Mapping Broadband Health in America is a customizable data visualization platform that depicts the intersection of broadband and health data for states and counties in the U.S. The 2023 release of the mapping platform now includes data on maternal mortality and severe maternal morbidity in response to the “Data Mapping to Save Moms’ Lives” Act, in addition to updated chronic disease data and opioid-related mortality and prescription rates.
The mapping platform uses color-coding, filtering, and data overlays to permit users—from the casual observer to the sophisticated researcher—to create customized maps intersecting the broadband and health variables of interest. With these design features, users can explore gaps and opportunities in the broadband health space, the status of broadband or health across levels of the other, as well as the characteristics and context of these areas by selecting relevant demographic, socioeconomic, and access variables. It is also open data and open source so that users can conduct their own analyses and add functionality that addresses their specific needs and priorities. This section describes the methodology underlying the mapping platform, including the variables, data sources, calculations, and design features. Visit www.fcc.gov/health/maps/developers and https://www.fcc.gov/reports-research/maps/connect2health/data.html for more information.
Maternal Health
In Phase 1, the approach to intersecting broadband and maternal health reflects an initial focus on two outcome variables, severe maternal morbidity and maternal mortality up to one year postpartum. As a complement to that outcome data, the Task Force also consulted with the Centers for Disease Control and Prevention and reviewed the available literature to identify relevant risk factors and Social Determinants of Health that influence maternal health outcomes and where broadband-enabled interventions might help bridge the gaps. Information on the resulting conceptual model and the measures included in the Phase 1 effort and planned Phase 2 additions can be found here with a brief summary below.
Data on maternal mortality comes from death certificates submitted to the National Vital Statistics System and is accessed through CDC WONDER—the Wide-ranging Online Data for Epidemiologic Research. The platform incorporates three measures related to maternal mortality. The first is the Maternal Mortality Rate, which is defined as “the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and the site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes.” The second is the Late Maternal Death Rate, which includes deaths that occur after 42 days but up to one year postpartum. Both rates are calculated per 100,000 live births, also using data from CDC WONDER. The third is “Maternal Deaths,” which reflects counties reporting maternal deaths (up to 42 days postpartum) or no maternal deaths. The CDC does not report mortality data when there are between one and nine deaths in a given geographical location for privacy and confidentiality reasons; these counties are marked as “suppressed” in the platform. The CDC also does not report rates when there are fewer than 20 deaths due to statistical unreliability.
Data on severe maternal morbidity come from the Healthcare Cost and Utilization Project’s Fast Stats, which are sponsored by the Agency for Healthcare Research and Quality. They reflect the number of women experiencing unexpected outcomes of labor and delivery (21 indicators) per 10,000 in-hospital deliveries. For the initial effort, data for this measure are available at the state-level for intersection with broadband. This information on maternal mortality and severe maternal morbidity can be intersected with broadband to show areas where broadband can be leveraged through telehealth and other broadband-enabled solutions to improve health outcomes and address disparities.
Additional data on risk factors, maternal health care, and Social Determinants of Health come from various sources, including from the Health Resources and Services Administration, the March of Dimes, and the U.S. Census Bureau. These data are available in the maternal health module of the platform for intersection with broadband or as filters that can be applied to explore disparities in the broadband health picture in specific areas and populations.
We note that, by design, the maternal health module reflects new and additional functionalities that reflect the evolution of thinking on the intersection of broadband and health (i.e., the visualization of double burden counties that have higher health need and limited connectivity resources and other county categorizations); enhancements to the analytics of the mapping platform to better support policy and decision-making (i.e., tables that reflect the number of counties and population in each county group); and innovations to respond to data limitations when visualizing mortality data (i.e., incorporation of data on risk factors for intersection with broadband connectivity). As such, the layout and features of the maternal health module may differ from those for opioids or chronic diseases.
