What doctors on the front lines wish they’d known a month ago • Hacker News Board

Re: What doctors on the front lines wish they’d known a month ago

#81

Post by IkmoIkmo » Sat, Apr 18, 2020, 3:23 PM UTC

shalmanese wrote: post #65

jordanbeiber wrote: post #63

Earlier quoted context omitted.

My wife runs a home for old people with dementia. A lot (100s) of deaths where old people have died of flu-like symptoms is counted as covid deaths. Teating has not been possible until just recently.

So couldn't it just be that New York is admitting many more sicker people into their ICUs vs Sweden's ICUs performing better?

Can be lots of things, there's lots of evidence that data quality is extremely poor right now, but not enough to be able to say accurately what all the data should be.

For example, in the Netherlands, in a blood test about 3% of the population was found to have antibodies, while the number of people confirmed to have corona via tests is at 0.15% of the population. We're looking at a 1:20 confirmed/actual rate.

That tells you testing is way off, about 20x so. The total tests performed is roughly similar to the US by the way, about 1 in 100.

Then there's death rates. The Netherlands currently saw 2 in 10.000 confirmed deaths from corona. This is about double the rate in the US. However if we look at excess deaths compared to a 3-year average for the same period (2017-2019) we see about triple the amount. Apparently 2 in 3 die of corona without it being tested or confirmed. (likely in nursing homes where mortality rates are often around 30% per year, and it's not uncommon for staff to find a patient died in their sleep for example).

And the Netherlands is a small country with high-quality public infrastructure. It's quite likely that it is among the countries with the best record-keeping system, and is among the countries with the highest tests per population, and even here we see massive undercounting of infections and deaths.

It's good to keep looking at the numbers, but only with a huge caveat. I'm actually completely flabbergasted that the media report on infection numbers without ANY mention of a confidence interval, standard deviation or some kind of uncertainty measurement, as well as some basic stats around testing methodologies, in particular when reporting cross-country comparisons. This should be the norm, but instead we just get country comparisons, world maps and graphs displaying wildly inaccurate and limited data (which is absolutely useful, but criminal to report without caveats). Apart from a mention about 'Chinese data isn't reliable' or 'India isn't testing much, so the problem may be far greater', there's really no consistent discussion about this. There's lots of standard statistical tools to express uncertainty which have been used in science for decades, but they're completely absent in our reporting or discussions at the moment.

Re: What doctors on the front lines wish they’d known a month ago

#82

Post by jordanbeiber » Sat, Apr 18, 2020, 3:24 PM UTC

pbhjpbhj wrote: post #68

jordanbeiber wrote: post #62

Earlier quoted context omitted.

The data out there indicate a 50% ICU death rate in general. Ventilators seem to exacerbate the condition for covid patients. Sweden seem to have better success with intensive care than average. Are we perhaps intubating less. Train of thought ^.

FWIW not all ventilation is intubation, you can have non-invasive ventilation. Just in case you weren't aware, this makes stats hard to ascertain as some sources seemingly are interpreting orginal sources as if they were synonymous.

Thanks.

I was assuming the 80% death rate was intubated ventilation.

Re: What doctors on the front lines wish they’d known a month ago

#83

Post by epmaybe » Sat, Apr 18, 2020, 3:25 PM UTC

carbocation wrote: post #76

This is a perplexing article. Prone ventilation has been known to reduce mortality in severe ARDS from randomized, controlled trials for almost a decade[1]. Proning is great! It keeps people alive. We are doing a lot of it. It's not new. 1 = https://www.nejm.org/doi/full/10.1056/NEJMoa1214103

I agree it was a weird focus of the article based on the title.

There are also problems associated with proning, including but not limited to: body habitus, loss of airway, hypotension, loss of venous access. I imagine that proning isn't always an option for some patients, and was basically told as much by an ICU attending last week.

I guess I'm trying to say that proning isn't a magic bullet, it's just one of many tools in a doctor's toolbox to try and save lives.

Re: What doctors on the front lines wish they’d known a month ago

#84

Post by barkingcat » Sat, Apr 18, 2020, 3:29 PM UTC

misja111 wrote: post #78

killIdeas wrote: post #67

Oxygen and laying prone is not exclusive to hospitals. I'm wondering if we can get oxygen tanks ready for home use. Reduce hospital load, reduce healthcare costs, get rid of the need to flatten the curve. (Waiting for a my job to begin, waiting for my son to get his surgery)

>> get rid of the need to flatten the curve. This. A policy of flattening the curve will probably have to last for one or two years until either a vaccine is found or there is enough herd immunity. This will not only destroy our economy but the isolation will be a psychological challenge for many as well. We have to take one step back and think why we wanted to flatten the curve in the first place. And that is becaus…

conceptually, flattening the curve is meant to turn an exponential infection curve into something more manageable.

We are flattening the curve, yes, initially because hospitals don't have capacity, but the root of this issue is the exponential nature of infection transmission.

If we have the ability to build hospitals at a rate to match O(2^n) time then we can let everyone out and declare quarantine over! But as you know, an O(2^n) algorithm is extraordinarily hard to keep up with once n approaches any large number.

That's the curve we're trying to flatten - by changing the approach, changing the "algorithm" so to speak, by not allowing people to interact so they have less chance of infection.

If you are able to build a new hospital in a week, and then continue building a new hospital every day after that and then one every hour after that, and then a new hospital every few seconds after that, then you have a chance to keep up with an exponential curve (well, until the virus runs out of people to infect, so until it reaches 100% of the global population).

Re: What doctors on the front lines wish they’d known a month ago

#85

Post by pbhjpbhj » Sat, Apr 18, 2020, 3:33 PM UTC

DanBC wrote: post #53

downerending wrote: post #22

Earlier quoted context omitted.

Without disputing that (and thx for link), it's worth noting that the "flu" has been killing a large number of people year in and year out for as long as we can remember. The long-term death rate of COVID-19 remains to be seen. The final verdict on how bad it was will depend on whether or not lengthy immunity is attained (including to mutations).

> it's worth noting that the "flu" has been killing a large number of people year in and year out for as long as we can remember. > The long-term death rate of COVID-19 remains to be seen. At the moment we count flu deaths differently to covid-19 deaths. Counting deaths due to flu is hard. We've only just started this work for Covid-19 by putting in standards for death certification. These stats lag the real time cou…

Well in England&Wales the National Statistics Office figures show Now there's some controversy that UK government have been reporting other deaths as flu, basically hiding Winter deaths due, eg to poor elderly care, in flu figures. So other sources suggest far far higher flu rates; but this is going off death registrations.

https://www.ons.gov.uk/peoplepopulationandcommunity/birthsde...

Covid19 death rates for under 50s are something of the order 1:1000, 30x the flu rate in data I've seen most recent (Worldometer) but reported rates vary considerably.

Re: What doctors on the front lines wish they’d known a month ago

#86

Post by gdulli » Sat, Apr 18, 2020, 3:36 PM UTC

somewhereoutth wrote: post #70

nickthemagicman wrote: post #6

That if you're under 54 corona virus isn't any more deadly than the flu?

I think the critical thing is we don't know So best to err on the side of caution until this disease is more fully understood.

Also, comparing this to anything long-standing is flawed because it misses the point that we have an opportunity to solve this that we don't have with other situations.

Smoking habits and seatbelt habits and just about everything else is in a steady state. Nothing we can do can change those situations dramatically, we can only nudge them to a small degree slowly because those situations and habits are entrenched.

The point isn't that N people die from this so we should be cool with it because N is in the same ballpark as stuff we've accepted. The point is that we have a short window to solve the new situation before acceptance sets in.

Re: What doctors on the front lines wish they’d known a month ago

#87

Post by learc83 » Sat, Apr 18, 2020, 3:36 PM UTC

ethbro wrote: post #69

jordanbeiber wrote: post #62

Earlier quoted context omitted.

The data out there indicate a 50% ICU death rate in general. Ventilators seem to exacerbate the condition for covid patients. Sweden seem to have better success with intensive care than average. Are we perhaps intubating less. Train of thought ^.

Confounding factor: obesity rates and comorbidities From a glance, the US averages 3x the incidence rate of obesity as Sweden. Obese patients with existing health problems will result in substantially worse outcomes from Covid-19.

Where are you getting those numbers. From this study it looks like 20% of Swedish adults are obese. That's compared to 36.5% of US adults.

Also the swedish study is from 2012, so it's probably slightly higher now.

https://www.sciencedaily.com/releases/2016/06/160604050632.h...

Re: What doctors on the front lines wish they’d known a month ago

#88

Post by misja111 » Sat, Apr 18, 2020, 3:38 PM UTC

barkingcat wrote: post #84

misja111 wrote: post #78

Earlier quoted context omitted.

>> get rid of the need to flatten the curve. This. A policy of flattening the curve will probably have to last for one or two years until either a vaccine is found or there is enough herd immunity. This will not only destroy our economy but the isolation will be a psychological challenge for many as well. We have to take one step back and think why we wanted to flatten the curve in the first place. And that is becaus…

conceptually, flattening the curve is meant to turn an exponential infection curve into something more manageable. We are flattening the curve, yes, initially because hospitals don't have capacity, but the root of this issue is the exponential nature of infection transmission. If we have the ability to build hospitals at a rate to match O(2^n) time then we can let everyone out and declare quarantine over! But as you…

I was writing that we should both create hospital capacity and have some moderate flattening of the curve. Of course we shouldn't let the virus run free.

But unless you are hoping for a vaccine to be invented soon, which would of course be wonderful, you have to face the other scenario and that is that we have to slowly build up herd immunity. Herd immunity means that people have to get infected and a percentage of that will have to go to the hospital. No matter how you manage it, bigger hospital capacity simply means that you are able to reach herd immunity more quickly.

Re: What doctors on the front lines wish they’d known a month ago

#89

Post by SilasX » Sat, Apr 18, 2020, 3:45 PM UTC

carbocation wrote: post #76

This is a perplexing article. Prone ventilation has been known to reduce mortality in severe ARDS from randomized, controlled trials for almost a decade[1]. Proning is great! It keeps people alive. We are doing a lot of it. It's not new. 1 = https://www.nejm.org/doi/full/10.1056/NEJMoa1214103

Okay but then how is that something "they wish they knew" a month ago? They didn't know then?

Re: What doctors on the front lines wish they’d known a month ago

#90

Post by maxerickson » Sat, Apr 18, 2020, 3:46 PM UTC

misja111 wrote: post #78

killIdeas wrote: post #67

Oxygen and laying prone is not exclusive to hospitals. I'm wondering if we can get oxygen tanks ready for home use. Reduce hospital load, reduce healthcare costs, get rid of the need to flatten the curve. (Waiting for a my job to begin, waiting for my son to get his surgery)

>> get rid of the need to flatten the curve. This. A policy of flattening the curve will probably have to last for one or two years until either a vaccine is found or there is enough herd immunity. This will not only destroy our economy but the isolation will be a psychological challenge for many as well. We have to take one step back and think why we wanted to flatten the curve in the first place. And that is becaus…

We have to smash the curve so that we don't need the hospital capacity.

It turns out that getting the number of infected down close to zero would also be good for the economy.

I don't understand why so many people think that the best possible plan is for everyone to get infected. The best possibilities involve a few percentage points of the global population getting infected, no where near everyone.