Q&A of the Day – How accurate is COVID-19 reporting?
Each day I’ll feature a listener question that’s been submitted by one of these methods.
Email: brianmudd@iheartmedia.com
Twitter: @brianmuddradio
Facebook: Brian Mudd https://www.facebook.com/brian.mudd1
Today’s entry: Brian, you previously referenced the amount of deaths in excess of the average rate. Can you compare that to the present numbers. If there’s not a corresponding difference as before that would lead to a real suspicion that COVID death counting is being manipulated. Thanks for keeping us informed
Bottom Line: We’ve learned a lot about what we can and can’t rely on from the Florida Department of Health’s daily reporting. Based on a combination of data from the CDC, FAU’s COVID-19 tracker and the Florida Department of Health, we know the following:
- The positivity rate has been artificially high due to several labs only reporting positive cases
- Daily reported “new” cases contain positive tests which are often weeks old
- Reported deaths could be up to eight weeks old
This week brought about the most absurd example of this yet when the lab Niznik, which tests in Broward, Miami Dade and Palm Beach Counties dumped 46 days of testing at once inflating the daily totals by thousands of cases and once again showing the extensive flaws with the reporting of data. We should all be skeptical of what’s real and what isn’t at this point. This has become critical as a matter of not just public health but public policy as local governments and school districts are refusing to allow for the reopening of our economy and our schools until targets (some defined like target positivity rates, some not so much, like trends) are met. That’s the biggest problem in the here and now. To the point of your note however, what about the data over the long run? Is it proving to be accurate? The answer largely remains yes.
Throughout the pandemic I’ve periodically provided the CDC’s “excess death” data as a way to fact check COVID-19 reporting. CDC director Robert Redfield validated my use of excess death data for this purpose. For those who might not be familiar... The CDC’s excess death metric measures a five-year average for total deaths and adjusts for population changes. If no one had been diagnosed with COVID-19 and if no Floridian was ever listed as a COVID death, the stats would be the same. According to the CDC, data is incomplete for at least eight weeks as there’s a lag from deaths occurring and all death certificates being issued. This means the week ended June 20th is the most recent which contains mostly complete information for Florida. Let’s look at what we’ve learned using this CDC data.
- The first week Florida experienced excess deaths was April 4th
- Florida’s experienced excess deaths for every week with complete data since
Between April 4th and June 20th, Florida was expected to have 48,610 deaths. The actual total was 52,760, resulting in 4,150 excess deaths. On June 20th, the total number of coronavirus deaths reported by the Florida Department of Health was 3,147. This data illustrates Florida’s official tallies underreported related deaths by 1,003 deaths, representing approximately 2% of the total deaths during that window of time. It stands to reason that not all people ill with COVID-19 have been tested and therefore not all related deaths diagnosed either resulting in the slight underreporting of COVID-19 related deaths.
What this illustrates is that over the long run, once all reporting has been accounted for, the tallies have been highly accurate. The problem is the obvious. We have reasonably accurate data to review from the Florida Department of Health from June 20th. How are we supposed to make real time decisions about our economy, schools, etc. - by using data which won’t be mostly reliable until October 9th?
In conclusion, at least as of June 20th, there’s no evidence of manipulation of COVID-19 reporting in Florida. It remained slightly understated and consistent with what I’ve studied from the onset of the pandemic. What is being manipulated is public policy based on metrics that are unquantifiable in real-time based on current reporting.