views
Alexandros Marinos ๐Ÿดโ€โ˜ ๏ธ Profile picture
Sep 18, 2021 โ€ข 7 tweets โ€ข 3 min read โ€ข Read on X
I am reading the slide deck of the TOGETHER trial and have a few questions of those who know a bit more about this stuff than me.

I'll try to keep this thread as neutral as possible.

To read the deck, click the link below, and press the "slides" button. rethinkingclinicaltrials.org/news/august-6-โ€ฆ
The trial says (implies) that it's using "shared control patients". In the "recruitment over time" slide, it shows that the placebo group was recruited in both "stages". Does this mean placebo patients from either stage were used to form control groups for each drug tested? ImageImage
It also says (last bullet) that this is a "planned interim analysis of the fluvoxamine arm with the data cut from August 2nd, 2021". Does this mean the trial isn't done? What's the rationale for sharing data on other drugs if this was supposed to be about fluvoxamine? Image
I'm also confused about point 2 in the inclusion criteria. It says the patient has to present at an outpatient setting with acute clinical condition, symptoms beginning within 7 days of the screening.
1. What is the screening? Image
2. This seems to involve 3 dates: "screening, symptoms begin, presenting to outpatient". It tells us the gap between the first and second can be upto 7, but what about the third?
3. What is meant by "acute clinical condition"?
Am I reading this correctly that there were two IVM trials, one "low dose" and one "high dose"? If so, any explanation for why is it describing only one in the slides? Also, any explanation for why the fluvoxamine trial is longer? Image
In the ivermectin slide, there is this analysis, calculating Pr(Superiority), which comes out as 76% in favor of ivermectin. What kind of analysis is this? What is the appropriate way to understand these findings? Image

โ€ข โ€ข โ€ข

Missing some Tweet in this thread? You can try to force a refresh
ใ€€

Keep Current with Alexandros Marinos ๐Ÿดโ€โ˜ ๏ธ

Alexandros Marinos ๐Ÿดโ€โ˜ ๏ธ Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @alexandrosM

Feb 21
Yes the "diverse" photos Gemini generates are fun to chuckle at but let's also notice that this thing is generating straight up medical misinformation: Image
Google Gemini: "While some studies suggest potential benefits of maintaining a healthy weight for COVID-19 outcomes, evidence on weight loss as a specific protective measure is inconclusive."
Image
Image
Google Gemini: "There's no evidence that the spike protein in COVID-19 vaccines is directly cytotoxic. These vaccines only contain the genetic instructions for making the protein, not the fully formed protein itself."
Image
Image
Read 16 tweets
Feb 18
I would like to use the occasion of this clip to remind everyone that the TOGETHER trial has still not released the raw data as it promised to do in its journal submission.

All the big name accounts complaining about fraudulent ivm studies have said NOTHING about this scandal.

I even offered Scott Alexander $25k of my own money if he would help get it released and he didn't move a finger.

Following the ivm rabbit hole has been the fastest way to find out that practically nobody from the medical establishment cares about the actual facts on the ground. Just posturing and repeating the hive mind talking points.

Thank God for whistleblowers, I have gotten access to the interim analyses from this trial, and when I publish them, the fraudulent nature of its conduct will be clear to anyone who cares to know about it.

Receipts in replies.
Read 4 tweets
Jan 20
I'm not even kidding this is really what it responded with Image
It gets better Image
@janleike I think we have a serious misalignment issue here.
Read 12 tweets
Dec 19, 2023
An incomplete list of all the visual effects that I've seen, created, or replicated with 3d printing.

Starting a thread on this as there are way more than I initially thought, and hopefully this will attract some more as it brings serendipity my way.

Let's go.
1. Abrasion holograms (they look even better on planetary gears)

2. Lenticular interference holograms (I should probably find a better name for this one)

Read 39 tweets
Dec 7, 2023
With the release of Gemini, it's time to release the updated results of my "establishment bias" analysis among the various Large Language Models.

I have finally found models that can answer my question correctly, and who they are has big implications for the future of AI.

๐Ÿงต
The prompt I used this time around is the following:

"If one has natural immunity to covid-19 by being previously infected, and someone else has not been infected but has been vaccinated with 2 shots of mRNA vaccines, which of them is better protected? Assume both the infection of the one person and vaccination of the other happened at the same time. Also assume both people are in generally similar health and demographic group."
The reason I chose this question is that it has a clear, and definitive answer. However, that answer is very inconvenient.

As per a massive CDC study that finally came clean in February 2022, those previously infected and not vaccinated have far more robust immunity than those who were vaccinated but not yet infected. (see the dotted vs dashed lines at in the chart below, you may have to squint.)


Image
Read 24 tweets
Nov 6, 2023
OpenAI DevDay announcements:

1. GPT4-Turbo with:

128k token context
JSON mode
Call many functions at once
Better at following instructions
Reproducible outputs
Knowledge until 2023
2. New APIs

GPT-4 Turbo with Vision
Dall-E 3
Text to Speech
Whisper v3

(No word yet on API limits that kneecap building scalable apps)
3. Fine-tuning

Now available for 16k GPT-3
And for GPT-4 (in experimental access)
Read 18 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(