US Breaks 150,000 COVID deaths this week: a perspective in graphs and some simple calculations predicting a very dire August.

(sorry for the formatting, I am still getting the hang of this blog world)

So this is the overall state of things according to the database I believe is the least corrupt at this point (more on the incredible corruption in death and case reporting later). It’s 4:35 a.m. and I can’t sleep. All the dying going on might be something to do with it. And all my nurse and doctor friends in the hospitals dealing with things I can’t even imagine- war time things- battle ground things- stress, and exhaustion, and fear for their own lives, and the lives of their patients, and the lives of their families. You get numb at a certain point, except I’ve never been good at that particular thing- the numbness.

But I was going to talk to you about numbers. So let’s do that. 154 thousand deaths is a lot of deaths and unfortunately, I think we are just ramping up. It’s going to be a very gory August. We may be starting to level off in terms of daily new cases at a really shocking number if you put it in perspective, 68k new cases a day. If you recall back in March when Italy (and then New York City) hit the 30k total cases mark and we were panicking?— well now we are more than doubling that total number each and every day. It’s a little hard to imagine but the below graphic might help.

One way I think about predictions or models (more to come on the epidemiological modeling failings during this pandemic), is pretty simply using just same basic logic. People like to get fancy with R0 etc. but I have found that to be pretty useless in this case in making accurate predictions. I am planning a full out review article on this so I won’t get into all of that right now but walk with me through a simple way of thinking about things that has been very accurate in predicting for me.

We have all had experience with ratios in elementary school math- I think that is pretty relevant here since we don’t know that much about this virus and how to model its spread. We can use ratios to predict the future of what will happen based on the past behavior of the virus. You can see that the rough peak of daily new cases back in April was a little over 25k/day in the US. The current peak is ~75k. So more or less, we have 3x the cases that we had then. Things have not changed much since then in terms of relevant parameters- Trump keeps going on about testing being higher but most insurance is not covering testing for asymptomatic people still and the percent positive rate keeps going up so I’m going to ignore that for the time being. We also have no reason to believe that the virus itself has gotten any less deadly by mutation and any advances with drugs or treatment in hospitals is cancelled out by hospitals being very overwhelmed and very understaffed at this point. So, for the sake of argument- lets say apples to apples with case numbers. Now lets compare that with daily deaths– to predict the upcoming daily death rate.

So because deaths lag behind cases (due to the time it takes people to die, deaths to be reported, and certificates signed off on)– the deaths are just rising now- no where near their peak. So that takes me back to ratios– if the peak of daily cases back in April was 1/3 of what it is now then we would expect the peak daily deaths to be 3x its former peak. Its former peak was ~3k/day so I expect that in August we will see ~9k/day people dying. That’s about 300,000 people in one month 30 days x 9k/day:( I’m not sure the US understands that that’s about to happen. Schools opening is ludicrous.

Of course, there is some room for noise here and these are rough calculations. But I think my logic is sound and my predictions have been right so far in this pandemic. So I’d say buckle up, it’s going to be a rough month. The only sunshine on the horizon is that I do really think herd immunity will/is happening- ahem NY being all pale and not increasing I will get to that grand reveal later this week.