Arkansas

State: AR

Dates affected: 02/25/21

Describe the issue: On February 25, 2021, Arkansas updated their Ever hospitalized and Ever on ventilator after the time of our daily update. We are patching it from the graphic on their official Arkansas Department of Health twitter.

Links: https://twitter.com/ADHPIO/status/1365102980496830466/photo/1

State: AR

Dates affected: 02/25/21

Describe the issue: On February 25, 2021, Arkansas updated their Ever hospitalized and Ever on ventilator after the time of our daily update. We are patching it from the graphic on their official Arkansas Department of Health twitter.

Links: https://twitter.com/ADHPIO/status/1365102980496830466/photo/1

State: AR

Dates affected: 01/13/21

Describe the issue: On January 13, 2021, Arkansas updated their COVID-19 dashboard after the time of our daily update.

State: AR

Dates affected: 01/13/21

Describe the issue: On January 13, 2021, Arkansas updated their COVID-19 dashboard after the time of our daily update.

State: Arkansas

Issue: On September 18, 2020 Arkansas stopped reporting Total/Positive/Negative Antigen tests (specimens) and began reporting Total/Positive/Negative Antigen tests(people). We carried over the last available value, but the metric has not returned, so the value is out of date and potentially confusing. We will be keeping the time series as provided by the state, but removing the carried over values from September 18, 2020, forward.

State: Arkansas

Issue: On September 18, 2020 Arkansas stopped reporting Total/Positive/Negative Antigen tests (specimens) and began reporting Total/Positive/Negative Antigen tests(people). We carried over the last available value, but the metric has not returned, so the value is out of date and potentially confusing. We will be keeping the time series as provided by the state, but removing the carried over values from September 18, 2020, forward.

State: AR

Dates affected: 01/10/22

Describe the issue: On January 10, 2021, Arkansas updated their COVID-19 dashboard after the time of our daily update.

State: AR

Dates affected: 01/10/22

Describe the issue: On January 10, 2021, Arkansas updated their COVID-19 dashboard after the time of our daily update.

State: Arkansas

Issue: On January 9, 2021, Arkansas did not update until after our publish time. Patching metrics from updated dashboard and state page.

Sources: https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-20210109-230914.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-secondary-20210109-230930.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-tertiary-20210109-231010.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-quaternary-20210109-231049.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-quinary-20210109-231123.png

State: Arkansas

Issue: On January 9, 2021, Arkansas did not update until after our publish time. Patching metrics from updated dashboard and state page.

Sources: https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-20210109-230914.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-secondary-20210109-230930.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-tertiary-20210109-231010.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-quaternary-20210109-231049.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-quinary-20210109-231123.png

State: AR

Dates affected: 01/06/21

Describe the issue: On January 6, 2021, Arkansas updated their COVID-19 data dashboard after the time of our daily update.

State: AR

Dates affected: 01/06/21

Describe the issue: On January 6, 2021, Arkansas updated their COVID-19 data dashboard after the time of our daily update.

State: AR

Dates affected: 1/3

Describe the issue: On January 3, 2021, Arkansas updated their data dashboard after the time of our daily update.

State: Arkansas

Issue: Arkansas's dashboard did not update before our publish time. Patching in all metrics except cumulative hospitalizations and on ventilators, which we updated from the state Covid-19 page.

Source: dashboard (will update with screenshots when available)

State: AR

Dates affected: 01/01

Describe the issue: On January 1, 2021, Arkansas updated their COVID-19 data dashboard after our daily update.

State: AR

Dates affected: 12/24

Describe the issue: On December 24, 2020, Arkansas did not update their COVID-19 dashboard by the time of our daily update.

State: AR

Dates affected: 12/09

Describe the issue: On December 9, 2020, Arkansas updated their COVID-19 dashboard after the time of our daily update.

State: AR

Dates affected: 11/28

Describe the issue: On November 28, 2020, Arkansas did not update their Total antigen tests (people), Positive antigen tests (people), Negative antigen tests (people), Total PCR tests (specimens), Negative PCR tests (specimens), Positive PCR tests (people), Probable cases, Cases (confirmed + probable), Negative PCR tests (people), Recovered, Deaths (confirmed + probable), Deaths (confirmed), or Deaths (probable) before the time of our daily update.

State: AR

Dates affected: 10/7

Describe the issue: On 10/7 the AR dash was not updated by the time we published. Cur. hosp and cur. icu were used from the page graphic, however we have decided that going forward we will follow source notes and carry over these metrics if necessary. We should patch 10/7 cur. hosp and cur. icu from the dash in addition to other values which we normally source from the dash.

Links: https://covid-tracking.slack.com/archives/C01B0EGJ3J9/p1602164560246400

State or US: Arkansas

Describe the problem On 9/20, we updated recoveries using the value reported on the dashboard, not the graphic on the state page. This caused the recoveries time series to increase and then drop. Backfilling with the graphic value in order to have a more consistent time series while we investigate which source is the best/most consistent one.

Changing from 67853 to 66397

Link to data source

State or US: Arkansas

Describe the problem There were issues with broken/unupdated/lagged hospitalization and case/test values on 9/18 and 9/19. Current screen captures do not record testing data from those days.

