https://abs.twimg.com/emoji/v2/... draggable="false" alt="🧪" title="Test tube" aria-label="Emoji: Test tube"> CASE POSITIVITY IN FL https://abs.twimg.com/emoji/v2/... draggable="false" alt="🧪" title="Test tube" aria-label="Emoji: Test tube">

I wanted to focus on 2 Qs I& #39;ve gotten recently

https://abs.twimg.com/emoji/v2/... draggable="false" alt="📌" title="Reißzwecke" aria-label="Emoji: Reißzwecke">The diff in the 2 measures reported by FL result from excluding previous + for 1 of the measures. What % are excluded?

https://abs.twimg.com/emoji/v2/... draggable="false" alt="📌" title="Reißzwecke" aria-label="Emoji: Reißzwecke">The weekly report removes repeat testers w/i the week. What& #39;s the impact?

1/
Quick review:

MEASURE 1: case positivity
- numerator: unique people who test +
- denominator: unique people tested

MEASURE 2: % positivity for new cases
- same calculation but exclude (from numerator and denominator) anyone who tested + prior to the date being assessed.

2/
To understand diff in the measures, I focused on the % excluded from the numerator and denominator of MEASURE 2.

I used data from 9/27 thru 10/3, as listed on the 10/4 state and county reports.

3/
The % removed from the NUMERATOR (+ people who previously tested +) was 29.6% for the week

- daily % removed varied from 26.7% to 32.4%

The % removed from the DENOMINATOR (+/- people who previously tested +) was 8.1% for the week

- daily % removed varied from 7.7% to 8.8%

4/
So, a higher % of people who tested + are being removed from the numerator (about 1 in 3) than the % of people tested from the denominator (about 1 in 12).

That& #39;s why the % positivity for new cases is LOWER than the case positivity.

What about the WEEKLY TESTING REPORT? https://abs.twimg.com/emoji/v2/... draggable="false" alt="👇" title="Rückhand Zeigefinger nach unten" aria-label="Emoji: Rückhand Zeigefinger nach unten">

5/
Around 60-65% of people tested daily in FL have been tested previously.

So, we tend to see a lot of repeat testers in the daily positivity measures (both of them).

Florida releases ONE DATASET weekly that de-duplicates people tested during a given week.

6/
This report gives numerator and denominator for the "% positivity for new cases" (MEASURE #2).

If one were to ADD the daily numerators and denominators from the week, they would calculate:
- 16,191 / 365,705 = 4.43%

The weekly report calculates:
- 16,191 / 347,123 = 4.66%

7/
The NUMERATORS are the same, which makes sense.

As soon as a person tests +, they are not included in any subsequent day& #39;s calculations (for this positivity measure).

In that sense, the numerator of the "% positivity for new cases" is always de-duplicated, even daily.

8/
But the DENOMINATORS are different since DOH de-duplicates:
- people who tested negative multiple days in the week
- people who tested negative and then positive during the week

How much did de-duplication remove?
- 18,582 of 365,705 (or 5.1%) of the denominator

9/
So, would seem that, if this week is representative of most weeks recently:
- about 1 in 20 who are tested in FL are being tested multiple times WITHIN A 7-DAY PERIOD

Also, the % positivity from the weekly rpt tends to be https://abs.twimg.com/emoji/v2/... draggable="false" alt="⬆️" title="Pfeil nach oben" aria-label="Emoji: Pfeil nach oben"> than the daily avg due to removal of repeat neg.

10/
You can follow @JasonSalemi.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: