COVID Numbers Mystery

Here's the mystery:

Reported COVID deaths in the US have been declining nicely for 4 weeks now.

(Image by CDC)   Details   DMCA

Consider this: before you can die from COVID, you have to contract the disease. In fact, it takes about 3 weeks on average to kill a patient. So we'd expect that the number of people coming down with COVID would have peaked 3 weeks earlier than the deaths peaked. That would have been the last week of March.

But here we are in May, and the number of reported cases is first showing signs of coming down, and very slowly at that:

(Image by World In Data)   Details   DMCA

If we look at this graph, the number of cases was still rising rapidly the last week in March, when we expected infection rates should have started declining.

What sense can we make of this? One interpretation is that doctors are learning much better how to treat the disease, so even though more cases keep coming in, they are preventing more of them from dying.

Another possibility is that the virus is mutating and becoming less lethal. This is the expected course for most viruses. Over time, they tend to become more infectious but less damaging to the host. But this usually takes more than a few weeks to happen.

Or it may be that the numbers are not being reported accurately. Hospitals can claim a 20% bonus on their billings (for doing exactly the same procedures) if they report that a patient has COVID. Perhaps they've started over-reporting COVID cases.

This article argues that the numbers are being goosed, maybe in order to scare us.

Here's another reason to look under the hood and raise questions. (I beg your patience with some meathematical reasoning.) Every COVID patient was, presumably, infected by an earlier COVID patient. If each existing patient infects more than one new patient, then the numbers rise faster and faster. Technically, we expect to see a rising exponential curve. On the other hand, if each existing patient infects less than one new patient, then the numbers fall rapidly toward zero. This is a falling exponential curve.

The only way the curve could stay flat for long periods of time would be if the number of new people infected were to be exactly one, no more and no less. We'd expect this to be a rare and unstable situation.

Readers, I invite your comments and theories about what's going on. As for me, I suspect we're not being told the truth.