In structuring the new maternal health functionality (and planning for Phase 2), we were also mindful that telehealth and other broadband-enabled solutions and technologies offer new and exciting opportunities to help address the maternal health crisis. Remote monitoring, telesonography, self-operated ultrasound and robotic sonography are all emerging broadband-enabled technologies that show promise in this area. The March of Dimes recently observed that, “In obstetrics, telehealth exists in nearly every aspect of care from remote observation of ultrasound recordings to postpartum blood pressure tracking, however, development of evidence-based practices may lag with technology uptake.” For example, broadband can be leveraged in:
- Providing preconception, prenatal, and postnatal care to reproductive age and pregnant women through telehealth, when appropriate, particularly in maternity care deserts or to women who experience limited access to care due to geographical, financial, physical, cultural/social, or other reasons. Studies show that fewer prenatal care visits are associated with increased maternal mortality;
- Providing telehealth services for mental health and substance use prior to, during, and following pregnancy, especially where access to mental health providers is limited;
- Remote monitoring of chronic conditions and other risk factors for reproductive age and pregnant women;
- Facilitating participation in childcare education classes, maternal health support groups, and community services or networks;
- Enabling access to online information and resources on family planning, pregnancy, childbirth, and the postpartum period.
However, the success of these solutions relies on available broadband at adequate speeds in the areas where women of reproductive age who are at-risk of poor maternal health outcomes.
Broadband
The broadband variables were chosen to provide insights on what can be enabled and delivered through the broadband health ecosystem of network, devices, and applications, and to identify exactly where gaps and opportunities exist. The selected variables include: broadband access, rural broadband access, Internet adoption, and upload and download speeds. The map also includes functional characteristics such as the number of broadband providers in an area and the percentage of the population served by each broadband provider and speed tier. Internet speeds are measured in Megabits per second (Mbps) and the tiers correspond to a set of speed ranges (e.g., 0-1 Mbps, 1-3 Mbps, 3-4 Mbps, 4-6 Mbps, etc. up to >1000 Mbps).
The map uses data on fixed broadband at download speeds of 25 Mbps and upload speeds of 3 Mbps. These are the current minimum broadband thresholds set by the FCC. It also displays information on the availability of higher speeds which are relevant to more complex health applications and use cases. The fixed broadband data in the current mapping platform come from the FCC’s Form 477 data program released in February 2020 and covering data submissions by all facilities-based broadband providers on the status of broadband as of June 2019, as well as revisions through September 1, 2020. (The Commission denominates this dataset as “Form 477 Broadband Deployment Data – June 2019 (version 2).” Internet adoption metrics are based on data released in the FCC’s 2022/2023 Internet Access Services Reports and capture the extent to which consumers subscribed to fixed connections at 200 kbps or higher in one direction at a household level in 2019.
For the broadband access metrics, the FCC releases data on the number of broadband providers and maximum download and upload speeds available at the census block level. Consistent with the Commission’s methodology in other contexts, a census block was considered to have access if at least one broadband provider advertised speeds of 25 Mbps download and 3 Mbps upload of higher. The population in each census block considered to have broadband access was noted and totaled across the county. By dividing the total population of census blocks with broadband access by the total county population, we were able to derive the proportion of the population in a county with broadband access. Rural broadband access is similarly calculated by dividing the total population in rural census blocks with broadband access by the total population of rural census blocks in the county.
Download and upload speed tiers were calculated by determining the proportion of population that has access to each tier as its maximum available offering. To calculate this, the highest speed tier offered in each census block was noted, the population in that census block was attributed to the speed tier, and the population of census blocks for each speed tier was totaled and divided by the total county population. Proportions for broadband access and speed tiers are calculated using the FCC Staff Block Estimates that provide estimates on the population for each block for the same year.
Chronic Disease
The data on chronic disease reflect four critical dimensions of health (i.e., population health outcomes, access to care, quality of care, and health behaviors) where broadband connectivity may be used to enable effective and cost-saving interventions. Population health outcomes include adult obesity and diabetes prevalence, health-related quality of life measures (such as self-reported poor/fair health and sick days in the past month), and premature death. Access to care is represented by the number of primary care physicians per 100,000 population, as well as the total number of primary care physicians, dentists, and mental health providers in each county. Quality of care is represented by preventable hospitalizations. These are hospitalizations for diagnoses that are usually treatable in an outpatient setting, indicating that the quality of care in these settings was sub-optimal. Relevant health behaviors include adult smoking, excessive drinking, and physical inactivity. We also include variables on selected community factors—injury deaths and severe housing—that provide insight on the physical and safety environment in a county.
