Aaron Hartman, MD (May 5, 2020) explains why recent University of Washington COVID-19 mortality projections for the U.S. rose (e.g., to ~134,000 from ~74,000 in days). He attributes the jump to changing model inputs—not model failure—primarily states' reopening (increasing transmission), revised infection-fatality estimates (U.S. data suggesting ~0.6% vs earlier higher rates), and large national weighting from hotspots (New York City). He flags specific UW-model limitations: its forecast horizon ends Aug 4, it doesn't include emerging treatments (e.g., remdesivir) or vaccines, seasonal/behavioral mitigation (masks), and local heterogeneity. Takeaways: expect a summer rise as society reopens, focus on local/state projections (UW site has locality views), and view national spikes cautiously.
Questions this video answers
About 134,000 deaths through August 4, 2020.
About 74,000 deaths.
About 2.4 to 3 million deaths (early, high-end projections).
Approximately 0.6% (U.S. estimate mentioned).
The UW model forecast window stops on August 4, 2020.
Chapters
It's great to be with everybody once again tonight. I'm going to talk briefly about projections, and this is mainly in response to a family member who texted me this question. And the question basically is, why have the reported mortality in the United States jumped to about 134,000? So for those of you don't know, today the University of Washington changed the projections for deaths in our country through August 4, and it went up to 134,000. Now, to put that in perspective, a couple days ago it was at 74,000. So in the matter of a couple days, the projected deaths in our country have more than doubled. The lowest of these projections from the University of Washington have been as low as about 64,000. And my family member asked me the question, why are we seeing more than a doubling in a day or two? Is this cause for alarm? Should we be concerned about this? Is something new happening? And the answer is no, no and no. If you've been following our videos for the last two months, and I was actually looking at this today, the very first video I put up was January 30th. This is a projection we've been kind of talking about. After isolation, we were talking about once we start going back out into culture and society, we will see a jump. We should expect to see a jump. So before I go into that, we have to talk about these models. These models, whether it was the World College model, which we were tracking before COVID 19 came into the United States, or if it was the University of Washington model, which this is referring to, models are only as good as the data it's put into them. Garbage in, you know, garbage out kind of sort of thing. I'm not saying these are bad models. We need models to figure out what's going on. But the problem with these models is they only are as good as the information put into them. So if you recall, about six weeks ago, eight weeks ago, we were projecting 2.4 to 3 million Americans were going to die from this. So why was it 2.4 million then? And it dropped to 64,000. Now it's more than doubled to 134,000. And the reason for that is what's being put into the models. The original data was expecting us to have a mortality rate from this disease anywhere from the 11% in Italy to the 5.4% in Wuhan, China. Well, we found out in our country the number is probably closer to 0.6, still about five to six times worse than the flu, but not the 5.4%. The Wuhan kind of expectations. So what happened was over the last six to eight weeks, we saw these numbers decline come down and down and down. As we saw that, hey, it's not quite as bad. Every state's not New York City, every state's not New York. Hey, look at Florida, they're not quite that bad. Hey, look at California. And what happened was the University of Washington projections came down and down and down. What's happened in this past week is that states are starting to reopen and states are starting to project to reopen in the next week or two. So what's happening is the University of Washington is now taking this into account, saying, look, as states reopen, we expect people to now spread the virus a little bit, we expect some more infections and the expectation of that is we'll see a spike again. And this is something that we've been talking about for the last couple weeks as we reenter society and culture and we should expect to see an increased incidence of infections and deaths. So this spike in the University of Washington is not anything unexpected. And I want you all to realize that. I'm not sure what the news cycle will say tomorrow or the next day or the next day. They may not even mention this, who knows? But this is to be expected. As we kind of start going back out, we will see the number of people getting infected and dying increase. That's again to be expected. There are issues with models and the very first issue with this model that I have is it stops at August 4th. So what does that mean? August 4th, everything magically disappears. August 4th, we have the cure. August 4th, we don't care anymore. That's a huge, huge flaw with this model is it's not looking at the fall, it's not looking at the winter, it's not looking at next year. That's the first problem with this model. The second problem with this model is not taking into account medical changes. Literally. We now have a medication that's FDA approved Remdesivir to treat this disease and we did not have that a week ago and we have not actually started using it. Widespread and it's not really playing into mortality rates at all because it's relatively new. We have a whole pipeline of new medications being studied, being deployed as we speak and none of this stuff is being taken into account. I just recently did a video, a four part video series that's going to be put out there talking about treatments, modalities and then our new understanding of this disease and how it actually