{ILI}_{it} = 100 \cdot \frac{Y_{it}}{N_{it}}\]

  • Weighted ILI (wili) is a population-weighted average of state-level percentages calculated by the CDC for multi-state regions:

    \[\text{wILI}_{rt} = \sum_{s \in r} \left( \text{ILI}_{st} \cdot \frac{P_s}{P_r} \right)\]

    where \(P_s\) is the population of state \(s\) and \(P_r\) is the total population of region \(r\). State-level population weights are not applicable, so the CDC does not compute state-level wili. Delphi excludes wili at the state level.

  • Counts (num_ili, num_patients, num_providers) pass through directly from the CDC. num_ili is also available by age stratum.

Smoothing

All signals are unsmoothed discrete 7-day weekly totals or ratios.

Uncertainty

Uncertainty intervals and standard errors are not published by the CDC for ILINet data. The dataset includes num_patients and num_providers as direct measures of reporting volume and network coverage.

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 past weeks as delayed provider reports arrive.

Geographic Handling

All four geographic levels (nation, hhs, census_division, state) are computed natively by the CDC and ingested directly without regional aggregation (fill_method = 'source').

At the state level, the CDC reports New York as two separate jurisdictions: New York City (nyc) and New York State excluding NYC (ny_minus_nyc). When both are reported for a given week, Delphi reconstructs a combined statewide ny record (fill_method = 'nyc_plus_ny_minus_nyc').

Raw counts are summed directly:

\[Y_{\text{NY}, t} = Y_{\text{NYC}, t} + Y_{\text{NY-NYC}, t}\]

The unweighted percentage (ili) is recomputed as a pooled ratio:

\[\text{ILI}_{\text{NY}, t} = 100 \cdot \frac{Y_{\text{NYC}, t} + Y_{\text{NY-NYC}, t}}{N_{\text{NYC}, t} + N_{\text{NY-NYC}, t}}\]

Weighted ILI (wili) is omitted at the state level because the CDC only calculates population-weighted averages across multi-state regions.

Relationship to V3

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

In V3, age-stratified counts were flattened into separate columns (num_age_0 through num_age_5). In V5, age stratification is organized under the age_group key column of num_ili.

V3 Field V5 Signal Extra Key (age_group) Notes
wili wili all State-level wili excluded in V5
ili ili all Exact match
num_ili num_ili all Exact match for total ILI encounters
num_providers num_providers all Exact match
num_patients num_patients all Exact match
num_age_0 num_ili 0-4 Mapped to 0-4 age stratum
num_age_1 num_ili 5-24 Mapped to 5-24 age stratum
num_age_2 num_ili 25-64 Historical overlapping age bucket
num_age_3 num_ili 25-49 Mapped to 25-49 age stratum
num_age_4 num_ili 50-64 Mapped to 50-64 age stratum
num_age_5 num_ili 65+ Mapped to 65+ age stratum

What changed in V5:

  • Age Stratification via Keys. Rather than maintaining separate column names, all age groups are unified under the num_ili indicator and categorized using extra_keys.
  • Exclusion of State-Level wili. The legacy V3 endpoint substituted unweighted ili into wili when weighted ILI was missing at the state level. V5 removes this substitution: wili is published only where the CDC computes it (national, HHS, and census divisions).
  • Discontinuation of Pre-2010 Imputation. The legacy V3 endpoint provided sensor-fusion regression estimates for state-level data prior to 2010w40. V5 serves direct CDC observations without synthetic historical imputation.

Schema

Columns

Column Key Type Data Type Description
signal Primary Key string The name of the indicator (wili, ili, num_ili, num_providers, num_patients).
report_time Primary Key date The date Delphi ingested the CDC weekly file (YYYY-MM-DD).
geo_type Primary Key string Geographic level (state, census_division, hhs, nation).
geo_value Primary Key string Location code (e.g., us, 1–10, 1–9, pa, ny, nyc, ny_minus_nyc).
age_group Primary Key string Age group category (see Extra Keys below).
fill_method Primary Key string source for directly ingested rows; nyc_plus_ny_minus_nyc for reconstructed statewide NY.
reference_time Primary Key date Saturday week-ending date of the surveillance week (YYYY-MM-DD).
value Value Column float Percentage or patient/provider count.

Extra Keys

The age_group key applies to num_ili and accepts the following values:

  • all: Total ILI encounters across all ages (matches num_ili total).
  • 0-4: Patients aged 0 to 4 years.
  • 5-24: Patients aged 5 to 24 years.
  • 25-49: Patients aged 25 to 49 years.
  • 50-64: Patients aged 50 to 64 years.
  • 65+: Patients aged 65 years and older.
  • 25-64: Coarser historical age band used prior to the 25–49 and 50–64 split.

For all other signals (wili, ili, num_providers, num_patients), age_group is always set to all.

Missingness & Privacy

  • Weighted ILI (wili) is omitted at the state level because the CDC only calculates population-weighted averages for multi-state regions.
  • ILINet is a voluntary reporting network. If no providers report within a jurisdiction during a week, that observation is omitted from the dataset rather than recorded as zero.

Limitations

  • ILI measures symptoms (fever with cough or sore throat) rather than laboratory-confirmed influenza. Other circulating respiratory pathogens, including COVID-19 and RSV, contribute to ILI counts.
  • Provider participation varies across states and clinical specialties. Jurisdictions with more pediatric practices often report higher baseline ILI rates.
  • Shifts in healthcare-seeking behavior, insurance coverage, and telehealth usage can influence visit volumes independently of disease spread.

Lag & Backfill

  • Initial weekly data is published by the CDC on Fridays, six days after the surveillance week ends.
  • Backfill is common during the first two to four weeks as participating clinics submit delayed reports. Revisions can be accessed via the /archive/ endpoint or by querying with snapshot_date.

Source and Licensing

ILINet data is published by the CDC Outpatient Influenza-like Illness Surveillance Network. The data is in the public domain.