Day 42: These DPH Numbers are Just Crazy Now or Suddenly Honest or Miraculous?

First things first, there is no way that the numbers from DPH are on the mark today. Hospital admissions and deaths falling to single digits and the total number of infections rising by only 185? The number rose by over 1,000 the day before! Since nobody in Atlanta is taking these numbers seriously, why should we believe ANY of these numbers from the beginning. Nonfeasance like this makes the memes of the “tinfoil hat”-wearing, government collusion crowd look more and more compelling. Naturally, I can’t find any to post right now because the censors have taken them all down. One recent one shows a scarecrow half under a steam roller with the caption that the “victim” was killed by COVID-19. Another shows the aftermath of a fiery car crash with a similar caption. These memes imply that deaths not caused by COVID-19 are being attributed to the virus simply to justify the government’s prior martial law actions. Depending upon the jurisdiction, I can believe that some bureaucrats would be enticed to attribute unknown causes of death to COVID-19 (no doubt Federal monies will be doled out on a per-victim basis). However, I have a hard time believing that medical professionals would be enticed to do so. The more likely answer is that the data is moving faster than the collectors. Like Lucy Ball trying to put chocolates into boxes coming at her from the ever speeding assembly line, those counting COVID-19 simply can’t count fast enough. And in life, everyone has a boss and everyone has an agenda whether they know it or not, and they are trying to make everyone happy.

What is frustrating, is that we citizens just want to know if this pandemic is getting better or not. For us in North Georgia, things really look about the same. No bodies in the streets. No riots. No looting. The cops are still on patrol, and there are no bands of roving bandits setting up roadblocks. So you can forgive people if they need a bit more proof that there actually is (or was) a problem while we sit at home on house arrest. The tool of politicians is rhetoric, and rhetoric is just a bunch of wind coming out of someone’s mouth designed to persuade the inattentive listener. We want to see results, and the way to show results are to provide the data to back it up, not just tell me that “they” say the numbers are getting better. In a time where the word transparency is used and overused, this process has been–and continues to be–rather opaque. The Governor makes one report, the President says too soon, and a local hospital cries out that they are about to be overwhelmed this week, yet only 6 people were added to the hospital roles yesterday statewide! They can’t all be right at the same time, can they?

Even if you take the numbers at face value, the infection total released daily is misleading. Since it is Day 41 of the quarantine, the “total” number from DPH can’t be accurate. According to Johns Hopkins, if a patient is going to recover from COVID-19, then the patient will do so in no more than 21 days (2-14 days to show symptoms, followed by 7 days to run the course of the virus and recover). Anyone with a positive test from more than 21 days ago is either recovered or deceased. Therefore, it means that the total confirmed cases definitely does not equal the number of people who are currently sick. What we really want to know is how many sick people there are, and if that number is going up still, staying the same, or going down. There absolutely is no way to determine this from the DPH numbers as we are provided. That also means that any graph of total infected is absolutely misleading and there is no real way to correct it, so I’m abandoning it.

Let me give an example of what I’m talking about. DPH reports the total number of COVID-19 tests is 24,401 as of Sunday, April 26, 2020. Twenty-one days ago (the earliest possible point of infection), the total positive tests were 6,647, and as of today 912 people have died. Subtracting those who must have recovered and those who have died leaves 17,754 people still in some stage of infection today. If we use a more likely time period like 2 weeks ago, then the number of people in some stage of infection today goes down to 10,037. That number is less than half of the total number reported. But this is all just guesswork on my part. The people who have those numbers aren’t sharing them with us, so we don’t know.

As to the other metrics that rely upon the total infected, I’m abandoning them as well. Since the DPH numbers seem to be ridiculous anyhow, any quantitative analysis is foolish. When you look at any number analysis day-over-day, it appears to be Winnie-the-Pooh up, down, up, down. Therefore, one way to correct for that is to use averages to settle things down in one form or another, so I turn to that process now. Fig 1, below, shows the daily number of infections. Fig 2, below, shows the change in the average number of cases for the whole pandemic. Yesterday it was up sharply while today it is down sharply. More Winnie-the-Pooh data. Far from settling the data down, this graph is very sensitive to daily change. I’m not epidemiologist, but this up, down, up, down does not look like the result of a natural phenomenon to me. Fig 3, below, shows the number of reported deaths. More Up, Down, with an unbelievably low number today of 5. The experts were off by +27 and I was off by +26. If true, this is great news. But it is more likely that the reported data is off. If the reported data is off, then any analysis arising from this data is going to be off.

Now I feel like Lucy myself. I just completed this post for Sunday, and Monday’s numbers just came out. Maybe I should say that I feel like Charlie Brown and the other Lucy has just pulled the ball up.