Without any checks and balances, it would be equally reasonable to assume that the large number of calls occupying outreach services that aren't required may have the reverse effect. While very well meaning, we really do need to learn the lessons of untested policy roll out.
#rstats is so back 💪🤩
Honestly one of my favourites
Behold. The quadruple blinded trial: associationofanaesthetists-publications.onlinelibrary.wiley.com/doi/10.1111/...
So gooooood
A digital CAPTCHA verification window titled "Select all squares with PIPES" against a plain white background. The window contains a 3×3 grid of numbered squares, mixing literal hardware, smoking pipes, and programming syntax.
These captchas just keep getting harder #rstats
Ahh. Always makes my day seeing this.
🚩 the positive predictive value of an article that uses "leveraging" and "AI" in the title being bullshit approaches 1.
Through the looking glass. It has never been more important to have real expertise. I'm pretty worried about the AI slop future of healthcare (and research)
When training data is reddit, this is the kind of nonsense you can expect to see. Since everything is now AI slop, we should expect these models to get much worse from here!
Does anyone ever think about how supervision and mentorship are somewhat missing from healthcare. It's basically all self directed, with some form filling along the way. I can't help but feel that we could be doing it so much better.
Hahaha 😂 ok ok totally fair.
My god. This is a terrible idea for ECHO.
Underrated comment.
Lol. STH; obviously an excellent hospital. But I certainly wouldn't put it head and shoulders above the rest. Also kinda depends on what's wrong with you.
Erm ok. So if we are summarising the article as "exercise is better than no exercise" I suppose it is evidence based. But it equally isn't really adding anything of value to the discussion. There's a lot of great stuff in exercise/sports science. This was just a bit meh really.
There are some rare scenarios where the stars align, and there is genuinely some quality of the domain that can be exploited to give a real causal inference. But that feels to me to be the exception rather than the rule.
Internally I think of CI methods as: generally thoughtful retrospective analysis. Likely some unprovable assertions. Raw data still likely garbage. But good indicators of prevailing associations to help guide experimental research (time and cost prohibitive).
Yikes. Really? The original article isn't great. The advice to avoid in vogue fads is obviously sensible. But the specific exercise advice at the end really isn't very good or evidence based. The irony being that the plank as an exercise is somewhat driven by a trend itself rather than biomechanics
While I think the comments on silly trends are spot on (snake yoga), the exercise advice actually isn't very good. I could critique a lot from an evidence based perspective. A trivial example would be plank as an isometric hold is a highly inefficient exercise. It isn't a well researched piece.
Gell-Mann amnesia strikes again!
I deeply miss the time when I didn't hear trumps name every day.
Data.table almost made me drop R very early on. It was just the worst.
Analyse as you Randomise.
This recent RCT of an "AI stethoscope" claims the technology "shows promise" for diagnosing cardiovascular conditions.
It does not.
It is a textbook example of the risks of conducting unprincipled 'per protocol analyses'. Once again, peer review at a major medical journal has failed.
🧵 1/
Wow. A friend just looped me in on what's happening over on the other site. I have absolutely no regrets about deleting that account. 🤯
I'm at the stage, where I think any therapy applied to "sepsis" is just a bit silly. It's like having a treatment for "cancer". We need to be more specific and treat biological entities. Also... That dose is absolutely wild.
😭😭
😂
