The Future Healthcare Journal debate focuses on whether AI might replace clinical decision-making in our lifetimes. Please read and vote (second link) And please share to convince RCP that Bluesky is the future! #MedSky #AI www.sciencedirect.com/science/arti... forms.office.com/pages/respon... - ThreadSky

lawtontri.bsky.social • 114 days ago

The Future Healthcare Journal debate focuses on whether AI might replace clinical decision-making in our lifetimes.

Please read and vote (second link)

And please share to convince RCP that Bluesky is the future!

#MedSky #AI

https://www.sciencedirect.com/science/article/pii/S2514664524015686

https://forms.office.com/pages/responsepage.aspx?id=_UITNMt_rkqMJxSNJB3wRyzDKIJtBoRJsYN1Oe4GzKVUOUpXMEsySVQ3RUgxRUoySzYwTDFDOVNDVi4u

Comments

dr-shenanigans.bsky.social•114 days ago

Voted.

I think it will be implemented....with a big old helping of if it goes wrong it's a named doctor's fault

lawtontri.bsky.social•114 days ago

On the plus side, that's one of the areas of near-universal agreement. So we need to support healthcare staff using AI to have meaningful understanding and control rather than any expectation they might just "rubber stamp" a machine recommendation

benecal.bsky.social•114 days ago

📌

veraaraujosoares.bsky.social•114 days ago

I would say it will aid decision making. A human in the equation will probably always be needed, particularly if we are to also engage in shared decision-making. I still cannot see, but can be persuaded, this process occurring between a machine and a human being. Of course perceptions change…

dobbyjog.medsky.social•114 days ago

My tuppence worth: no

☺️

lawtontri.bsky.social•114 days ago

My view is similar, but with a side order of "but somebody will try to make it happen anyway"

gpollara.bsky.social•114 days ago

yeah agree. we need to make sure that its introduction is assessed as a clinical intervention like a drug or diagnostic test would be.

What should not happen is that AI decision making tools are just bolted onto an EHR by the respective providers for people to use freely!

lawtontri.bsky.social•114 days ago

Absolutely. We're just finishing a project that's been exploring some of the ways these things could be used, and (spoiler alert) none of the successful models involve the machine making the decision itself (and we did test that, within the limits of plausible regulation)

gpollara.bsky.social•114 days ago

very interesting, thank you.

and apologies, i didn't mean to imply you wouldn't test them in the right regulatory setting. All I meant was that sometimes features pop up in EHRS that impact care without much testing or regulation (I've seen examples of this).

lawtontri.bsky.social•114 days ago

I imagine we all have! The regulatory framework is fascinating for systems which merely alert a human healthcare worker. I'm sure most of us dealing with electronic records have to deal with a variety of pop-ups every time we select a patient.

dobbyjog.medsky.social•114 days ago

Oh 100%

But we won’t have it for the same reason we don’t have pilotless planes.

lawtontri.bsky.social•114 days ago

Trains would be a start! I find it fascinating that in other safety critical industries they're more cautious - presumably because train and plane crashes attract more headlines.

dobbyjog.medsky.social•114 days ago

Yep - I guess it’s partly perception of risk (flying through the air in a pressurised metal tube is pretty tangible) and partly the headlines.

bonj.bsky.social•114 days ago

AI uses so much water. I wonder what will be gone first. Our water supply or us. https://www.watertechnologies.com/blog/artificial-intelligence-using-ton-water-heres-how-be-more-resourceful

Comments

Posting Rules

Reply