Coronavirus and You - Page 85

On March 27 2020 06:41 LegalLord wrote:

That exponential is really frightening. And given that there's no real solidarity nationally in terms of what the proper level of quarantine is, I think it's going to get significantly worse before it gets better. The current case load is almost certainly a significant underestimate of the reality in the country.

On March 27 2020 08:45 Amui wrote:

My excel sheet puts it about 33% increase in cases per day since Mar 2nd, tracking to within +-15% of actual total. I suspect that testing will start falling behind though. That will put them at millions of moderate to severe cases(the tested ones) in a couple of weeks, and tens of millions of infected, even if they had a full quarantine going. Only a matter of time at that point before almost everybody in the US will catch it.

There's been no coordinated effort at a federal level, and depending on the state it's anything from a lockdown to keep everything open, spice must flow.

The people in the states who realize what's going to happen are powerless, and the people who have no fucking idea what is going on(Trump and co.), are going to end up killing more Americans than any war ever fought by the US at this rate.

On March 27 2020 08:53 Bagration wrote:


Quick question as I'm curious - in your excel model, what is your range of expected casualties for the USA? And for the world?

Thanks - and appreciate the work that you're doing. In times like these, more analysis is always nice

On March 27 2020 09:34 Amui wrote:

DISCLAIMER: I AM NOT A DOCTOR AND THIS IS NOT A FORMAL SCIENTIFIC STUDY.
I'm an engineer who does some data analysis/statistics and root cause analysis' for my work, but my sources are reddit articles and worldometers, and tracking the numbers there.


Mean time to onset of symptoms that I read was somewhere between 5-6 days, and mean time til death was almost 20 days. People don't keel over immediately, so there's going to be a ~2 week lag between infected numbers and deaths. US also has a ridiculous number of preexisting conditions (Obesity/diet related) so they could be in trouble there as well. Not a lot of states are taking it seriously at the moment from what I can read so I expect it to continue spreading almost unhindered in a lot of places.

Low end would be a hundred thousand or so. I fully expect them to shoot past this number officially. Unofficial deaths may be some percentage points higher, but at some point they will just assume most people who died did so of covid19 related causes. Final tally should be within double digit percents of the total.

The highest I would expect it to hit is 5-10M. This is what happens if healthcare is overwhelmed, and a low double digit number of the US population gets infected. I don't expect it to go beyond this even in a worse case scenario.

If you infected everybody in the US at once, @10% mortality it's 33M.

On March 27 2020 09:20 ggrrg wrote:
A “report on the risk analysis in civil protection 2012” for the German government is currently circulating heavily on German websites. The spiked interest is due to the document including a risk analysis section on a fictional pandemic virus “Modi-SARS”. + Show Spoiler + The document was compiled by various German government institutions alongside the Robert Koch-Institut (basically the Germany version of the CDC). Unfortunately, it is gaining traction for all the wrong reasons with conspiracy theories popping up all around ranging from the document actually being the currently real prediction for Covid-19 to it being a government plan to decimate the population…
Anyway, while reading there was one thing that struck me as extremely interesting, namely the prediction of the graph for infection progression – how fast, how many people will get infected at a certain point in time.
First of all, the fictional virus used for the risk analysis has all the properties and effects of the 2002-2003 SARS coronavirus including a 10% mortality rate (1-50% depending on age), 2/3s requiring hospitalization with 20-30% (of hospitalized people) requiring ICU treatment and 14% (again of hospitalized people) requiring ventilation. Other relevant properties: infectiousness starts 3 days after acquiring the virus; patients are infectious for around 13 days; average duration of hospitalization is 19 days; basic reproduction number assumed as r0 = 3; every survivor acquires immunity for 360 days and can be infected again by a mutated version of the virus after those 360 days. The only difference between SARS-CoV and “Modi-SARS” is the time till onset of infectiousness (much earlier for the fictional virus).
And this is the amount of infected people plotted over time they calculated. X-axis: days after first infection, Y-axis percent of the German population (assumed as 100% = 80 million):

Legend:
- symptomatic cases - hospitalized
- ICU treatment - dead
- beginning/end of intervention (measures taken to reduce r0)

First and foremost, the graph cannot be used as a prediction for the development of the Covid-19 pandemic for the obvious reason that the mortality, hospitalization as well as ventilation rates are all much higher for the fictional virus than they are for Covid-19. Additionally, the model assumes that immunity after recovery lasts only 360 days and that there will be 3 waves of infections.

