Virginia

States: Alaska, California, Washington DC, Georgia, Kentucky, New York, Ohio, Oregon, Texas, Virginia, Washington

Issue: We are removing negatives that were created from mixed units (specimens minus cases or test encounters minus cases) for states that are using explicit totals in our main total test results field (called totalTestResults in the API). See the Data FAQ for additional explanation.

States: Alaska, California, Washington DC, Georgia, Kentucky, New York, Ohio, Oregon, Texas, Virginia, Washington

Issue: We are removing negatives that were created from mixed units (specimens minus cases or test encounters minus cases) for states that are using explicit totals in our main total test results field (called totalTestResults in the API). See the Data FAQ for additional explanation.

Target Date: Backfilling data from as far back as Jan-26th, 2020 Overview: Antigen and antibody totals and positive encounters, will be filled in the specimens fields.

Target Date: Backfilling data from as far back as Jan-26th, 2020 Overview: Antigen and antibody totals and positive encounters, will be filled in the specimens fields.

State or US: VA

Describe the problem VA added antibody testing encounters data and antigen test encounters that we're not capturing on 11/23. (We didn't capture the antigen because we were unsure about lumping with antibody) We can get historical data for it. https://www.vdh.virginia.gov/content/uploads/sites/182/2020/11/VDH-Antibody-Test-Reporting-Insights.pdf

Link to data source https://www.vdh.virginia.gov/coronavirus/coronavirus/covid-19-in-virginia-testing/ https://data.virginia.gov/resource/3u5k-c2gr.json

State: VA

Dates affected: 3/19-12/1

Describe the issue: On April 26, 2020, we began reporting Virginia's Ever hospitalized as calculated from {"Combined number of confirmed positive COVID-19..." + "...discharged"} from the official Virginia Hospital and Healthcare association public dashboard. On December 2, 2020, we revised out time series of Virginia's Ever hospitalized to align with what the state reports on the state dashboard.

State or US: Virginia

Describe the problem This is a number that was retrieved once from reporting and carried for one week https://www.facebook.com/GovernorVA/videos/218402129225741/ reporting an errant cumulative number for one week is skewing the national numbers. We should either remove the numbers, or carry them forward.

Link to data source

Additionally, VA has a 4 day gap in cumulative hospitalizations: 2020-04-14 - 2020-04-11

State or US: Virginia

Describe the problem Virginia provides total tests encounters and PCR test encounters. We relied on previous outreach and were using the difference between the two numbers as antibody positives & totals, but based on recent outreach it appears that this number might also contain antigen tests, so we have to delete the previous capture.

Action Plan Remove time-series of positive and total antigen in VA

Related Issue https://github.com/COVID19Tracking/issues/issues/840

State or US: VA

Describe the problem CTP started reporting probables generally around 7/28 for many states. VA has been reporting probables since it started separating out confirmed and probable case counts on 4/21. We can backfill probables using positive - confirmed until 4/21 and cross-referencing with screenshots.

We can backfill antibody positives from the same data source that's used for PCR total and positives and antibody totals

VA has a full time series of test encounters but we started collecting it beginning of May. Need to backfill, need to decide how to backfill

State or US: Virgina

Describe the problem Starting in early May, Virginia has been reporting their tests in units of test encounters. CTP has been capturing this data in the Total Tests (PCR) column. CTP now added a column to capture tests reported in test encounters units so we are moving the data to the correct column.

Link to data source https://www.vdh.virginia.gov/coronavirus/2020/05/07/how-is-vdh-calculating-the-number-of-people-tested/

Before

1283409 1267512 1257605 1240339 1219165 1205004 1191304 1176180 1164515 1151604 1136568 1123927 1107360 1091615 1073463 1056148 1036010 1025066 1007314 987188 969145 951174 937148 917501 904423 886748 873787 858312 842124 828017 811187 801205 783516 765889 750631 735207 725327 712350 706588 696520 685657 671560 655958 642371 633705 625663 615290 603870 590404 574013 565835 555607 543186 528914 509930 499599 491884 483024 476573 466597 457623 448172 405025 395972 388480 381220 377484 368857 361519 351354 340197 328889 322568 313283 303430 296321 285273 275074 265279 256913 242292 230683 223433 216890 210965 201183 194677 188874 184118 176635 185551 180084 171239 167758 157957 150952 143220 136558 127938 127938 122778 119065

Need to amend historical total-deaths, historical confirmed-deaths and historical probable-deaths on 5/13 and prior.

Confirmed-deaths missing or incorrect on 5/13 and prior (ex. on 5/2, we recorded 454 confirmed-deaths, but historical screenshot shows 572).

Note historical screenshots for VA from COVID Tracking Project external-facing website. https://covidtracking.com/data/state/virginia#historical Confirmed and probable began to be reported in the dashboard starting with the 4/21 12:19pm screen capture (not 6am one). Here is the first one where we see confirmed and probable called out. https://covidtracking.com/screenshots/VA/VA-20200421-121931.png

As per conversation today at the publish shift, yesterday's negative numbers do not fit our definition of "Negatives (people and cases)" as the "Total people tested" was used (due to confusing language). The new procedure now uses "Unique people tested" instead.

Previous number: VA [Negatives] : 95471 (=112450 - 16979)

Should be: VA[Negatives] : 84365 (=101344 - 16979)

Virginia positiveIncrease is negative for 2020-04-05:

library(data.table) covidtracking.dt <- fread('https://covidtracking.com/api/v1/states/daily.csv') covidtracking.dt[, date := as.Date(as.character(date), '%Y%m%d')] covidtracking.dt[order(date, decreasing=FALSE)][ , .(date, state, death, deathIncrease) ][deathIncrease < 0] date state death deathIncrease 1: 2020-04-05 VA 51 -1 2: 2020-04-07 DC 22 -2 3: 2020-04-22 AR 42 -1 4: 2020-04-23 PA 1421 -201 5: 2020-04-24 KS 111 -1 6: 2020-04-24 TN 168 -2

VA had some data reporting issues on 4/21 and 4/22. Data needs to be updated to the respective 6PM screenshots.

Some issues with VA data for 4/12 and 4/11 so far as observed by the 4/13 morning data entry crew:

1. Guidelines for updating current hospitalizations have updated to include both pending and confirmed hospitalization cases (if the state reports it). However, the number from 4/12 seems to indicate that only hospitalization numbers from confirmed cases were reported. On 4/11, the number of hospitalizations was 1252, but, the next day it dropped to 751, which is closer to the number of current hospitalizations for cases only on the dashboard. Unfortunately this is only circumstantial evidence since no screenshots for the hospitalization dashboard were recorded.

2. The number of current ICU and ventilation were also added together for 4/11 and 4/12. However, ICU numbers already include those on ventilation, which means those numbers should be separate. Specifically, the numbers should be: 4/11 Curr. ICU: 426 | Curr. Vent: 283 4/12 Curr. ICU: 440 | Curr. Vent: 294 These numbers were inferred from the formula on the spreadsheet.

Additionally, is there any way we can add screenshots for this page (https://www.vhha.com/communications/virginia-hospital-covid-19-data-dashboard/)? We currently take hospitalization data from it.

Let me know if there are any issues 👍