COVID-19

Data sourcing

The primary data source is the CDC Case Surveillance Restricted Access Detailed Data. This dataset allows for detailed breakdowns by race, age, and even specific ten-year age intervals. We previously used the New York Times COVID Dataset as a benchmark to evaluate the comprehensiveness of state data in our tracker, however the New York Times itself now relies on the CDC dataset making comparisons less useful.

Age and Demographic Data Analysis

Age categorization involves ten-year age buckets, ranging from 0-9 up to 80+. The case-level data also can provide a cross-sectional breakdown by both race and age.

Geographical Distribution and Reporting

National Data

National statistics are based on summations of all state level cases.

However, if state data is not accurately provided, these national aggregations may be incomplete and potentially skewed.

County Data

Specific figures might be concealed in counties with low case counts to protect the privacy of affected individuals. The foundational data provided by the CDC is detailed at the case level. Therefore, when a county doesn't report cases for a certain demographic, it's uncertain whether the county truly lacks cases for that group or if there's a lapse in accurate demographic reporting.

Time-Series and Temporal Analysis

Utilizes the cdc_case_earliest_dt ("CDC Case Earliest Date") field from the CDC Restricted dataset. Data is categorized based on the earliest of several factors:

1. symptom onset,
2. positive test date, or
3. report date

to the CDC. Monthly charts represent the incidence rate, indicating the number of new cases reported that month. We categorize each COVID case, death, and hospitalization based on the month and year recorded in the dataset.

However, it's crucial to highlight that, for deaths and hospitalizations, our data reflects the month when the case was initially reported, not the actual month of the death or hospitalization event.

In our per 100,000 and inequitable distribution metrics, we only account for confirmed deaths and hospitalizations.

Therefore, if our data indicates 'zero' deaths or hospitalizations for a specific demographic in a given month, it's possible there were unreported or unconfirmed events for that group during that time frame that the CDC hasn't noted. If any geographic area shows no data for cases, deaths, or hospitalizations for a particular demographic throughout the pandemic, we exclude that demographic from our visual representations, under the assumption that data collection for that group might be lacking.

Addressing Missing and Suppressed Data

Many of the source data case records aren't broken down comprehensively, may not specify hospitalization or death statuses, or might lack the complete demographic details needed to fully understand COVID-19's overall impact.

Vaccination Data Compilation and Analysis

Vaccination definitions might vary across datasets, with terms like “at least one dose” and “total doses administered” being used.

Due to the lack of a consolidated national vaccine demographic dataset, multiple sources are use across different geographic levels:

National Data

Derived from the CDC vaccine demographic dataset.

State Data

Extracted from the Kaiser Family Foundation's COVID-19 Indicators dataset.

County Data

At the county level, we utilize the COVID-19 Vaccinations in the United States, County dataset. This dataset provides only the total number of vaccinations administered in each county, without the racial and other demographic breakdowns we normally present. It is our hope to standardize our vaccination data on higher quality, more detailed recent data in the near future.

Demographic Population Estimates for Vaccination Data

A note about the 2019 American Community Survey (ACS). While the American Community Survey (ACS) is a valuable resource for many demographic insights, it has its limitations in the context of our study. The ACS doesn't provide the specific age-segmented estimates that the CDC offers.

National Estimates

We use the CDC's population numbers for our national figures, especially when considering regions like Palau, Micronesia, and the U.S. Marshall Islands, for which obtaining accurate data can be challenging.

State and County Estimates

Accurate population estimates are essential for understanding the distribution of vaccinations and pinpointing disparities, especially among indigenous and underrepresented groups. For the state level, we base our primary data on the ACS 2019 estimates to calculate each state's total vaccinations. It's noteworthy that state-reported population categories sometimes differ from census categories, leading us to lean on the Kaiser Family Foundation's (KFF) tabulations. The Kaiser Family Foundation provides detailed population counts for specific demographic groups at the state level, including Asian, Black, White, and Hispanic populations.

However, since the KFF data doesn't comprehensively cover indigenous groups, we supplement with ACS 2019 estimates for American Indian and Alaska Native, as well as Native Hawaiian and Pacific Islander groups.

