Tuesday, April 21, 2020

Clarke, Oconee And Barrow Counties All Add COVID-19 Cases On Monday; NE Health District Adds 24 Cases; Oglethorpe Adds 1 Death

***State Has Large Increases In Cases And Deaths***

Clarke County added four new COVID-19 cases on Monday, Oconee County added three, and Barrow County added nine, contributing to an increase of 24 cases over a day earlier in the Northeast Health District of the Department of Public Health.

The increase of 24 is lower than the 36 cases added on Tuesday of last week but higher than the 13 cases added on Monday. Only Morgan and Elbert counties in the 10-county district did not add cases.

The rolling average for the number of cases added in the Northeast Health District with the noon Tuesday Daily Status Report of the Department of Public Health dropped from 23.9 to 22.1 as a result of the decline from Tuesday a week ago.

The Northeast Health District added one death, a 94-year-old Oglethorpe man without known underlying conditions. That brings the number of deaths attributed to COVID-19 in the Daily Status Reports to 25 in the Northeast Health District, up from 24 on Monday.

The state added 934 new confirmed COVID-19 cases in the noon Daily Status Report, 66 deaths, 229 hospitalizations, and 3,812 tests.

That brings the total confirmed cases to 19,881, the number of deaths attributable to Novel Coronavirus to 799, the number of hospitalizations to 3,779, and the number of tests to 88,140.

A simply plotting of the number of cases, deaths, and hospitalizations shows continued growth in the disease in the state.

The rolling averages for number of added confirmed cases and deaths does suggest a possible flattening of confirmed cases but not of deaths.

The noon Daily Status Report for Tuesday includes new charts showing seven-day averages for confirmed cases and deaths that plot dramatic drops in recent days in both cases and deaths, but the raw data released by the Department of Public Health show obvious errors in the computation behind those graphs.

Charts

Chart 1 below plots the cumulative number of confirmed COVID-19 cases across time for the 10-county Northeast Health District (blue line, right-hand vertical axis) and for Oconee and Clarke counties (green and orange lines, respectively, left-hand vertical axis.)

The table at the bottom includes the data for all 10 of the counties beginning with the discovery of the first case in the region (in Clarke County) on March 14.

The number of deaths for each of the 10 counties is listed in the box at the top, left of the chart.

The data have been updated to include the numbers included in the noon Tuesday Daily Status Report.

Chart 1 (Click To Enlarge)

Chart 2 below shows the actual number of new reported confirmed COVID-19 cases across time, with the blue line showing the actual count and the orange line showing the rolling or moving seven-day average.

The lines have been updated to include the numbers included in the noon Tuesday Daily Status Report.

Chart 2 (Click To Enlarge)

Chart 3 below summarizes the cumulative data from the state of Georgia on four indicators: number of confirmed COVID-19 cases, number of deaths attributed to COVID-19, number of tests for the disease, and number of hospitalizations for treatment of the disease.

The charts begin on March 2, when the first cases were discovered in Atlanta. Measures of tests and hospitalizations are included from the point at which the Daily Status Report provided those data.

Each of the four charts has been updated to include the numbers included in the noon Tuesday Daily Status Report.

Chart 3 (Click To Enlarge)

Chart 4 below provides the actual number of confirmed COVID-19 cases and deaths attributable to the disease going back to March 2 for the state of Georgia. Those data are plotted with the blue lines.

The orange lines are based on computation of the rolling or moving seven-day averages for confirmed cases and deaths.

The two charts have been updated to include the numbers included in the noon Tuesday Daily Status Report.

Chart 4 (Click To Enlarge)

The fifth chart below shows the computation behind the rolling average for a single day, April 20, the final data point shown in the two new charts in the noon Daily Status Report on Tuesday.

I have listed the total number of cases and total number of deaths for the eight days leading up to and including April 20. These numbers are taken from the Daily Status Reports at noon of each day.

It is not possible to go back and check these daily reports online because the Department of Public Health overwrites the old Daily Status Report with a new one each day.

I have asked for the raw data from the Department of Public Health but have never received them.

When I realized that the old Daily Status Reports were being replaced with the new ones, I cut and pasted those reports into a word processing document, and I have a complete record going back to nearly the beginning of the Daily Status Reports.

I use that procedure to create the charts I present in this post each day. The data for April 20 shown in the table below are the data for that date shown in Chart 4.

Chart 5 (Click To Enlarge)

I began computing rolling or moving averages when I realized the bias in the Daily Status Report based on day of reporting. The rolling averages are based on seven days of data, as shown in the final chart.

It is simply impossible to have a rolling average of new cases of 428.1 when no number that low appears in the list for the seven cases for which it is supposed to be the average.

Even sighting down the lists of deaths it is easy to see how unlikely it is that the rolling average would be 23.6. It is simple mathematics.

The charts in which these data are presented in the noon Daily Status Report for Tuesday show other oddities.

The first confirmed case of COVID-19 in the state was on March 2, but the chart shows a rolling average going back to Feb. 2. The first death was reported on March 12, but rolling averages are reported back to March 5.

6 comments:

Rosemary Woodel said...

I am so glad you are doing this. They'll probably get it right by October.

amanda said...

covidtracking.com started taking screenshots of the DPH page in mid-March, if it helps: https://covidtracking.com/data/state/georgia

Emily Grace said...

This is absolutely fantastic. Will you keep updating this?

Lee Becker said...

Emily,
I have been doing daily updates since the Daily Status Reports were first issued last month.
I plan to continue to do these reports.
Thanks.
Lee

Lee Becker said...

Amanda,
I have used covidtracking.com. The screen shots taken initially were only of the first page of the reports. I appreciate your sending me back to the site, because I see now that the screen shots are of the entire report. I have all of these materials except for the 7 p.m. Easter Sunday report, which I missed, I am happy to be able to get it from this source.
Thank you for your suggestion.
Lee

IPingUPing said...

Someone just pointed me to your post. I am seeing similar inconsistencies in the data and just posted some analysis to Reddit (https://old.reddit.com/r/CoronavirusGA/comments/g64pdw/analysis_there_are_three_kinds_of_lies_lies/). Glad to share data if you need it.