We have paid, and continue to pay, a huge economic price for it, but new data from an unexpected source gives compelling evidence that social distancing is working against COVID-19, as well as a providing a possible surprising dividend.
Categories
The CoronaSphere Lounge Episode 16: Social Distancing is Working
5 replies on “The CoronaSphere Lounge Episode 16: Social Distancing is Working”
1. As Dr. Birx pointed out in last night’s briefing, the current death rate reflects people who were infected two weeks ago. So when looking at these graphs, resist the temptation to make a direct correlation between “lockdown orders” and the downward slopes. Look at where the downward slopes begin and then look back one to two weeks to see what was implemented then. It seems to me that the big drops occur one to two weeks after schools were closed and all of the large events were cancelled. It’s certainly true that the lines continue to decrease and that much of that is attributable to the more restrictive social distancing being implemented. But what that suggests to me is that, looking towards next year’s flu and the next pandemic, we won’t have to take quite such draconian and economy-killing measures to slow the spread.
1a. The President and his task force doctors have repeatedly brought up the idea that following some of the guidelines during flu season in the future could substantially reduce “normal” flu deaths.
1b. It’s not just flu deaths that the shutdown has reduced. The original intention for the Navy hospital ships was for them to take non-Covid-19 cases to relieve that pressure on the local hospitals and allow them to focus on Covid-19. They were particularly going to emphasize trauma (battlefield surgeons being more skilled than most in that area). However, because no one is out driving, the number of trauma cases has dropped so much that they are instead going to use the ship for Covid-19–at least in NYC. So the social distancing is dramatically reducing “normal” levels of death by other causes. (Would be interesting to see the number of gunshot wounds in Chicago or Baltimore, for example.)
2. Could this dataset be used to examine VDH’s hypothesis in a recent article (which Bill mentioned a few episodes ago) that Covid-19 came to California last fall through the hundreds of thousands of people travelling between China and California, and its population has already achieved herd immunity?
The Hong Kong data matches this. Their closing of their borders, and social distancing and mask wearing drove influenza to basically 0.
Though, Influenza’s fatality rate from the start of the season last year, to February 1st, was only 20k to 52k. If there were basically 3 months in that data, and there were 2-3 months left to go in the season, then you would have to stop SARS-2 quickly and keep it that low for it to be any benefit to health. Not to mention that in the handful of places that have overwhelmed health systems, there’s non-flu non-SARS-2 deaths that are because of SARS-2 as well, but won’t get counted in the totals.
It’s promising that we might get this peak hammered down though.
Unfortunately, I’m in the “mild atypical” area of the map, and a lot of people were still out and about a lot in the last 2 weeks, so we might have a longer rough time. We’re also on a Steady state on the 7 day trend.
Covid-19’s theme song: I’ve been everywhere by Johnny Cash
Bill,
to what extent are these temps representative of the entire population? Seems to me that the sample might be highly weighted toward the young (teach absorbed) and affluent (own and use the watch). If that is the case, the conclusions could apply mostly to the specific population included in this data rather than the general population.
It may still apply to the general population, but that assumes their social distancing mirror the sample. I suspect the sample group has a much higher compliance with social distancing than those in the population as a whole. I suspect a higher percentage of the sampll population is working from home, ordering delivery, etc.
If I’m correct, the peak may be delayed and the peak might be higher than implied by this sample data.
It still strongly implies that social distancing has the positive effects you showed.
I agree that self-selection bias in this dataset needs to be addressed. I also agree that overall it probably doesn’t invalidate the data, but might skew it a little bit.