District of Columbia

@ChristineRuana @alexismadrigal We're trying! Without the ability to compel the states to share information (like the CDC could) — we're doing our best to fill in the information. -@alexismadrigal

We are tracking the number of people who have been tested here, which is different from "number of tests," which is what the CDC has released so far.

Washington has tested the most people and has the most cases (267). NY now second for cases at 216.

Our daily 4 pm update is out! Across 50 states and DC we've tracked 16,502 total tests: 1,953 positive, 13,419 negative, and 1,130 pending.

Before trying to interpret this data, please read over our notes about how each state reports data differently: https://t.co/NFXhO4aZOz.

Our daily 4 pm update is out! Across 50 states and DC we've tracked: 2,450 positives 17,071 negatives 1,236 pending.

Total tests: 20,795

Before trying to interpret this data, please read over our notes about how each state reports data differently: https://t.co/NFXhO4sAG7.

Daily 4pm update. Across 50 states and DC we've tracked: 3,173 positives 22,548 negatives 2,242 pending.

Total # of people tested: 27,963.

Before trying to interpret this data, please read over our notes about how each state reports data differently: https://t.co/NFXhO4sAG7. https://t.co/tnECTqQvrr

@jgdean14 @Mightymouth5413 More importantly, a CDC could force standardization! We just can't make people do that. -@alexismadrigal

End of day, 3/17, we have confirmed cases in all 50 states + DC and 3 territories, almost 6000 in total. WA and NY are now over 1000 cases each. Of the smaller states, Louisiana, Colorado, and Rhode Island look like hot spots. https://t.co/LzKJBXTD0J

@davidgaliel @AdrienneLaF @TheAtlantic The CDC is the only entity I know that explicitly is reporting that way. And they, explicitly, used to report PEOPLE tested. Check the Internet Archive, you'll see that changed in March. But maybe you have some sources we don't and I'd be happy to hear that-@alexismadrigal

Before our update, we wanted to show you why we do this. Tracking testing data in the U.S. has been a challenge. The CDC is currently publishing incomplete data that lags several days behind state reports. We collect and clean the state data. Here’s how the two datasets differ: https://t.co/Eu8pbiZ5Rf

In a presser, @VP Pence just gave two important testing tidbits:

1. He gave a total test number of 313,000 (we show 294k, for a 94% capture rate).

2. He said all labs are now required to submit their results to the CDC. We hope that includes negatives. #weneedthedenominator

@elonmusk @teslaownersSV @RenataKonkoly @sdunbabin @jonkay @Quillette The CDC is not reporting data in anything close to real-time and relying on it will make your analysis a couple weeks late. -@alexismadrigal

@announcer_box @NateSilver538 This is definitely something we're looking at. The original impetus was the CDC cuts off at 4pm, but that has not turned out to be the main issue. -@alexismadrigal

Deaths rose today in our dataset by 2,194 to 47,980.

Note New York City is reporting 4,453 more confirmed and probable COVID-19 deaths than the state government is for the city.

If you add those deaths into our dataset, our count (52k) falls between the CDC and JHU. https://t.co/dznPM4bhln

The death toll today is 1,510. That brings the total deaths in our dataset over ~77 thousand.

This does not include 5,136 probable COVID-19 deaths reported by NYC, but not New York State, nor does it address excess mortality as reported by the CDC here: https://t.co/4ybLKisg6u https://t.co/upQWMTFN37

Our testing data collection is proceeding — but/and we hope to be able to sort things out with the CDC over the coming days.

The reported death toll was 1,741, bringing our dataset over 80,000.

This does not include 5,057 probable deaths reported by New York City but not the state, nor does it address excess mortality as reported by the CDC here:

https://t.co/4ybLKisg6u https://t.co/PCJMSdieQw

1,645 deaths were reported today, bringing the total in our dataset to 81,729.

This does not include 5,054 probable deaths recorded by NYC, but not published by New York State, nor does it address excess mortality as reported by the CDC here: https://t.co/4ybLKisg6u https://t.co/WTdXIsfeTa

1,286 deaths were reported today, bringing the total in our dataset to 83,015.

This does not include 4,820 probable deaths recorded by NYC, but not published by New York State, nor does it address excess mortality as reported by the CDC here: https://t.co/4ybLKisg6u https://t.co/zK1qCAM17t

We generated as complete a report as could comparing the CDC data with the numbers we compile from the states themselves. https://t.co/euC0pIKyXH

There is a known difference between what states report and what the CDC says it is reporting. Some states report "people tested" while the CDC reports "specimens tested."

4 states report both—so we used them to calculate a ratio to adjust the "people tested" numbers. https://t.co/waQEhbTTEY

The upshot: we will continue compiling and publishing data as we have been.

We're also ingesting the CDC data.

And we'll be working hard to understand and close the gaps we see.

Ultimately, one dataset should exist, and it should be managed by @CDCgov. That's the goal.

@AquaShena We don't know if the CDC is doing that or if it's something else.

Yes to the racial data tracker in the API.

No to county data, at least for now.

-@alexismadrigal

The CDC currently reports 99,031 deaths.

In general, the discrepancies in death counts between different sources are small and explicable.

The death toll stands at roughly 100,000 people today.

