U.S. Coronavirus Data: Frequently Asked Questions
By The New York Times
What is the source of this data?
Since late January 2020, The New York Times has tracked coronavirus cases and deaths as they have been announced, using data released by countries, states and local health officials. Times staffers work around the clock to create, verify and maintain this data.
This data is used in Times reporting and graphics, sometimes as cumulative figures, new daily totals or per capita counts.
The U.S. case and death data is available to the public on Github.
The Times uses testing and hospitalization data published by the U.S. Department of Health and Human Services, and it has used data collected by The COVID Tracking Project. The Times uses case and death data from Johns Hopkins University for most countries.
How often are the virus tracking pages updated?
The Times’s virus tracking pages are updated multiple times a day, and county-level risk is updated every morning. Most states update their data at least five days a week, though some states and many counties report less frequently.
Testing and hospitalization data for the previous day is typically updated daily. County-level testing and hospitalization data is updated weekly when available, typically on Mondays.
Why do numbers and risk levels differ from what I’m seeing from my state or county government?
The Times data is sourced from states, counties and regional health departments. Local officials often report earlier than the states do, so this can be a source of variation. Displayed risk levels are based on an analysis of coronavirus case and testing data by The Times and public health experts, which may not align with states’ own risk assessments.
Jurisdictions typically count cases and deaths based on a person’s residence, rather than the location where they tested positive or died, though there are exceptions to this in Alaska, Hawaii and Vermont.
State and local counting methods can vary. To learn more about the data for a specific state, see the About the Data section on each state page.
The Times also makes some adjustments to create a more uniform dataset across jurisdictions. For example, some states do not include cases in correctional institutions in county totals, and The Times includes these cases in the appropriate county when possible.
In some specific situations, such as an outbreak on a cruise ship or military vessel, The Times has tracked cases based on the locations where patients were treated.
There is also variation in the way states and counties report probable cases and deaths, which is described below.
Is data for previous days or months updated retroactively?
The U.S. Department of Health and Human Services frequently makes historical corrections to data about hospitalizations and testing. These ongoing adjustments are reflected on our pages.
Data about cases and deaths from The Times’s database is not changed retroactively except to correct errors, like when a state or county provides a corrected number the following day, or to incorporate new data sources, such as including probable cases.
Why is my county at a particular risk level?
The Times developed risk level guidance with public health experts at Johns Hopkins Bloomberg School of Public Health and Resolve to Save Lives, which is an initiative of Vital Strategies. The risk levels are determined for each location based on recent data about cases and testing. This is an independent analysis unrelated to states’ own risk level assessments.
Some locations are assigned an unknown risk level if there is insufficient data available. To view a detailed explanation of the thresholds for each risk level, view the “About the risk levels” section on risk pages.
What are probable cases and deaths, and are they included in the data?
While confirmed cases and deaths are counts of people whose coronavirus infections were confirmed by a molecular laboratory test, probable cases and deaths count those who meet criteria set by state and federal governments — including verified exposure to the virus, having symptoms common to it, and other test results that suggested infection. Many governments include only probable cases identified through antigen testing.
Confirmed cases and deaths are widely considered to be an undercount of the virus’s impact, so many states and counties began reporting probable coronavirus cases after the Centers for Disease Control and Prevention issued guidance to do so in April 2020 and in August 2020.
Not all governments report these the same way. Some release only confirmed cases and deaths; some report confirmed and probable numbers separately; and others report a combined confirmed and probable figure. If a jurisdiction reports probable cases and deaths, these are typically presented on the health department’s website, though sometimes these are not clearly labeled. The Times verifies this information with officials and includes probable cases and deaths when reported, as recommended by the C.D.C.
Probable cases and deaths can include people who test positive from a rapid antigen test. Public health officials say that cases based on antigen tests are most likely undercounted across the country.
To see whether a state includes probable cases and deaths, visit the individual state pages listed at the bottom of this page.
Why doesn’t the page show how many people have recovered?
Data on the number of recovered patients or currently active cases is not consistently available at the county level. The number of people who have recovered, as tracked by some states and counties, is an estimate based on an assumption that people diagnosed with Covid-19 have recovered after a particular length of time if they have not died. The criteria varies by state. Health officials do not typically follow up with people to find out when they have recovered.
How are daily averages calculated?
Case and death data for the U.S. is presented as a rolling average with a flexible time range to account for volatility in daily case counts and holiday reporting schedules. The average for states and counties are calculated over at least seven days of reported data. If the most recent seven-day window includes days with no data reported, the period is extended to older days until it includes seven days of reported data.
Some day’s figures may count as more than one day of reported data. For instance, data reported on a Monday following two weekend days with no data would count as three days of data. The count from a day following non-reporting days is always averaged over that day and the non-reporting days that precede it.
For national case and death counts in the U.S., the average is calculated from the sum of the average number of cases and deaths reported each day in all states and territories. This average may not match the average that results from the daily total of U.S. cases and deaths. This is to account for irregularly timed case and death reports at the state level.
Figures for other countries will continue to use a simple seven-day average.
Why do spikes sometimes appear in the daily case and death charts?
The Times data is based on the date cases and deaths are announced, as this is the closest to real-time data that is available from states. Governments sometimes revise data or report a single-day large increase in cases or deaths without specifying when those cases and deaths occurred, which can cause an irregular pattern in the daily reported figures. The Times is excluding these anomalies from seven-day averages when possible.
To see a detailed list of all reporting anomalies, visit the individual state and county pages.
Why do cumulative counts sometimes decrease?
Health officials frequently remove or reassign cases and deaths after receiving new information, resulting in small decreases in total state and county tallies. Common reasons include removing duplications or cases and deaths that turn out to involve people who live in other jurisdictions. Occasionally, jurisdictions report larger decreases after changing case or death definitions.
Which states are included in each U.S. region?
In the chart showing trends by geographic region, the Northeast region includes: Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, West Virginia and Washington, D.C.
The South region includes: Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, Texas and Virginia.
The Midwest region includes: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota and Wisconsin.
The West region includes: Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington and Wyoming.
How do Covid-19 deaths compare with other causes of death?
The Times has examined government mortality data to determine the number of “excess deaths” that occurred in 2020, and it has compared those deaths with official coronavirus death counts. This U.S. page and international page were updated regularly during the first year of the pandemic.