works related to inflammation in the small arteries. These are all new things we're learning. We literally had no idea about this when it first popped up. We assumed it was going to be super, super lethal and we had wildly exaggerated projections. In retrospect, they were over projections. But we did the right thing. We locked down, the numbers have come down. So we're going to see new projections spike. An expectation that we're going to go back to the old data lines. The problem is we have a new road now. The new road is we know that we can social isolate. We know that Americans can do responsible things. We know that we can actually stay at home when we're told to stay at home. And we can actually flatten this curve. And that's new information that these models will not take into account. The new information in regards to the new medications and even a vaccine which we did not know as of a week ago. There's some new data that we actually might have a vaccine come fall based on some new capsid information about either capsid or based on the proteases and the COVID 19. So there are actually some really great things happening. Point being is this model does not take any of these things into account. It does not take into account that the virus won't spread quite as quickly in the summertime and also is not taking into account that if people do actually start wearing face masks as we go back out there, it's assuming we're just going to go back out there without any kind of precautions going on. We'll just flood the streets, so to speak. So I think the model is useful to say, hey, we should expect a spike and it's going to be like a slow dribble throughout the summertime. Some places like population centers, that is New York City might have a significant spike. Other places like California that have not had a big spike initially may not. We don't know. We need to remember that the United States as a whole, if you live in Virginia, which is where I live and my family member lives, hopefully you're watching this right now, is that we're not going to see. We need to look at Virginia data, not United States data because one third of all these cases to date are in New York and most of those are in New York City proper. We cannot expect what happens in New York City to happen in Richmond, Virginia or Northern Virginia or any other place in the country. I think that's the take home thing is not to be, not to look at the data national the spike, but look at your local area and the University of Washington has a great resource there. The other thing is when you're looking at that website, realize there's a big shaded area of color. That's the, that's the, the area of variation. And you'll see that's a huge, huge variation. So again, I expect we should expect to see a spike. We should expect to see that as we go back out. It doesn't mean that reintroducing ourselves to culture and society slowly graded the young, healthy people going out first, the low risk people going out first. It doesn't mean that's a bad plan. It means we should expect to see infections. The reality is a vaccine is not going to be here this month or next month. We now know this, this virus is not as deadly as we originally thought it was. As we get more data from testing, the mortality rate is only going down and that's the most important thing. It's not the projected deaths but the actual mortality because projected deaths deaths have already been as high as 2.4 million, as low as 64,000. That's a wildly variable range. So take the take home points are one, expect to see a spike which we've been talking about. Two expect projections to reflect that. Three, don't get upset when you hear this in the media because we expect this. Look more to your locality, go to the University of Washington website, see what they say there and see what your numbers are and to see what you should be expecting in your local area. Most states in the United States are not going to see massive spikes because most states do not have huge population centers like New York City. That's another thing to take into account is that when you have a model for the country, New York City, which represents one third of all cases, is going to represent over 30% of all this sway. So literally they're going to have a massive impact on these national data which does not apply to you locally wherever you are. So look at the University of Washington site, look at where you live, what state, and then look at your locality and see if you can hone things in even more specifically to you. Anyway, hopefully that was helpful. Do expect to see the spike, look at these numbers. I'll be curious to see how the media reports this next day or two and people's reactions to it. But ultimately this is nothing we did not expect. If you've been watching our videos, it's not new, it's just now it's out there and people are going to start asking questions, questions about, feel free to post questions to this link. We'll continue to do updates. Please continue to share these with family and friends. We want to get good information out there. Continue to ask questions. I do have a Q and A that we'll be doing sometime soon hopefully next one to three days answering some questions that have already been asked. Everybody stay safe, be well, keep your head in the game and realize that ultimately we have already come so far. We've already had so many new modalities and things happen that we have never even imagined four to six weeks ago. We're going to get through this together as a country, as a nation, as a community, as individuals, as families. Don't let the hype so to speak spiral you but ultimately stay informed. Again these are the three values of Richmond integrative and functional education, encouragement and empowerment and starts with education. Get some good facts, find reliable places you can get information and avoid the chaos. Take care and be well.
