Region Notes

On November 3, 2020, Georgia started reporting antigen positive cases in their COVID-19 Daily Status report, necessitating changes so we could include them in the Case counts. JHU includes antigen positive cases as probables. While GA started reporting out on antigen cases, it appears that GA's daily status report only includes the molecular test-confirmed cases in their cumulative case counts.

On November 9, 2020, we modified the case data for Puerto Rico to line up with the Commonwealth's revised 'probables' reporting. Previous reporting included serology results in the probables field reported on their dashboard. On Novermber 7th, reporting changed to define probables as anitgen testing-backed results. Historical data has been corrected through March. This modification reduces the overall the case count for Puerto Rico; the cumulative count dropped by approximately 38k on the 9th.

On October 30, 2020, our data source for test data, the Covid Tracking Project, changed its definition of the contents of its totalTestResults field for Delaware. We use this data for our positivity calculation denominator. This will will likely reduce the state's positivity percentage because people are deduplicated people less frequently, resulting in a bigger denominator.

On October 30, 2020, our data source for test data, the Covid Tracking Project, changed its definition of the contents of its totalTestResults field for Nebraska. We use this data for our positivity calculation denominator. This will will likely reduce the state's positivity percentage because people are deduplicated people less frequently, resulting in a bigger denominator.

On October 30, 2020, our data source for test data, the Covid Tracking Project, changed its definition of the contents of its totalTestResults field for Nevada. We use this data for our positivity calculation denominator. This will will likely reduce the state's positivity percentage because people are deduplicated people less frequently, resulting in a bigger denominator.

On October 30, 2020, our data source for test data, the Covid Tracking Project, changed its definition of the contents of its totalTestResults field for Utah. We use this data for our positivity calculation denominator. This will will likely reduce the state's positivity percentage because people are deduplicated people less frequently, resulting in a bigger denominator.

On October 30, 2020, our data source for test data, the Covid Tracking Project, changed its definition of the contents of its totalTestResults field for Vermont. We use this data for our positivity calculation denominator. This will will likely reduce the state's positivity percentage because people are deduplicated people less frequently, resulting in a bigger denominator.

On October 30, 2020, our data source for test data, the Covid Tracking Project, changed its definition of the contents of its totalTestResults field for Wisconsin. We use this data for our positivity calculation denominator. This will will likely reduce the state's positivity percentage because people are deduplicated people less frequently, resulting in a bigger denominator.

Our data source for test data, the Covid Tracking Project, changed its definition of the contents of its totalTestResults field for Florida. We use this data for our positivity calculation denominator. This will will likely reduce the state's positivity percentage because people are deduplicated people less frequently, resulting in a bigger denominator.

Previously, the totalTestResults field included Positive Cases + Negative Test Results. Now, the field includes Test Encounters, which are defined as the total number of people tested per day via PCR testing. The count for this metric is incremented up by one for each day on which an individual person is tested, no matter how many specimens are collected from that person on that day. If an individual person is tested twice a day on three different days, this count will increment up by three.

Recently, we have fielded inquiries about inconsistencies between JHU's and Florida's positivity calculation. There are several potential factors. Per Florida’s data definitions, there is an inconsistency in how the tests are deduplicated. For total cumulative negatives, all tests are deduplicated so people only show up once as either positive or negative. For weekly negatives, which is used in Florida’s weekly positivity calculation, tests are deduplicated for people tested that same week, allowing people to duplicate from week to week.

Also, Florida calculates positivity only using Florida Residents, while we use the total number of people tested in Florida. Ultimately, it is difficult to pinpoint the discrepancy because Florida does not make all of the variables that feeds into these calculations available, nor does the data definition provide the calculation for its Positivity Rate of New Cases by Week.

We detected an anomaly in our data source's reported total tests for Iowa on October 16. As a result, we have excluded that date's data from our 7-day rolling averages.

On October 15, 2020, Nevada switched from reporting tests in specimens to reporting tests in test encounters, and also stopped reporting negative tests (people). This caused the negative test numbers in our source data to freeze. It has temporarily been corrected in the source data, and will alter the percent positivity for October 15-20 the next time the site is updated.