For one thing, only half of states report them despite direction to do so from the CDC. Also, although probables currently account for less than 1% of the nat'l total, they make up 8-20% in the states with the highest proportion of probables, and 75% of all cases in Puerto Rico.

The variation between states/territories is also fascinating. DC told the Washington Post it doesn't report probables because it doesn't have any. Delaware, meanwhile, with a comparable total case count, reports nearly 1,000 probable cases.

States reported 630 deaths. We are still seeing a solid national decline.

Death reporting lags approximately 28 days from symptom onset, according to CDC models that consider lags in symptoms, time in hospital, and the death reporting process. https://t.co/LBmcot3h9a

While doing the research for this change, we learned more about the two different ways states determine COVID-19 deaths. Both have merits: one is a little more accurate, the other is a lot faster. CDC advises one, but uses both. States are in a bind.

For the record, we compile current hospitalizations for 48 states and DC now. There are variations in how states count their COVID-19 patients. Because these numbers measure strain on the hospital system, we record both confirmed and suspected COVID-19 hospitalizations.

Here’s what’s going on. As states switch from the long-established CDC system to the new HHS one, some hospitals appear to be having trouble reporting their numbers. That led both Texas and California to place advisories on their dashboards. https://t.co/fbq1Su5Nz2

But here’s some great news about hospitalization data: We now have figures for current COVID-19 hospitalizations for 50 states and DC + Guam and Puerto Rico. Once the data settles down from the systems changeover, this should be a very solid national number.

Hospitalizations fell a bit again. CA and TX (and maybe other states) are still not getting complete data due to the HHS changeover. We’ve frozen the CA and TX figures for now — and will reevaluate tomorrow. We’re hoping the CDC-HHS changeover issues get cleared up this week.

And finally, deaths from the beginning to yesterday.

(Along with your periodic reminder that New York State still is not reporting over 5000 probable deaths recorded by New York City and the CDC.) https://t.co/k0HugoRg4Q

We recorded a considerable drop in hospitalizations today. The new HHS hospitalization dataset had been tracking much higher than our numbers; today, it experienced a precipitous and unrealistic decline. Short story: The data remains mushy since the CDC-->HHS changeover. https://t.co/907rNsK96h

Update for our data users - we continued to implement changes to our original totalTestResults API field, resulting in a ~600k cumulative test increase since March.

We updated test totals for 14 states: AK, AL, AR, AS, CA, DC, GA, ID, KY, MN, NH, SD, VA, and WA.

Not only are the units confusing, the formulas for calculating test positivity also vary. We did the math for five states using three methods validated by the CDC and the results are wildly divergent. https://t.co/QbAqz0TybN

In the month of October, only Georgia and Hawaii saw decreases in hospitalization. California held steady. There were increases in hospitalizations in every other state and DC. https://t.co/tIwKEeNBeJ

Hospitalizations are also up in 34 states plus DC from 2 weeks ago. https://t.co/dF1ewRt2my

NHSN, which reports LTC data to the CDC, announced the removal of “suspected” cases, and a shift from “confirmed cases'' to “positive cases'' which can now be determined by antigen testing or PCR. This change will make it harder to decipher LTC case data:https://t.co/gJqykALk7p

Our project is not tracking vaccine data, but the CDC is now reporting two topline vaccination metrics for the US. We’re also looking at the kind of vaccine data states have published so far and have outlined our hopes for federal reporting:

https://t.co/DXbuK4R4l8

20 states provided no update: AK, CA, CT, DC, ID, KS, KY, LA, MA, MI, MN, MP, NC, ND, NH, OH, RI, SC, SD, UT, VT.

7 other states made partial updates.

Here's a snapshot of what the key COVID-19 metrics look like in Washington, DC as of today. https://t.co/FBdIpmae7a

Last week, South Carolina released federal vaccine data that shows which long-term-care facilities have received doses. They are the only state providing this transparency on the vaccine process.

The CDC and other states should follow their lead.

https://t.co/xLH7J8gg67 https://t.co/uUUo1vek3J

The CDC and HHS are now publishing data that is much more comparable to the figures we have been compiling from states since last spring. Gaps and inconsistencies remain, and we will continue to analyze federal data and document what we find through the end of May.

The CDC has a new COVID Data Tracker Weekly Review that is simple and useful, as well as a good starting point for understanding what is available in federal government data repositories. https://t.co/wq8IeDIQXH

As we wind down data compilation, we want to help everyone understand how to find and use federal COVID-19 numbers. Today, we’re going over CDC case data. https://t.co/6cm838CYO1

The CDC aggregate case data is extremely similar to the state-reported data we have been capturing. https://t.co/7qFQCzLg7j

You can download this data from @CDCgov and on https://t.co/XOJY7f5vSx in a variety of formats, including CSV and XML. It’s also possible to filter, sort, and visualize the data on the CDC website without downloading it. https://t.co/jgQXeGonj3

(NYS and NYC both report deaths to the CDC. There are about 8.7k deaths that NYC reports which are not counted by NYS, and because we use the state’s reported numbers, our dataset does not reflect those additional deaths.)

We also wrote up a brief introduction to the easiest ways to keep up-to-date on the pandemic using CDC and other federal data. https://t.co/quJMBxIjod https://t.co/bQl39SXLdS