ivanonymous

It's a bad reply!

Narrowly because it's actually unfair point by point, as you say.

And slightly dishonest, because Scott does in fact entertain ideas that are scandalous to an expected arrival from the NYT. (Though not, I would say, for the reasons they might assume.)

But also more broadly, it doesn't take the opportunity to make intelligible to that arrival his insider's view of these related-but-distinct phenomena (his blog, the Rationalists, "Grey Tribe", the counter-techlash). Which the article predictably muddles.

Perhaps saddest to me, by responding defensively (however understandably!), he reinforces the emphasis in the piece on this sort of culture war stuff. Rather than perhaps channeling this new interest towards the meta-science and history and psychiatry and the like.

Besides close monitoring of people on O2 to adjust amounts and intubate if that time comes, there are potentially other components of supportive care that might alter outcomes, like IV anti-virals (e.g. remdesivir) or anti-coagulants, e.g. heparin. Some anti-coagulants can be given at home, but the strong doses being discussed also come with high risk of bleeding, benefitting from bed rest and close observation. For a good slice of the sickest COVID patients, hospitalization is helpful and hard to replicate at home, even with our limited therapies.

That said, oxygen and heparin at home could be good for a lot of patients. Improvements in prognostication may help distinguish that group in the future.

That ~90% figure is wildly misleading.

It's based on 282 deaths among the 320 ventilated patients who either died or were discharged.

But 831 patients were still on ventilators!

It's a snapshot taken too soon.

Estimates of mortality on vents very widely still, partly because of real underlying variation in practices or population, partly because the data is really messy.

This email series from Mass General is a good orientation (discusses this study, links to discussions of previous studies): https://mailchi.mp/db30d9d2cb24/tz4idnzryr-4406129?e=acf498e...

Any vent strategy being debated on Twitter (early! late! APRV!) is also being discussed by working pulmonary critical care docs. Judging those discussions or the variations in practice as an outsider is hard. But relative silence on Twitter doesn't equal mindless orthodoxy.

An outright cure at a year strikes me as unlikely, given the record with anti-virals.

But improvements in supportive care (anticoagulants?) and lucky repurposing might win meaningful reductions in morbidity (some permanent) and mortality within months.

I'm not sure how to weight that possibility. (Metaculus has settled on a likely ~5% relative risk reduction for chloroquine, which is no game changer.) But it seems underweighted in general in discussions about different paths to herd immunity.

That framing implies that all the relevant actors have committed to strategies that they'll see through to the end.

Learning from what's happening in Sweden should help guide decisions elsewhere.

The process of comparison is messy and politicized and in some sense always potentially premature, but I don't think for all that it should be forsworn.

My understanding is that there are variable lags up the chain, and some states don't report deaths for months. NYC as it happens is a relatively prompt reporter. Not sure if there is any public information locally before it's reported upwards, e.g. if you could send volunteers to county health offices.

This was discussed fairly widely recently when people were interpreting recent decline in year over year deaths as evidence that COVID responses were saving lives (e.g. less driving). But it was almost all just typical reporting lag.

Demonstrated by an animated graphic: https://www.reddit.com/r/dataisbeautiful/comments/fxau85/the...

Sampling uncertainties aside, this is super sensitive to mis-estimating the false positive rate. Which they seem fairly likely to have done, using an estimated test specificity of 100% (with error bars) in one branch of their analysis after getting 30/30 negatives against pre-COVID serum samples. Too small a sample to get that confident in an atypically high specificity, especially if that collection of samples was disproportionately low antibody (e.g. summer blood rather than post cold/flu season blood).

That's to my mind the most consequential misestimation of the test characteristics, but Balaji Srinivasan details more:

https://medium.com/@balajis/peer-review-of-covid-19-antibody...

I don't think the CDC/WHO recommendations started as a noble lie, just as a medical/bureaucratic orthodoxy, consistently expressed before any shortage.

The orthodoxy drew too sharp a distinction between airborne and droplet modes of transmission, it didn't anticipate the possibly extensive asymptomatic spread of this new virus, and it tried, given the low bandwidth of public health messaging in the old days, to speak simply and authoritartively in favor of the interventions with the strongest evidence for the largest effects. Even to healthcare workers, the emphasis in the setting of something like flu was on handwashing. Masks in clinic went on coughing patients, not us. For a public less consistently exposed, the number needed to mask to prevent a single infection was judged too low to bother. Masks in Asia were seen as public health theater, like spraying fog machines in the streets. And maybe most dangerously, most complicating a reversal, expressing an understanding of this mildly counterintuitive finding - covering faces doesn't stop respiratory viruses - was taken as a mark of scientificness. (And remember, it's still largely true, the effect is probably fairly small, N95 or cloth mask no matter. The virus spread well in places with mask-wearing, and dampened more quickly there for reasons other than masks.)

The CDC/WHO people are under pressure to change deeply held and strongly stated beliefs in public, and then to live with the implication that their confident error cost lives. Relative to that, communicating that we should wear masks, but leave the respirators for medical workers, is easy peasy.

For fun background on the orthodoxy, including a case where a single flu patient infected dozens of people on a plane in the 70s without managing to persuade people of some meaningful amount of airborne/fine aerosol spread: https://twitter.com/rkhamsi/status/1244659064350670848?s=19

These workers would be dependent contractors in Germany or Canada, I think. Defining that in-between category in the US would allow for some useful compromises. Especially because full employee comes with so much baggage here, e.g. health insurance.

https://www.wolfbaldwin.com/articles/small-business-articles...

I'm curious as a non-expert what's specifically worrying about their privacy model?

My intuition is that rapid adoption of a relatively transparent privacy-preseving option could preempt more heavy-handed approaches to what could be a very valuable public health intervention.

I don't want to discourage you, but please don't take your experience as a justification to ignore medical evidence and advice.

Kidney function is most often estimated indirectly, based on measurements of the amount of a chemical called creatine in the blood. Higher functioning kidneys lower creatine levels more effectively. But lowering the amount of creatine generated by your diet, by cutting out animal products, will also lower the creatine measured in your blood. That change in measurement does not reflect an actual improvement in your kidney function, I'm sorry to say.

At lower levels of kidney function, attempting to game the numbers in this way, by minimizing the creatinine you consume and by over-consuming water, could be dangerously misleading. And it shouldn't distract from measures that do stop the decline in kidney function, like blood pressure control and avoidance of certain kidney-damaging meds (e.g. high dose ibuprofen).

Good luck!