PopHIVE Data Source Documentation

Abcs

CDC monitors invasive bacterial infections that cause bloodstream infections, sepsis, and meningitis in persons living in the community through Active Bacterial Core surveillance (ABCs). ABCs conducts laboratory- and population-based surveillance for invasive pneumococcal disease (IPD). ABCs serotype data are used to measure the impact of vaccine use in the United States on vaccine-type IPD. This table reports IPD case counts in the ABCs catchment area by serotype for years 1998 through 2022. Cases are grouped into the following mutually exclusive age groups: age <2 years old, age 2-4 years old, age 5-17 years old, age 18-49 years old, age 50-64 years old, and age >=65 years old. ABCs methods and surveillance areas reporting IPD cases has changed over time. Given these changes, trends in serotype distribution by year and age group should be interpreted with caution. The all-site summary presented here is calculated based on the 8 sites that consistently report to ABCs and differs from the All-site measure provided by the source. Additional information on ABCs methods and surveillance population is available at https://www.cdc.gov/abcs/methodology/index.html. Analyze and visualize data using the ABCs Bact Facts Interactive Data Dashboard at https://www.cdc.gov/abcs/bact-facts-interactive-dashboard. ABCs IPD Isolates were serotyped by Quellung, PCR, or whole genome sequencing (WGS). Cases without an isolate available or with mixed serotypes reported are listed on the table as MISS. Additionally, non-typeable IPD cases are shown as NT. Zero cell rows were not included in this dataset. Minor changes to previous years serotype data can occur as additional isolates and serotype data become available. Cases were excluded from this dataset if the ABCs site did not perform surveillance in the catchment area for a full calendar year. As a result, cases were excluded from the following sites: TN, 11 counties, Jul-Dec 1999; CO, 5 counties, Jul-Dec 2000; CA, 2 counties (aged <5 years), Oct-Dec 2000.

Sources

• Data Source | Centers for Disease Control and Prevention | API/Data Location
• Data Source | University of Louisville / Pfizer | API/Data Location

Variables

data.csv.gz

uad.csv.gz

Atlas AMR

No standard data files found.

BRFSS

The Behavioral Risk Factor Surveillance System (BRFSS) is the nation's premier system of health-related telephone surveys that collect state data about U.S. residents regarding their health-related risk behaviors, chronic health conditions, and use of preventive services. Established in 1984 with 15 states, BRFSS now collects data in all 50 states, the District of Columbia, and three U.S. territories, completing more than 400,000 adult interviews each year. BRFSS provides state-specific data on health conditions including obesity, diabetes, depression, and health behaviors such as heavy drinking, physical activity, and tobacco use. Data are available by age, sex, race/ethnicity, and education level. BRFSS is a critical resource for public health surveillance and policy-making at both state and national levels.

Sources

• Data Source | Centers for Disease Control and Prevention | API/Data Location

Variables

data_survey.csv.gz

data.csv.gz

CDC Cfa Rt

Weekly estimates of the effective reproduction number (Rt) for COVID-19, Influenza, and RSV at the US national and state level. Rt represents the average number of secondary infections generated by one infectious person and is a key indicator of epidemic trajectory. Values above 1 indicate growth; below 1 indicate decline. Data include the median Rt estimate, 95% credible intervals, and the probability that Rt exceeds 1.0. Through May 2026, estimates were produced using EpiNow2 with independent state-level models; from June 2026 onward, estimates use hierarchical generalized additive models (HGAM) with spatial partial pooling across states. Coverage begins September 2024.

Sources

• Data Source | CDC Center for Forecasting and Outbreak Analytics (CFA)

Variables

data.csv.gz

Census

The 2020 Census Urban Area to County Allocation File maps Census-defined urban areas (urbanized areas with population ≥ 50,000 and urban clusters with population 2,500–49,999) to counties. Each row represents a county-urban area intersection and includes the share of county population, land area, and housing units within that urban area. Summing across all urban areas for a county yields total percent urban. Counties not appearing in the file have no urban area and are considered 100% rural. Urban area boundaries are based on the 2020 decennial Census and reflect the updated delineation methodology that increased the minimum population threshold for urbanized areas from 50,000 to 50,000 and revised cluster definitions.

Sources

• Data Source | U.S. Census Bureau | API/Data Location
• Data Source | U.S. Census Bureau | API/Data Location

Variables

data_county.csv.gz

data_state.csv.gz

data_zcta_2019_2020.csv.gz

data_zcta_2021_2022.csv.gz

data_zcta_2023_2024.csv.gz

CMS Mmd

The Mapping Medicare Disparities (MMD) by Population Tool is an interactive map that displays chronic disease prevalence, costs, hospitalization, and preventive care utilization data for Medicare Fee-for-Service beneficiaries. Data are available at the national, state, and county levels, stratified by age, race/ethnicity, and sex. Condition prevalence rates are calculated using ICD-10 diagnosis codes in Medicare claims data, following the Chronic Conditions Warehouse (CCW) definitions. The tool covers over 30 chronic conditions including diabetes, hypertension, COPD, heart failure, and mental health conditions, as well as preventive service utilization metrics. Data are updated annually.

Sources

• Data Source | Centers for Medicare and Medicaid Services (CMS)

Variables

data_state_county_age_by_race.csv.gz

data_state_county_age_by_sex.csv.gz

data_state_county_age.csv.gz

County Health Rankings

The County Health Rankings & Roadmaps (CHR&R) program, a collaboration between the Robert Wood Johnson Foundation and the University of Wisconsin Population Health Institute, ranks the health of nearly all counties in the nation. The annual rankings measure vital health factors including high school graduation rates, obesity, smoking, unemployment, access to healthy foods, the quality of air and water, income inequality, and teen births that influence how long and how well we live. Data are released annually and cover all U.S. counties. Raw data are publicly available through the Zenodo repository (https://doi.org/10.5281/zenodo.18157681) and span release years 2010 through the present.

