doced.bsky.social
Doctor in Anaesthetics & Intensive Care Medicine | Part time academic | R enthusiast | Sci-Fi fan | Dad
https://doced.github.io
472 posts
1,126 followers
160 following
Regular Contributor
Active Commenter
comment in response to
post
Even if there isn't a biological sequelae for them, it's probably a week a year or so extra in taking leave to look after them. Which scaled up, has a pretty big societal economic implication.
comment in response to
post
My chief concern isn't long covid per se, but the long term (decades) consequences of routine exposure to inflammation (myocarditis, vasculitis etc.). My kids don't get flu every year, but they do get covid at least once a year, and typically have 5-6 days of high grade fever.
comment in response to
post
10% doesn't add up. It would be all we talk about at school drop off, and would be in the school bulletins. I'm in a bunch of large parent WhatsApp groups, and considering the trivialities of childhood that are often discussed, it would be impossible for 10% long covid to never come up.
comment in response to
post
Very informative. Thank you.
comment in response to
post
Thank you. I will read with great interest. I can see how the risk of acute severe illness is low, but I worry about the long term health implications. What are we storing up for future generations?
comment in response to
post
I’m reluctant to take notice of US policy (including CDC) at this time. Agree with @chrischirp.bsky.social that if there is a benefit to the individual, but not deemed cost effective at scale, that I should be able to buy privately.
comment in response to
post
You can imagine how I feel about these companies sending a 23 year old with no prior experience or domain knowledge to review complex healthcare systems and tell us how to do it, for a huge sum of money.
comment in response to
post
One of my all time favourite Nigella recipes. 🤩
comment in response to
post
Adam, is there a good review of the evidence in this area (COVID vaccines risk benefit across age groups). My kids routinely get COVID each year and have a high grade fever for a week. I find it hard to understand how that regular exposure could be better than vaccination? @chrischirp.bsky.social
comment in response to
post
This is such a good question. But I must give way to Andy who has answered it so perfectly aligned to my own thoughts that I literally have nothing further to contribute 😂
comment in response to
post
I’ve found a lot of @f2harrell.bsky.social ’s work on the subject very informative. Regression modelling strategies was great. And BBR filled more gaps.
comment in response to
post
I quite like odds ratios, but I’ve become accustomed to them because of how common they are in my domain. I accept all the arguments against them.
comment in response to
post
🙃
comment in response to
post
Hadn’t thought about that but you’re absolutely right.
comment in response to
post
I've just skimmed through some of my interest areas...this resource is insanely good. Wow.
comment in response to
post
All totally fair points well made.
comment in response to
post
I really only got it done in no small part because a super enthusiastic supervisor took the time (unpaid and in their own time of course!) to take me on as an extra. She knows who she is, and that she’s a total legend.
comment in response to
post
Though I suppose they aren’t called “guessing tubes” for nothing!
comment in response to
post
I see your point, and of course detection and interpretation of physical signs is imperfect. But US (especially echo) does have a much steeper learning curve and training requirement to be competent than hearing crackles through a stethoscope.
comment in response to
post
I do like DAGs for expressing our current best knowledge about a system. But then I really like RCTs because it mitigates the need to control for any of that complexity.
comment in response to
post
Who “owns” FAMUS? FICM and ICS relations are… unique 🫣
comment in response to
post
I tried for years to do FUSIC. Just couldn’t make it work at any centre while moving every 3-6 months. Eventually had to just dig in deep to do it. Hopefully that can change for others in the future. Hubs sound like a really good idea.
comment in response to
post
I would be really surprised if they will be mandated while ICS holds the accreditation process. Quite clear that they (ICS) aren’t an appropriate body to hold this, but I don’t see them giving up that gravy train any time soon!
comment in response to
post
Sorry, I mean after randomisation only arrows going into treatment are that of treatment allocation. Arrow from treatment to outcome is what we are evaluating.
comment in response to
post
Which is very very easy to do!
comment in response to
post
I don’t see why you need a DAG for that if you are planning to randomise, as the only arrows that remain after randomisation are treatment into outcome. Scientific thought is needed: does this therapy have a mechanistic theory to work in this case.
comment in response to
post
I basically did it by coming in on my days off for 6 months. It’s *very* hard to do in rotational post unless at a center for excellence.
comment in response to
post
Having recently done FUSIC heart, I think it’s probably not quite sufficient as a qualification. The more I learn about echo the more I realise just how complex it is.
comment in response to
post
When I was a medical student on ICU, the consultant did a massive (solo double handed) SLR to a patient and their BP went from 60 systolic to 120. That was the single moment I thought “I want to be an intensivist”.
comment in response to
post
This is so interesting. Thanks for sharing!
comment in response to
post
The governance stuff is certainly interesting. Since I’ve gone from “training” to “accredited” (only FUSIC level) I’ve felt much higher burden of responsibility to “get it right” which has actually made me want to get a lot more training.
comment in response to
post
I’m not sure how causal models would help unpick that. And of course, if randomising, then there is only one arrow going from treatment to outcome. I agree that this lumping issue is going to absolutely trash statistical power though.
comment in response to
post
But there's a nuanced argument to be had. For example, if there was a highly conserved common biological pathway for inflammation (e.g. IL-6 response to LPS) then one could reasonably expect a treatment targeted to that pathway to work across a broad number of "diseases" that trigger the syndrome.
comment in response to
post
I also think REMAP-CAP is problematic in the way it is being conducted and interpreted. I really don't think it's appropriate to have influenza and bacterial pneumonia in the same RCT for steroids. I've always struggled to understand this rational. It seems a lump too far!
comment in response to
post
I enjoyed the video, thank you for sharing. I continue to think you make excellent points, many of which I agree with. But I feel the conversation isn't moving forward. There are assumptions on both sides of the argument. I'm interested in trying to prove/refute the argument in practical terms.