Jump to content

APPROPRIATE ACTIONS OR EXTREME MEASURES?


Recommended Posts

4 hours ago, markeb said:

 

You can run epidemiological models that are much more complicated than a single random distribution population. I assume most of the models being run at Hopkins, for instance, are using multiple subpopulations and do not assume random mixing. Takes a lot more time, even with good computers, but would much better fit the observed behavior over last 6 months or so. The challenge is projecting forward if you over fit the curve.

There's a saying about models:
All models are wrong, some models are useful.

 

One of the big challenges is determining what societal behaviour looks like from a disease transmission perspective.  Unfortunately, many of the models are far more simplistic than you assume, and often do use very basic parameters to represent human interaction.   That's how the presumption of herd immunity was determined (for vaccinations) that says 60% of the population is the magic number.   A study that tries to make better representations of human interactions puts that number far lower.

https://science.sciencemag.org/content/369/6505/846

 

On the modeling side, the last 6 months of data really doesn't help as it's been nothing but abnormal behaviour-wise.  It's an interesting data point for what happens in a lock-down, but it does little to suggest a path out of COVID.  I found this New England Journal of Medicine article interesting in regards to modeling:  https://www.nejm.org/doi/full/10.1056/NEJMp2016822

In all mechanistic models, epidemics can die away in two ways: either the disease runs out of fuel because there are no longer enough susceptible people to infect, or something changes to slow or halt transmission — for example, the number of contacts is reduced by dramatic physical distancing interventions. Since this latter mechanism slows the spread of disease without changing the number of people at risk, Covid-19 models agree that almost all populations are at risk of disease resurgence when societies reopen.

 

The current situation is like taking shelter in a bad horror movie.  The killer is still out there waiting for you when you start to think the coast is clear and try to leave.  Epidemiologists (Fauci included) have offered nothing in terms of a non-vaccine long-term strategy that gets us past COVID19, and two of the major vaccine trials were paused the last I saw. 

 

Link to post
Share on other sites
4 hours ago, phoenix_dream said:

The interesting part of my observations is that 1) I live in an area with a low infection rate, 2) the people in question are not necessarily in my usual 'bubble' - most are acquaintances at best and/or family/friends of family/friends. and 3) being retired I don't interact with nearly the amount of people I did in my working days so my base of exposure is much smaller than 10 years ago. I absolutely get the point you are making, but add all these factors together and it makes the statistics somewhat suspect I believe.

Gotcha.   With you not being 'core' to their social circles, it's hard to say who else they know directly that has been infected.  May very well be clusters of cases, which would be expected given that it takes time in close proximity to receive an inoculating dose. 

 

I don't personally know anyone who has had it.   DW has a co-worker that did, and I've heard stories of people who know people, but couldn't tell you any names.  It's simply not in our social circle right now.  

 

Expand that to a cruise ship and combine it with the fact that we tend to be creatures of habit, often frequenting the same locations at the same times each day, I would expect that spreading events on a ship should be well-contained with even moderate precautions.   

Link to post
Share on other sites
25 minutes ago, LGW59 said:

Arm chair scientist....UGH

 

Critic that doesn't think people should be able to independently research topics of interest that they may actually know something about.....UGH

Link to post
Share on other sites
59 minutes ago, D C said:

 

Critic that doesn't think people should be able to independently research topics of interest that they may actually know something about.....UGH

 

Select and copy user name, go to your profile, open ignored users, paste, select all.

 

Easy...

Link to post
Share on other sites
1 hour ago, D C said:

There's a saying about models:
All models are wrong, some models are useful.

One of the big challenges is determining what societal behaviour looks like from a disease transmission perspective.  Unfortunately, many of the models are far more simplistic than you assume, and often do use very basic parameters to represent human interaction.   That's how the presumption of herd immunity was determined (for vaccinations) that says 60% of the population is the magic number.   A study that tries to make better representations of human interactions puts that number far lower.

https://science.sciencemag.org/content/369/6505/846

 

On the modeling side, the last 6 months of data really doesn't help as it's been nothing but abnormal behaviour-wise.  It's an interesting data point for what happens in a lock-down, but it does little to suggest a path out of COVID.  I found this New England Journal of Medicine article interesting in regards to modeling:  https://www.nejm.org/doi/full/10.1056/NEJMp2016822

 

The current situation is like taking shelter in a bad horror movie.  The killer is still out there waiting for you when you start to think the coast is clear and try to leave.  Epidemiologists (Fauci included) have offered nothing in terms of a non-vaccine long-term strategy that gets us past COVID19, and two of the major vaccine trials were paused the last I saw. 

 

Dave I am enjoying some of your colorful quotes (in red) in your post.  And your reasoning too.

Taking shelter in a bad horror movie.  LOL.  Have you seen the current Geico commercial on this very thing?  A bunch of teens have a car ready to leave the horror scene but talk themselves into hiding in the chainsaw shed.  😀

Link to post
Share on other sites
4 hours ago, D C said:

There's a saying about models:
All models are wrong, some models are useful.

 

We could have worked with some of the same people. My favorite saying about models.

 

I had some very smart, young Jedi Knights working around me at my last job who were amazing at working models, including a lot of subpopulation work and different interaction rates. I was a consumer of their work, and the guy who tries to describe what I could use. Definitely not the programmer or modeler! They were brave enough, and proven right, BTW, to challenge an almost religious adherence to a previous modeler when they recognized his work was based largely on fitting the curve of a previous outbreak, and that limited the ability to project onto other outbreaks. And we had the ability to run 5000 or more Monte Carlo simulations...

 

I tend to the easy 1-1/R0 equation, but really, as you suggest, you're really dealing with some summation of 1-1/Rn across different subpopulations. And if goal is to stop transmission, rather than stop disease, you're probably closer to 1-1/R0, but interactions between populations, and their ability to transmit, makes it complicated.

 

A discussion for a different discussion board, I'd think.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Forum Jump
    • Categories
      • Forum Assistance
      • Q&A: Cruise Insurance with Steve Dasseos of TripInsuranceStore.com
      • New Cruisers
      • Cruise Lines “A – O”
      • Cruise Lines “P – Z”
      • River Cruising
      • ROLL CALLS
      • Digital Photography & Cruise Technology
      • Special Interest Cruising
      • Cruise Discussion Topics
      • UK Cruising
      • Australia & New Zealand Cruisers
      • North American Homeports
      • Ports of Call
      • Cruise Conversations
×
×
  • Create New...