The reason why I am showing this graph and what I find extremely interesting is the following:
- The fictional virus has a basic reproduction number of 3. The current SARS-CoV-2 has an estimated r0 of 1.4-3.9 according to the sources used for its wiki page + Show Spoiler + or alternatively 2.0-3.3 according to the RKI + Show Spoiler +. Basically, the current virus may be about as infective (possibly even more) as the fictional virus.
- The model and the calculations used for the graph actually assume that preventive measures to reduce the rate of new infections are implemented 40 days after the first case and last for 360 days! The measures are not completely listed in the document but examples listed are: quarantine of contact persons of infected cases, isolation of infected people, closing down schools, canceling major events, personal protection for working people at high risk, general hygiene recommendations. While the preventive measures are in place, the basic reproduction number is assumed to be only 1.6!
- The fictional model uses the premise that only 5% of the infected experience a mild case of the sickness and only 2% of the infected are asymptomatic.

I cannot comment on the accuracy of the model, but if we assumed that it is somewhat accurate and take under consideration that it uses a virus with a similar infectivity as SARS-CoV-2 and assumes preventive measures used (partially unknown, but at least the listed ones are being taken during the current pandemic), we can say that, while the number of hospitalized and dead may be well off, the number of infected with SARS-CoV-2 is accurately depicted in this graph. There are still unknowns that may result in slower infection rates than plotted such as the measures currently taken being more effective than the ones the model uses or the r0 of SARS-CoV-2 being on the lower end of the current estimates. At the same time, it may be the exact opposite: the measures taken being less impactful than the ones in the model and r0 being on the high end of the current estimates. On top of that, unless I am making a logical error here, the percentage of mild and asymptomatic cases in SARS-CoV-2 (~80%) being significantly higher than the one in the fictional virus would result in infections spreading faster than in the model.

Anyway, if we assumed (based on r0 and preventive measures being similar) that the progression of Covid-19 is going to be the same as the plotted progression, this would result in a peak of ~6 million infected at the same in Germany alone at day 300 after the first infection (the graph seems to be showing around 7.5 mil, but the table on the next page of the risk analysis assumes “~6 mil” as a peak for some reason). The current estimate for the percentage of severe Covid-19 cases is 20%, which would mean that at some point in time Germany would need 1.2 million hospital beds for Covid-19 patients only. The last estimate for critical Covid-19 cases was 5%, which in turn would mean that Germany would need 300.000 ICU beds at the peak (possibly the same number of ventilators as well) again only for Covid-19 patients.

Currently, Germany is said to have around 26 ventilators for 100k/population. This would mean that there are around 21k ventilators in the country, which would be horribly insufficient for the number of people that would require one if the calculations in the graph are even remotely applicable to the Covid-19 pandemic.

Personally, I feel like it is fair to assume that there are no significant errors in the model itself considering that the RKI was involved in its creation. I also feel like, it is not unlikely that the progression would be similar for Covid-19 given that the r0 used for the fictional virus is similar to the estimated one of SARS-CoV-2.
The only big unknown I can think of that could change the progression in a major way is the impact of the preventive measures taken. Possibly the lockdowns and restrictions on everyday life are much more impactful than anything used for the model. Even something like the ability to do remote work from home may change the course of progression in a notable way and is most likely something that was not considered back in 2012. On the other hand, it is a bit unsettling to see that the model assumed 360 consecutive days of preventive measures taken…

On March 27 2020 11:59 Vindicare605 wrote:


Right now it's still mostly impacting large urban areas. It hasn't hit the rural areas yet, but when it does it's going to be bad. Hospitals in rural areas have limited supplies and manpower even during normal periods, when this gets going there they are going to be hurting.

It's been in LA now for almost a month, and Garcetti (our Mayor) has told us that lockdown will continue into May since it doesn't look like it's slowing down.

On March 27 2020 14:55 Gina wrote:
So we're still walking in a wood nearby, and the number of people there pretty much trebled. It's still much easier to social distance there as long as you avoid the popular spots like the easiest tree to climb. But one thing I noticed was that with schools closed down and most people taking the kindergarten kids away voluntarily, it's the grannies who are taking up the slack. Of course it's typical for Russia, but very unfortunate in this particular case.

On March 26 2020 12:27 Yurie wrote:
Oh well, got flu symptoms. Hard to tell if it is the flu or this disease since the symptoms are the same, just different probabilities. Had cold like symptoms for three days, woke up shivering from fever in the middle of the night now. So no real concern thus far.

Finally realised why people bunkered up on toilet paper. To blow their noses. :p Without the need for that I had enough for weeks and didn't even consider buying any.

Stopped following this a while ago so never heard about ibuprofen. Guess I'll take it safe and stop that now. There are good alternatives so no reason to take the risk.

On March 27 2020 20:34 Firebolt145 wrote:
Well Boris Johnson now has covid. Not surprising.

On March 27 2020 20:34 Firebolt145 wrote:
Well Boris Johnson now has covid. Not surprising.

On March 27 2020 20:37 Kipsate wrote:

Better hope Covid doesn't get to the House of Lords, the average age there is.....well.