Sources

• Data Source | University of Wisconsin Population Health Institute | API/Data Location

Variables

data_county.csv.gz

data_state.csv.gz

Delphi Doctors Claims

The Delphi Doctor Visits signal estimates the percentage of outpatient doctor visits with COVID-related diagnoses based on claims data from health system partners. CMU Delphi receives de-identified medical insurance claims data covering a significant fraction of United States healthcare visits. The signal is calculated as the percentage of outpatient visits with COVID-related ICD-10 diagnosis codes (U071, U072, B9729, J1281, Z03818, B342, J1289). Data are smoothed using a Gaussian linear smoother to reduce day-to-day noise. This signal provides near-real-time insight into COVID-19 activity at the community level based on actual healthcare encounters. Data are available at state and county levels with approximately 3-4 day lag from the date of service.

Sources

• Data Source | Carnegie Mellon University Delphi Research Group | API/Data Location

Variables

data.csv.gz

Delphi Hospital Claims

The Delphi Hospital Admissions signal estimates the percentage of new hospital admissions with COVID-19 or influenza diagnoses based on electronic medical records and claims data from health system partners. CMU Delphi receives de-identified hospital admission data covering a significant fraction of United States hospitals. The signals track inpatient admissions with relevant ICD-10 diagnosis codes for COVID-19 and influenza. Data are smoothed using a Gaussian linear smoother to reduce day-to-day variation. This signal provides near-real-time insight into severe respiratory illness at the community level based on actual hospitalizations. Data are available at state and county levels with approximately 3-4 day lag from the date of admission.

Sources

• Data Source | Carnegie Mellon University Delphi Research Group | API/Data Location

Variables

data.csv.gz

Delphi ILI Fluview

Influenza-Like Illness (ILI) surveillance data from the CDC's ILINet network, accessed via the CMU Delphi Epidata API. ILINet is a collaborative effort between the CDC, state and local health departments, and approximately 3,000 outpatient healthcare providers across all 50 states, Puerto Rico, the District of Columbia, and the U.S. Virgin Islands. Providers report the total number of patients seen and the number presenting with ILI (defined as fever >=100F plus cough and/or sore throat without a known cause other than influenza). Data are available from 1997 week 40 onward at national, HHS regional, and state levels. Both weighted (population-adjusted) and unweighted percentages are provided. This long-running surveillance system is the primary source for tracking seasonal influenza activity in the United States.

Sources

• Data Source | Carnegie Mellon University Delphi Research Group | API/Data Location

Variables

data.csv.gz

Delphi NHSN

Weekly hospital respiratory data reported to CDC's National Healthcare Safety Network (NHSN), accessed via the CMU Delphi Epidata API. NHSN is the nation's most widely used healthcare-associated infection tracking system, collecting data from hospitals across the United States. The data includes COVID-19, influenza, and RSV associated hospital admissions aggregated to national and state levels. Data collection became mandatory for hospitals in November 2024; prior to this, reporting was voluntary with variable participation rates.

Sources

• Data Source | Carnegie Mellon University Delphi Research Group | API/Data Location

Variables

data.csv.gz

Epic Chronic

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 visits, diagnoses, immunizations, laboratory results, and other clinical data. Due to privacy protections, counts fewer than 10 are suppressed and imputed. Coverage extends across all U.S. states and territories. Note that county-level and city-level stratifications could differ markedly in total sample size due to high levels of missingness of county data in some states.

Sources

• Data Source | Epic Systems

Variables

county_no_time.csv.gz

county_year.csv.gz

state_no_time.csv.gz

state_year.csv.gz

Epic Hepb Vax

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 visits, diagnoses, immunizations, laboratory results, and other clinical data. Due to privacy protections, counts fewer than 10 are suppressed and imputed. Coverage extends across all U.S. states and territories. Note that county-level and city-level stratifications could differ markedly in total sample size due to high levels of missingness of county data in some states.

Sources

• Data Source | Epic Systems

Variables

data.csv.gz

Epic Injury

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 visits, diagnoses, immunizations, laboratory results, and other clinical data. Due to privacy protections, counts fewer than 10 are suppressed and imputed. Coverage extends across all U.S. states and territories. Note that county-level and city-level stratifications could differ markedly in total sample size due to high levels of missingness of county data in some states.

Sources

• Data Source | Epic Systems

Variables

heat_year_county.csv.gz

monthly_injury.csv.gz

yearly_injury.csv.gz

Epic Resp Infections

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 visits, diagnoses, immunizations, laboratory results, and other clinical data. Due to privacy protections, counts fewer than 10 are suppressed and imputed. Coverage extends across all U.S. states and territories. Note that county-level and city-level stratifications could differ markedly in total sample size due to high levels of missingness of county data in some states.

Sources

• Data Source | Epic Systems

Variables

monthly_tests.csv.gz

no_geo.csv.gz

quarterly_gas.csv.gz

weekly.csv.gz

Gtrends

Google Health Trends data accessed via the Google Health Trends API, processed and collected using Yale DISSC's gtrends_collection framework. The data represents the probability of a short search session including a specific health-related term within a geography and timeframe, multiplied by 10 million for readability. Search volumes are provided at the DMA (Designated Market Area) and state level on a weekly basis. This data source enables tracking of public interest in health topics such as RSV, overdose, and naloxone as potential early indicators of disease activity or public health concerns.

Sources

• Data Source | Yale Data-Intensive Social Science Center (DISSC)

Variables

data_dma_year.csv.gz