PopHive Claims (Epic Cosmos)

Attribute Details
Source Name pophive
Data Source Epic Cosmos via PopHIVE
Geographic Levels nation, state, hhs
Temporal Granularity Weekly, week ending Saturday
Reporting Cadence Irregular (biweekly to monthly updates)
Temporal Scope Start 2018-01-07
Temporal Scope End Ongoing
Extra Key Columns age_group
License PopHIVE Attribution / CC BY 4.0

Table of contents

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

Overview

Epic Cosmos is a collaborative research platform containing de-identified patient data from over 300 million patients across more than 1,600 hospitals and health systems using Epic electronic health record systems. Data is accessed via SlicerDicer, a self-service analytics tool. The dataset includes emergency department encounters, diagnoses, immunizations, laboratory results, and other clinical data.

Delphi ingests aggregated emergency department encounter data curated by the PopHIVE platform.


Indicators (Signals)

Indicator Name Pathogen or Disease Metric Type Description
all_n_encounters_ed All Conditions Count Total emergency department encounters over the reference week.
covid_n_ed COVID-19 Count Emergency department encounters with a COVID-19 diagnosis.
covid_pct_ed COVID-19 Percentage Percentage of emergency department encounters with a COVID-19 diagnosis.
flu_n_ed Influenza Count Emergency department encounters with an influenza diagnosis.
flu_pct_ed Influenza Percentage Percentage of emergency department encounters with an influenza diagnosis.
rsv_n_ed RSV Count Emergency department encounters with an RSV diagnosis.
rsv_pct_ed RSV Percentage Percentage of emergency department encounters with an RSV diagnosis.

Estimation

Metric Definition

Count signals report the total number of emergency department encounters observed during the reference week.

Percentage signals report the share of those encounters diagnosed with the specified condition for the same location, age group, and week.

Temporal Handling

Dates in this dataset represent surveillance epiweeks, where each reference date (reference_time) corresponds to a Saturday week-ending date.

Upstream updates are expected biweekly, but the release schedule is irregular. Intervals between releases have ranged from weekly to monthly, with occasional publication delays exceeding four weeks.

Geographic Handling

PopHIVE natively reports national (nation) and state (state) data directly. These levels use the source fill method. Upstream national values are computed by PopHIVE from unsuppressed encounter records.

Delphi calculates HHS regional values (hhs) by aggregating state-level data using population weights. When computing HHS regional values, the fill_method column distinguishes the imputation method applied to missing or suppressed state values:

  • source: Direct upstream reporting without imputation (available at nation and state levels).
  • zero (or fill_zero): Treats missing or suppressed state values as zero during aggregation.
  • ave (or fill_ave): Fills missing or suppressed state values with the population-weighted average of reporting states in the HHS region before aggregating.

Schema

Columns

Column Key Type Data Type Description
signal Primary Key string Signal identifier.
report_time Primary Key date Publication or release date (YYYY-MM-DD).
geo_type Primary Key string Geographic level (nation, state, hhs).
geo_value Primary Key string Geographic entity code.
fill_method Primary Key string Imputation method (source, zero, ave).
age_group Primary Key (Extra Key) string Age group category.
reference_time Primary Key date Reference week ending Saturday (YYYY-MM-DD).
value Value Column float Measured encounter count or percentage.

Extra keys

The age_group column stratifies data into population age categories:

Value Description
all All ages combined
<1 Infants under 1 year
1-4 Children aged 1 to 4 years
5-17 Children and adolescents aged 5 to 17 years
18-49 Adults aged 18 to 49 years
50-64 Adults aged 50 to 64 years
65+ Adults 65 years and older

An unfiltered query returns rows across all age groups. Queries can filter to a specific group using the extra_keys parameter (for example, extra_keys=age_group:18-49).


Missingness & Privacy

Counts under 10 are suppressed upstream by PopHIVE for patient privacy and replaced with a count of 5. These suppressed cells appear as 5 in the count signals. Percentage signals derived from these counts reflect this imputed value.

Unobserved weeks or locations without reporting health systems are absent from the dataset rather than listed as null.


Limitations

Epic Cosmos reflects data only from healthcare systems that use Epic electronic health records and participate in Cosmos. Data coverage varies geographically according to Epic market adoption. Regions with lower Epic adoption may not be representative of the broader population.


Lag & Backfill

PopHIVE releases data updates on an irregular cadence, typically every two to four weeks. Each release publishes a full revision of historical data, which updates prior reference dates.


Source and Licensing

This data source is provided by the PopHIVE platform and is based on de-identified data from Epic Cosmos.

The data can be re-used with appropriate attribution under CC BY 4.0. A suggested citation relating to this data is:

Results of research performed with Epic Cosmos were obtained from the PopHIVE platform (https://github.com/PopHIVE/Ingest).