Differences in projected numbers…
A few folks have asked why the numbers I laid out in this projection are considerably under our current case count. By my projections, we should have been around 5,000 cases today. Instead, we are over 11,000. I projected around 110 deaths. Instead we are around 160.
Is that a reason for concern? Should we be alarmed at having nearly double the number of cases? Should we be alarmed at having 50% more deaths?
The answer to both is “not necessarily.
My original projections were based on a growth rate of 20%. At that time, the cases had been doubling at a rate of every 4–5 days. The actual growth was bouncing between 1.15 and 1.25. I split the difference and assumed 1.2.
The actual growth rate, particularly over the past few days, has been dramatically higher. A subsequent post I shared on Facebook at 1 AM on the 16th noted we were likely to see the first day with 1,000 plus cases that day. I also suggested we would be over 2,000 new cases per day by the end of the weekend. We have dramatically exceeded expectations and have surpassed 2,000 new cases for each of the last two days. our daily growth rate has been closer to 40–45% each day.
Still, there is reason for cautious optimism.
There is good reason to believe that the numbers are considerably higher simply because we are testing more. If drive through testing and more lab capacity are able to catch more cases early on, that’s a good thing. That is especially true if we can catch people who are mildly symptomatic or asymptomatic. Catching the infection before symptoms go nuclear is a good thing. Those people can be treated or isolated.
For that reason, I am incredibly disappointed in the daily Presidential briefings. They have repeatedly discouraged people from getting tested unless they are symptomatic. That’s dangerous advice. This disease appears to be most contagious in the early phase. Identifying people early and taking them out of circulation is important. I understand the reason for suggesting that people not seek testing, given we still have limited testing capability. However, it creates a sense in people that they are not in great danger of passing the disease. They may be tempted to go about their daily activities rather that self-isolating.
What about the fatality rate?
This is where things get more tricky. The good news, looking just at the case count versus number of deaths is the CFR seems to be considerably lower than the 2.4 in my assumptions. With 161 deaths and 11,329 cases, our current CFR based on known cases is around 1.4.
The bad news is it is unclear if that is anywhere near accurate. If we are catching more cases early, we have no idea if they will eventually experience the onset of acute respiratory problems. While we may be pleased that the CFR appears low, we can’t extrapolate the current case count to deaths for some time. if we based the CFR solely on resolved cases, for instance, we have 161 deaths and 108 resolved cases. That would suggest a CFR of 60%, which is wildly outside the range of likely eventual outcome.
The reality is it may be a long time before we have a solid read on the CFR. Even estimates of the swine flu epidemic of 2009 (H1N1/09) are wildly variable ten years later. Somewhere between 700 million and 1.4 billion may have been infected with a death toll between 150,000 and 575,000 worldwide. Covid is estimated to be twice as transmissible and may be 10 or 20 times more lethal.