Thinking about AI being used in conjunction with insurance at all makes me physically cringe. If somebody was looking for a tool to remove responsibility for denying claims or applying bias, it’s the tool for the job.
I know that’s not the side that you’re talking about here, but still, I don’t think I trust a system that’s basically Giant Autocomplete to handle this
Fun Fact, United Healthcare has been rolling out automated claim handleing and denial for a couple of years now, leading to increased denial rates, even after their CEO was murdered.
Yeah, basically for every hour a doc spends with a patient, they spend 2 hours with the electronic medical record, even with scribes. A whole lot of effort is going to modernizing and automating this, as it will free up a whole lot of doctor and nurse time, which is sorely needed.
Er, what? On what knowledge or experience are you basing that assessment of their claim? The ratio can vary widely depending on the specific job, but they’re not wrong. I’m a psychiatric nurse practitioner at a psychiatric hospital. My previous role at this facility was in the ball park of 2-3 hours of chart time per 1 hour of patient face time. In my current role it’s more variable, but it’s still around 2 to 1. Mind you, that’s not all just typing notes or checking boxes - both roles necessarily involve a fair bit of reviewing of records too. And I’m not counting coordinating with the care team, communicating with family members, staff meetings that don’t pertain to patient care, mandatory training time, etc. etc…
On automation, I can’t speak for others but I don’t see much automation of my job happening before we have artificial general intelligence, which current “AI” plainly is not. Even using AI just to do automated visit summaries wouldn’t be useful for me given how much time I spend talking with patients about the content of their hallucinations and delusions, which is something a LLM simply can’t understand in any meaningful way that would allow for a logical/useful summary.
A lot of medicine is chart keeping and interacting with insurance companies.
Thinking about AI being used in conjunction with insurance at all makes me physically cringe. If somebody was looking for a tool to remove responsibility for denying claims or applying bias, it’s the tool for the job.
I know that’s not the side that you’re talking about here, but still, I don’t think I trust a system that’s basically Giant Autocomplete to handle this
Algorithms have been doing this for years now
Not basically, that’s exactly what it is.
Fun Fact, United Healthcare has been rolling out automated claim handleing and denial for a couple of years now, leading to increased denial rates, even after their CEO was murdered.
Yeah, basically for every hour a doc spends with a patient, they spend 2 hours with the electronic medical record, even with scribes. A whole lot of effort is going to modernizing and automating this, as it will free up a whole lot of doctor and nurse time, which is sorely needed.
Thats such a stupid claim that others become more stupid just from reading it
Er, what? On what knowledge or experience are you basing that assessment of their claim? The ratio can vary widely depending on the specific job, but they’re not wrong. I’m a psychiatric nurse practitioner at a psychiatric hospital. My previous role at this facility was in the ball park of 2-3 hours of chart time per 1 hour of patient face time. In my current role it’s more variable, but it’s still around 2 to 1. Mind you, that’s not all just typing notes or checking boxes - both roles necessarily involve a fair bit of reviewing of records too. And I’m not counting coordinating with the care team, communicating with family members, staff meetings that don’t pertain to patient care, mandatory training time, etc. etc…
On automation, I can’t speak for others but I don’t see much automation of my job happening before we have artificial general intelligence, which current “AI” plainly is not. Even using AI just to do automated visit summaries wouldn’t be useful for me given how much time I spend talking with patients about the content of their hallucinations and delusions, which is something a LLM simply can’t understand in any meaningful way that would allow for a logical/useful summary.