These data sets are drawn from the 2021 release of the County Health Rankings & Roadmap program—a collaboration between the Robert Wood Johnson Foundation and the University of Wisconsin Population Health Initiative. They reflect the most recent data on the metrics of interest from the Centers for Disease Control and Prevention, Health Resources and Services Administration, Centers for Medicare and Medicaid Services, and American Medical Association, among others.
Opioid-Related Mortality and Prescription Rates
Pursuant to a congressional request that the FCC measure the potential impact of broadband access on the opioid crisis and that we utilize the Mapping Broadband Health in America platform to “overlay[ ] drug abuse statistics with the level of Internet access to help address challenges in rural areas,” the current update extends the platform to include variables on both health outcomes and risk factors for drug surveillance. It also reflects the first phase of a planned multi-phase effort to add drug abuse data to the platform. This initial phase focuses on a subset of variables that provide critical, baseline information for policymakers, namely opioid-related mortality rates, opioid prescription rates, and trends in each.
The opioid-related mortality data are based on death certificates for U.S. residents and reflect an average, age-adjusted mortality rate over a five-year period (2015-2019) accessed using CDC WONDER—the Wide-ranging Online Data for Epidemiologic Research. The CDC does not release mortality data when there are fewer than 9 deaths for privacy reasons or fewer than 20 deaths because of statistical unreliability. Given the large amount of suppression in the county-level data on mortality from the CDC, especially when a single year is selected, our approach uses a five-year average. We also include data on the broader category of all drug mortality given that more than 70% of drug overdose deaths involve opioids. These approaches have the effect of increasing the number of counties that meet the minimum death counts for data to be released and displayed in the mapping platform, while still providing a current snapshot of the situation. We identify counties where the CDC does not release mortality data as “Not Reported” in the mapping platform.
The CDC also recognizes three distinct waves in the opioid epidemic. Starting in 1990, the first wave of the epidemic reflected an increasing number of overdose deaths involving prescription opioids. A second wave of the epidemic began in 2010 with a spike in overdose deaths involving heroin. The current wave began in 2013 with a rapid increase in overdose deaths due to synthetic opioids, notably fentanyl. In recognition of these distinct waves, the mapping platform now includes several sub-categories of opioid-related mortality. Users can select specific data on overdose mortality related to prescription opioids, heroin, or synthetic opioids.
Finally, given the role of prescription opioids in driving the opioid epidemic, we also include data on the opioid prescription rate. While deaths due to prescription opioids may no longer be the most common cause of opioid-related mortality, the prescription rate provides an important measure of risk due to the level of exposure to opioids that exists in a community as a potential gateway to opioid abuse or other opioids use. Thus, this data may provide an early indicator of potentially emerging areas of need and targets for prevention efforts. Data on opioid prescription rates for 2019 come from the CDC’s U.S. Opioid Dispensing Rate Maps accessed in 2023.
To help users assess trends over time and identify potentially emerging hotspots, we also calculate a percent change in the mortality and prescription rates. To calculate the percent change for the mortality data, we compare two consecutive, five-year periods (2015-2019 and 2010-2014). The percent change in opioid prescription rate reflects the most current data (2020) compared to the previous year (2019). Trends are calculated by subtracting the value for the baseline period from that for the current period and dividing by the baseline period to produce a percent change. If either or both values for the baseline or current period are missing, a percent change cannot be calculated; and we depict this in the mapping platform as “Unavailable.”
