West Virginia

State or US: WV

Describe the problem We can the history of West Virginia's Negative (people or cases). The number mixes units and is no longer needed since totalTestResults are no longer calculated via positive+negative.

Link to data source N/A deleting data

State or US: WV

Describe the problem We can the history of West Virginia's Negative (people or cases). The number mixes units and is no longer needed since totalTestResults are no longer calculated via positive+negative.

Link to data source N/A deleting data

State or US: WV

Describe the problem Backfill total tests (PCR) back to march from data scraped from WV's dashboard

Link to data source https://github.com/COVID19Tracking/covid19-datafetcher/blob/data/wv_lab_tests.csv

State or US: WV

Describe the problem Backfill total tests (PCR) back to march from data scraped from WV's dashboard

Link to data source https://github.com/COVID19Tracking/covid19-datafetcher/blob/data/wv_lab_tests.csv

State or US: West Virginia

Describe the problem CTP added the confirmed cases column on 4/29 and the probables column on 7/28. West Virginia started separating confirmed and probable cases on 5/22. Before 5/22, all cases reported on WV dashboard were confirmed.

CTP did not capture confirmed & probable between 5/22 and 6/4 Backfill: No because we do not have a source because screenshots do not capture hover.

Action Plan

1. Backfill probables from 6/5 to 7/27 by using positive - confirmed.

2. Fill the gap in the timeseries by carrying over confirmed from 5/22 to 6/4, because we know there were a minimum of that many confirmed. Do not touch probables because we don't know what was going on with those (we know there were some from Slack, but not how many because we cannot access the screenshots)

3. Backfill confirmed from 3/18/2020 to 4/29/2020 from the Positives field because we know they were the same.

State: West Virginia

Describe the problem: The state confirmed that WV has reported Total Tests in specimens on 7/7, so we are moving the Total Test data to the specimens metric from the people metric. Additionally, we missed recording the Total Test Numbers from 5/2-5/12, however we have screenshot of the WV dashboard from those days, and used them to fill in the missing data points.

Transposition issue on 7/17 recovery count from WV, resulting in red cell on 7/18

State or US: West Virginia

Describe the problem We've been historically recording "Confirmatory Lab Test" numbers under "Total PCR Tests (People)." However, from 7/7 outreach, the figures under WV "Lab Test Trends" tab refer to specimens tested, NOT people.

Link to data source https://dhhr.wv.gov/COVID-19/Pages/default.aspx

State or US: West Virginia

Describe the problem

1. Historical "Positive Cases (PCR)" goes back until 4/29. There's a gap from 5/22 - 6/4 (https://github.com/COVID19Tracking/issues/issues/555) that we may be able to fill now.

2. States Daily antibody test values go back until 5/23, but our long formulas sheet dates back to 4/12. The WV website has hover-over line graphs that date back to 3/17 for cases and 3/11 for lab tests.

3. Under the "Positive Case Trends" tab, cumulative historical cases on the WV website combine probable and confirmed cases, but in the graph underneath, historical daily cases are separated by probable and confirmed cases. Since WV updates data multiple times every day and backfills their own data consistently, current CTP historical values for cases and testing do not match with the hover-over line graphs on the WV website.

Link to data source Dashboard: https://dhhr.wv.gov/COVID-19/Pages/default.aspx Daily Press Releases (at 10AM and 5PM): https://dhhr.wv.gov/News/2020/Pages/default.aspx

State or US: WV

Describe the problem

Cases

NOTE: Between 5/23 and 6/4 the values for positive cases (PCR) are missing

Link to data source Discovered while working on #550

Death values are historically recorded in both the "Deaths" and "Deaths (Confirmed)" columns for WV. However, WV’s death values are unclear about what they represent, so they should only be recorded in the main "Deaths" field.

5/12: CTP started reporting same values for deaths (confirmed) and deaths.

https://covid-tracking.slack.com/archives/C012C4BEAR4/p1592489962311800

New dashboard has antibody tests in addition to diagnosis tests: https://app.powerbigov.us/view?r=eyJrIjoiNzhjMzUyZGItN2ZjMi00ZDQ2LWExM2MtMzE3ZTEzNTE2MzdiIiwidCI6IjhhMjZjZjAyLTQzNGEtNDMxZS04Y2FkLTdlYWVmOTdlZjQ4NCJ9&pageName=ReportSection9f5bc3be6e9ebda92968

Here's the dated CSV from the dashboard:

date,lab_results,antibody
"2020-03-11",1,0
...
"2020-04-05",711,1
...
"2020-05-18",2154,110
"2020-05-19",1793,195
"2020-05-20",3223,232
"2020-05-21",1446,313
"2020-05-22",2058,239
"2020-05-23",1092,56

These are daily numbers, will add a cumulative sum.

If, PCR testing needs to be backfilled as well, this table covers both.

wv_antibody.res.zip

From the external support message: ...there also seem to be discrepancies concerning cumulative positive tests. When I compare the amount of change between your totals for 5/11 and 5/12, I get 5 cases. But when I hover over the "Cumulative Number of Laboratory Confirmed Cases" on the DHHR site, the "cases" number for 5/12 (which I take to be "new cases being reported as of 5/12) is 14.

Note from GG on the support team: This is a bit too deep in the weeds for me, can you help explain? Thx!

Historical data should be updated in our system

https://dhhr.wv.gov/COVID-19/Pages/default.aspx

The data for West Virginia for cumulative hospitalization, cumulative ICU, and cumulative ventilator numbers come from multiplying the posted percentages by the total number of cases. But the percentages are based on "completed investigations", not total cases, so this data is unreliable and should be removed.

(Current hospitalization, ICU, and ventilator numbers come from a separate page that is up-to-date.)