Link to data source 9/18: https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-20200920-000419.png 9/19: https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-20200920-060438.png https://covid-tracking-project-data.s3.us-east-1.amazonaws.com/state_screenshots/AR/AR-secondary-20200920-120448.png

State or US: AR

Describe the problem AR updated its dashboard with new sources, we need to change our capture and patch our handling on 9/13 (and 9/12 for hospitalizations)

Negative Tests (PCR) and Negative (People or Cases) - PCR negatives from dashboard, no backfill because incomplete timeseries, but start using the numbers, patch yesterday and write a public note Total Tests (PCR) - PCR negatives + PCR positives from dashboard, no backfill because incomplete time-series, but start using the numbers, patch yesterday and write a public note Positive Tests (PCR) - leave the same (empty); the PCR number is cases Positive, Confirmed Cases, Probable Cases - start getting them from the dashboard and patch yesterday Antigens - submit outreach about the gap between their dashboard and the PDF, keep collecting yesterday’s Deaths for Deaths (confirmed and probable) - we will patch yesterday’s and start using the breakdown Cur hospitalizations - patch 9/12 from page graphic, add warning note Cur vent - patch 9/12 from page graphic, add warning note Cur ICU - keep as using dash Recovered - patch 9/12 and 9/13 from page graphic, add warning note

Data sources 9/13 tests 9/13 deaths

9/12 deaths SEptember 13 hosp graphics https://covidtracking.com/screenshots/AR/AR-20200913-180445.png September 12 Hosps graphics https://covidtracking.com/screenshots/AR/AR-20200912-180422.png September 12 cases https://covidtracking.com/screenshots/AR/AR-secondary-20200913-180451.png

State or US: Arkansas

Describe the problem Starting 9/3 Arkansas started providing the information about Total Antigen Tests, Positive Antigen Tests, Negative Antigen Tests and Probable Cases. This is to backfill the information from those days.

Link to data source Upload PDFs here please

State or US: [State name, or "US" if this is an issue with country-wide data] AR

Describe the problem AR has a table they have provided two days in a row with pos and neg antigen results. We should backfill and start collecting this data during pub shift (when available).

Link to data source https://covid-tracking.slack.com/archives/CUQ4MMTPD/p1599248272194300?thread_ts=1599248268.194200&cid=CUQ4MMTPD

https://covid-tracking.slack.com/files/U01489QMX99/F01A503QPRQ/total_cases_by_case_status_09042020.pdf?origin_team=TUPV6BFR9&origin_channel=Vall_threads

State or US: AR

Describe the problem AR shared a value for its total number of antigens verbally in a presser on 8/20 (10,358: see 2:33 in https://www.youtube.com/watch?v=_cqwYiMx7EY) significantly higher than what we found on the slides they post on their website a few days later on 8/25 (3,610 tests in https://www.healthy.arkansas.gov/images/uploads/pdf/briefing_slides_8_25.pdf). We have conducted outreach to ask why this discrepancy exists and for a full time series of the metric. In the mean time we are carrying over the old value (1836) from 8/20-8/24 and using 3,610 from 8/25 onward.

Link to data source https://www.healthy.arkansas.gov/images/uploads/pdf/briefing_slides_8_25.pdf

State: Arkansas

Dates impacted: pre-7/3 (inclusive)

Issue: Prior to 7/3, CTP was capturing total tests in Total PCR Tests (People), after 7/3 we started reporting total tests in Total Tests (PCR). But, we are still reporting negative tests in people, even though negatives are reported in specimens. We think we should copy negative tests from people to Negative Tests (PCR) KEEP all values in Negative (People or Cases) and copy values to Negative Tests (PCR) to reflect the fact that negatives are reported in specimens as confirmed by outreach. We also think we should move total tests pre-7/3 to Total Tests (PCR) from Total PCR Tests (People). We should also flag that the AR total tests number is calculated by (positives cases (people) + negatives tests (specimens)). :(

Action Plan:

• [x] Start Capturing in Negative Tests (PCR)
• [x] Move Total PCR Tests (People), captured from 3/6-7/3 to Total PCR Tests (People) AR Moved Values from Total Tests PCR People to Total Tests PCR.txt

Arkansas has been reporting Confirmed-Cases since Mar 27th when they started using the arcgis dashboard. Different than other states, AR reported Confirmed-Cases only on their State website instead of the dashboard for a while. Even up to the current day, AR uses the phrase "Total Cases" in its arcgis instead of "Confirmed Cases."

Please backfill confirmed cases back through Mar 27th.

Background:

Screenshots: http://covid-tracking-project-data.s3-website.us-east-1.amazonaws.com/state_screenshots/AR/

WS2 instructions:

States' Daily:

AR data was updated after the pub shift on multiple dates causing days of no change. For example "On June 13, the 6:30pm screen shot shows 12,095 cases but the data set shows 11,547."

The dates this affects are: 5/10, 5/23, 6/2, 6/4, 6/21, 6/26, 7/5

Front conversations

We have been recording AR's Total Tests as "Total PCR Tests (people)", however we received information from outreach that they are (partly) specimens. Need to move values to column "Total Tests (PCR)."

See --- AR 6/26 SvP Update --- in Slack for details.

Death values are historically recorded in both the "Deaths" and "Deaths (Confirmed)" columns for AR for only 6/15 - 6/23. However, AR's death values are unclear about what they represent, so they should only be recorded in the main "Deaths" field.

While potentially accurate to the state record, negative cumulative values cause problems for modeling. We could not verify the source of the death values as the AR "Death Maps" do not appear in the linked screenshots.

Deaths 4/20 - 42 4/21 - 43 (+1) 4/22 - 42 (-1) 4/23 - 45 ... 5/3 - 76 5/4 - 81 (+5) 5/5 - 80 (-1) 5/6 - 85

@thaynedye @dcmoyer

AR deathIncrease is negative for 2020-04-07:

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