Demographic and Socioeconomic
To provide characterizations and context for visualizations of the intersection between broadband and health, the mapping platform also includes other relevant demographic information and socioeconomic determinants, including population, rurality, education, and income (with different variables applicable to certain disease states). This information comes from various publicly available sources, including the U.S. Census Bureau, the FCC’s 2021 Broadband Deployment Report, and the County Health Rankings & Roadmaps.
Population is a critical variable underlying the demographic information and calculations in the mapping platform. We use the values for population from the U.S. Census Bureau as the primary source for population estimates. In addition to providing information on population statistics, the mapping platform also enables users to overlay population data in certain visualizations. We also use this data to calculate the number of people living in urban or rural areas and identifying as male or female for display in the chronic disease and opioids modules.
Given the importance of rurality to better understand the broadband health space, we also include information and functionality on rurality. The U.S. Census Bureau is also the primary source for information on urban and rural areas. The U.S. Census Bureau defines an urban area as a “densely settled core of census tracks and/or census blocks that meet minimum population density requirements, along with adjacent territory containing non-residential urban land uses as well as territory with low population density included to link outlying densely settled territory to the densely settled core” and encompass at least 2,500 people. Rural areas are all those not identified as urban areas. Users can filter by counties that are majority or minority rural or based on quintiles of rural population (0-20%, 20-40%, etc.). We define majority rural as those areas where 50% or more of their population live in rural areas as defined by the U.S. Census Bureau. Minority rural are areas where less than 50% of the population live in rural areas.
States, Counties, and Census Blocks
Map features are available at the state and county zoom levels. We chose to focus on counties for several reasons: (1) county level data are available across most health measures (e.g., diabetes, obesity, preventable hospitalizations) and connectivity variables; (2) counties are a discrete geographic unit with community governance that can potentially drive broadband economies and local health policy (i.e., they are neither too broad such as a state level geography nor too granular such as neighborhood level geography); and (3) counties are the building blocks for publishing many types of data (e.g., economic data) and for tracking progress and regional population and economic trends.
Map Design
The map includes three primary views: for Overview, Broadband, and Health. Below are some examples of data visualizations that can be extracted from the map:
- In the Overview tab, sliders enable users to select levels of broadband and health variables of interest, and colors are used to identify geographic areas that meet those parameters. The default setting for the slider values reflects the national average.
- In the Broadband and Health tabs, we assigned a color-coded value for each broadband variable (yellow to blue), maternal health and opioid variables (light pink to purple), and chronic disease variable (yellow to red) in every county and state. In the broadband and health choropleth maps (found under the corresponding views), the color of each state or county represents the corresponding level of broadband, maternal health, opioids, or chronic disease metric for that feature. For broadband, the color ranges correspond to ranges from 0-20%, 20-40%, 40-60%, 60-80%, and 80-100%. For maternal health, opioids and chronic diseases, the choropleth spectrum corresponds to equal intervals from highest to lowest value for that metric with quintile ranges calculated from the data.
- Users can also filter the data in the Broadband or Health views by the broadband or health variables of interest. For maternal mortality, for example, radio buttons allow users to display the broadband picture in maternity care deserts or in areas with low, moderate, or full access to obstetric care. For opioids-related mortality, slider bars enable users to visualize select areas based on their mortality rate as a function of the national average (number of times the national average). If looking at trends, they can select areas that are increasing, decreasing, or specify the change of interest. For broadband and chronic diseases, dropdown menus allow users to visualize areas based on prespecified breaks in the data.
In addition, the map provides supplemental information for each state and county in statistics, spider plots, charts, and graphs. (see below). These features were designed to help users who may be less familiar with the broadband health profile of targeted areas, to help characterize and contextualize the visualizations, and to facilitate better comparisons between counties.
Open Data, Open Source
The map platform allows the user to conduct further analysis by downloading the data in a number of formats including CSV. It also allows developers to integrate the data into their own platforms using JSON or XML data download formats. The code is also open source, and our "open integration" model allows users to integrate their own data to visualize its impact on the broadband health space. We encourage developers to leverage the platform for their own use. Please use the Community Engagement tab to share efforts in analysis and development using the platform and to identify other data or features that would be useful.