On our disparities bar chart, these specific population metrics stand out with a different color. Yet, the Unrepresented Race category poses challenges due to its varying definitions across states. As a result, direct comparisons become challenging. The Percent of vaccinated metrics for Native Hawaiian and Pacific Islander, and American Indian and Alaska Native, are juxtaposed with ACS 5-year estimates, while the Unrepresented Race lacks a direct comparison metric. For county-level insights, our data stems from the ACS 2019 population estimates. However, it's crucial to note that the CDC's county-level vaccine dataset only presents overall vaccination figures without any demographic breakdown.

Data Limitations and Specific Considerations

In our analysis, we account for unique state-specific reporting conditions. For instance, New Hampshire, after lifting its national COVID-19 emergency response declaration in May 2021, allowed vaccine recipients to opt out of state records, potentially skewing the data since then.

Additionally, there are disparities in state reporting methods; some states differentiate race and ethnicity, leading to varied percentages of unknown cases. We choose to display the higher metric on national maps and both figures on state pages for clarity.

The Kaiser Family Foundation's data collection primarily encompasses Asian, Black, White, and Hispanic demographics, limiting the scope of our per 100k metrics and demographic breakdowns at the state level.

Another challenge arises from the lack of a standardized definition for “vaccinated.” Typically, it refers to individuals who've received at least one vaccine dose. However, states like Arkansas, Illinois, Maine, New Jersey, and Tennessee report total vaccine doses administered, adding another layer to our comprehensive analysis.

• COVID-19 vaccinations: race/ethnicity: Available; age: Available; sex: Available
• COVID-19 cases: race/ethnicity: Available; age: Available; sex: Available
• COVID-19 deaths: race/ethnicity: Available; age: Available; sex: Available
• COVID-19 hospitalizations: race/ethnicity: Available; age: Available; sex: Available

COVID-19 vaccinations

Measurement definition: For the national and county levels, and for most states, this indicates people who have received at least one dose of a COVID-19 vaccine.

Clinical importance: COVID-19 vaccinations are an important tool for preventing the spread of the virus and protecting people from serious illness. However, vaccination rates vary significantly across different populations. Studying COVID-19 vaccinations in regard to health equity can help us to understand why these disparities exist and how to increase vaccination rates among all populations.

COVID-19 cases

Measurement definition: A COVID-19 case is an individual who has been determined to have COVID-19 using a set of criteria known as a “case definition”. Cases can be classified as suspect, probable, or confirmed. CDC counts include probable and confirmed cases and deaths. Suspect cases and deaths are excluded.

Clinical importance: COVID-19 has had a disproportionate impact on certain populations, including older adults, people of color, people with disabilities, and people living in poverty. Studying COVID-19 in regard to health equity can help us to understand why these disparities exist and how to address them.

COVID-19 deaths

Measurement definition: The number of people who died due to COVID-19.

Clinical importance: COVID-19 has had a disproportionate impact on certain populations, including older adults, people of color, people with disabilities, and people living in poverty. Studying COVID-19 in regard to health equity can help us to understand why these disparities exist and how to address them.

COVID-19 hospitalizations

Measurement definition: The number of people hospitalized at any point while ill with COVID-19.

Clinical importance: COVID-19 has had a disproportionate impact on certain populations, including older adults, people of color, people with disabilities, and people living in poverty. Studying COVID-19 in regard to health equity can help us to understand why these disparities exist and how to address them.

• UCLA HEALTH: COVID-19 exposes how Native Hawaiians and Pacific...
• National COVID-19 Resiliency Network (NCRN)
• COVID-19 and Equity - APHA
• COVID-19 and Health Equity: A Policy Platform and Voices

COVID-19 Vaccination Resources

• Building Trust and Access to the COVID-19 Vaccine in Communities of Color and Tribal Nations
• CDC - U.S. COVID-19 Vaccine Product Information
• APM Research: Death & Vaccination Statistics
• Bloomberg: COVID-19 Global Vaccine Tracker