U.S. COVID policy has likely killed up to ~2x as many vulnerables by following feelings and intuition rather than science.
Not all-cause deaths - just *COVID deaths.*
Math suggests a different, better approach.
THREAD 1/
Conversation
Replying to
2/ What feels good doesn’t always do good - the law of unintended consequences.
For example, reducing COVID transmission "feels" smart, but some researchers recently found moderate social distancing might *increase* total COVID deaths.
royalsocietypublishing.org/doi/10.1098/rs
2
28
92
3/ More specifically, modeling by and has found that encouraging *normal* transmission among under-40s, while significantly protecting the vulnerable, could cut total mortality by ~45%.
Quote Tweet
Replying to @WesPegden and @BallouxFrancois
Also as you say, age effects are likely even more dramatic than total size.
Basically, the epidemic will be over when it is over for 20-35 year olds. Whether this happens before or after a large number of 60-100 year olds have been infected depends on our policy decisions.
2
49
122
4/ I want to repeat that in case you missed it: the Chikina models finds that allowing <40s to mingle normally, while reducing transmission by 70% among >40s, leads to nearly half the total mortality of reducing transmission among everyone.
3
24
96
5/ Proponents of "reduce all transmission, all the time" fail to understand local vs. global optimization - optimal decisions can vary depending on time-scale. Ex. painkillers might make you feel better in five minutes, and much worse in 5 years.
2
17
78
6/ The usual counter is that increased transmission among low-risk young and middle-aged adults will "spill over" into older, more vulnerable populations - thereby creating more deaths.
This is true in the short term, but false in the long term.
Quote Tweet
Replying to @WesPegden @JuliaLMarcus and 3 others
Any time population immunity plays a significant role in controlling an epidemic, and "herd immunity strategies" were not part of influencing the population which had infection-conferred immunity, significant avoidable suffering and death has occurred. 6/
2
11
63
7/ We all know that models can be disastrously wrong (Ferguson, Hotez, IHME). But Florida provides a good real-world example.
1
9
59
8/ We here see mathematical evidence of Dr. David Katz’s argument in March: the easiest way to protect the vulnerable is/was to shelter them while the low-risk young build up herd immunity.
nytimes.com/2020/03/20/opi
1
14
86
9/ Alternative approaches assume - as has, repeatedly, been proven incorrectly - that most younger Americans will, either voluntarily or under coercion, give up months (or years!) of their lives to reduce spread of a virus that poses them very low risk.
1
6
59
10 “We have to accept human behavior the way it is, not the way we wish it to be... we must design our [systems] on [this] assumption.” - Don Norman, the foremost expert in human-centered design
2
13
74
11/ Even in the strictest U.S. states (CA), spread has not reached de minimis levels until acquired population immunity is sufficient to reduce spread.
So we can keep (unsuccessfully) fighting human behavior - or, we can make human behavior work FOR us.
1
21
88
12/ Some preposterously claim we can’t successfully briefly shelter the vulnerable (ex 0.5% of Americans in nursing homes who account for 40% of deaths), but CAN shelter everyone indefinitely. This fails basic logic.
2
15
111
13/ It is much more feasible to intensely shelter the vulnerable for a short period of time, than it is to shelter them for a long/indefinite period of time. Leakage happens, and it’s Pyrrhic / inhumane to isolate frail elderly for half of their remaining lifespan.
1
13
85
14/ It is one thing to debate tradeoffs, i.e. whether non-age-targeted lockdowns are worth the myriad costs (mental health, foregone medical care, starvation, re-emergence of preventable or treatable diseases, economic and societal devastation.)
1
6
51
15/ It’s entirely another thing to advocate society-wide lockdowns when data and modeling clearly suggests the most feasible, practical, life-saving approach - *for the vulnerable* - is to shelter them intensely but briefly while letting the young become literal human shields.
1
16
64
16/ In many areas of the country, it is mostly too late for this. But in other areas - or if transmission picks back up in fall/winter - we should heed this math. Let’s stop shaming college students, and start following science and data - unless we want more grandmas to die!
1
9
79
17/ The Chikina paper is available here (it uses some dated assumptions around IFR and hospitalizations, but this doesn’t change the model.)
1
6
46
18/18 See also this thread from . And please follow for balanced, rational, mathematical, evidence-based analysis of COVID-related issues.
Quote Tweet
[1/n] Well @gummibear737 pushed me into a position to have to explain our theory of how universal mitigations (for ex. lockdowns) kill over time more people than what Sweden did.
Show this thread
5
7
40