No numbers can fully convey the state of the covid-19 pandemic in the United States, but several metrics taken together provide a clearer view of what is happening now and what may be coming soon.
The data in these charts is gathered by The Washington Post from local and state government sites and from Johns Hopkins University (cases and deaths), the Department of Health and Human Services (hospitalization and testing) and the Centers for Disease Control and Prevention (vaccinations). Not all tracking sites use the same sources or report dates, so case and death numbers can differ.
Cases in the U.S.
New reported cases per day
At least have been reported since Feb. 29, 2020.
7-day average cases
The count of known new cases is the closest thing we have to a real-time gauge of the coronavirus’s reach.
It also can serve as a warning: If cases suddenly rise, hospitalizations and deaths almost certainly will follow. However, areas with little testing probably will appear to have fewer cases regardless of how rampant the virus may be.
The seven-day rolling average is the best way to view trends in new cases, because more than half of states don’t report new data every day. That is why certain days of the week show up as consistently higher or lower than others.
The chart above shows one-day spikes and drops that have occurred as states refined their data. For instance, a state may remove many duplicate cases all at once or reclassify a group of cases based on new criteria. Those anomalies are not factored into the seven-day averages.
|Place||Total reported cases per 100k||Avg. daily new cases per 100k||Change in avg. daily cases in last 7 days|
Covid-related deaths in the U.S.
New deaths reported per day
At least have been reported since Feb. 29, 2020.
7-day average deaths
Health officials believe the virus has killed more people than state totals indicate, especially early in the pandemic before testing and effective treatments were widely available.
A rise in deaths usually follows a rise in new cases by about a month. For example, after the delta variant caused a surge of new cases beginning in July 2021, the death toll began to climb in August.
As with the new cases chart, occasional single-day anomalies reflect one-time adjustments by states. They are noted but not included in the seven-day averages.
|Place||Total reported deaths per 100k||Avg. daily new deaths per 100k||Change in avg. daily deaths in last 7 days|
Reported covid-19 hospitalizations per 100,000 residents
The number of people in hospitals is key to understanding an outbreak’s effect on a community. If hospitals or their intensive care units are full, people seeking treatment — whether for covid or for something else — may have care delayed or even denied.
A rise in hospitalizations tends to follow a rise in new cases by a couple of weeks.
|Place||Currently hospitalized with covid per 100k||Currently occupied ICU beds per 100k||Change in hosp. from last week|
About this storyOriginally published March 27, 2020.
Recent changes on this page
August 17 Removed the “hot spots” chart in favor of a broader summary of cases, deaths and hospitalizations. Removed data on test positivity due to the increasing use of home COVID tests, most of which go unreported to health agencies. Removed vaccinations, as data on COVID vaccinations have become less instructive as the counts of those receiving initial shots or boosters drops.
August 12 Removed the county-level map, pending a review of how to best display inconsistently reported county-level data.
January 20, 2022 Switched the percent change column in the county tables to the percentage change in rolling average, rather than in total cases/deaths.
November 10 Redesigned page and added features.
March 3 Changed the data source for tests to the Department of Health and Human Services.
February 23 Changed the data source for hospitalizations to the Department of Health and Human services. See the methodology note for more details.
February 19 Marked anomalous days on the daily count chart, and reformatted the chart note to better track data anomalies.
February 12 Updated vaccination charts to reflect reported doses administered per day.
February 2 For five states with limited state hospitalization reports (WV, NV, ND, MS, and WY), data now comes from the Department of Health and Human Services, instead of state reports.
January 8, 2021 Added a data table and charts for vaccinations, and changed the vaccination metrics to display the most reliable data available.
December 23 Added vaccination data to state summaries where available.
December 15 Removed anomalous data from the rolling averages (such as backlogged cases or deaths reported in bulk on a single day), and other data improvements.
December 1 Updated the presentation of hospitalization and testing data, and added aggregate U.S. data for those indicators.
October 30 Added several notes clarifying which days states are expected to report data.
October 28 Switched to reported case counts from the Kansas Department of Health and Environment which only provides data updates on Mondays, Wednesdays and Fridays. This has also resulted in a one-day spike of reported cases on Oct. 28.
September 18 Switched to using the confirmed death counts for Bronx, Kings, New York, Queen's, and Richmond counties as reported by New York City, while continuing to use the state's reporting for deaths in all other counties. This has resulted in a one-day spike of 2,732 deaths. Read more about how NYC's methodology differs from NY state's. The Post has been using the city's probable death counts since April.
Given the difference in the methodologies between the state and the city, the Post feels that the city's numbers, which are derived both from positive blood tests and from deaths reported by the city's Office of the Chief Medical Examiner, provide a clearer indication of the fatality count than the state's approach, which uses numbers reported by hospitals, nursing homes, and other health care facilities. This means that the city's counting process is more likely to include deaths that occurred outside of care settings.
August 24 Replaced the modeled trend with a more standard 7-day rolling average of new daily cases and deaths.
August 20 Added a module to show the aggregate statistics in the U.S. and each state/territory over the past week.
July 29 Added hospitalization data and other page improvements.
July 2 Replaced the 7-day running average of new cases and deaths with a 14-day modeled trend. Added the week-over-week percentage change to the trends charts, using the modeled trend values. Also added additional columns to the data tables.
June 23 Added charts showing new daily counts in each state, ordered by the percentage increase in cumulative cases over the last week. Changed the default view of the page to confirmed cases per 100k.
June 11 Added an option to view change since last week to the map. The default view of the map is now deaths per 100k in the last seven days.
May 13 Added a line indicating the seven-day rolling average or reported cases and deaths to the national and state by day chart at the top of the page. The deaths total at the top of the page was revised to round the deaths number down to the nearest thousand.
May 6 Included revised data from New York City probable covid-19 deaths that attributes each death to the day it was first reported instead of on April 14.
April 24 The data on the page was revised to include Post-reported numbers. Reported data for New York City is now reported separately by county instead of being aggregated into one New York City total.
April 23 Date when states began reopening added to state charts.
April 21 Charts showing testing data for all U.S. states and territories were added to the page.
April 14 New York City adds nearly 3,700 probable covid-19 deaths to its total.
April 7 Labels showing the date state emergency and stay-at-home orders were declared added to the state charts.