Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

inflecthealth.medium.com

Post reply on HN
Print view View original

171–180 of 253 posts

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#171

Post by ansc »

firebirdn99 wrote: post #44

This really feels farcical, and kind of like vc bros hyping up a product. Considering this was a serious issue of life and death of a patient, it was surprising how it read like a prank or like the scene from Harold and Kumar, where Kumar plays a fake doctor and saves an ER patient through sheer luck. I know some doctors are full of baloney but c'mon. I really hope maybe that we create better doctors that have a lot…

Alright, I'm going to miss the point entirely here, but I'm swinging any way: if I recall correctly, Kumar had basically studied and become a doctor, but was scared to take the final test due to his fathers pressure yada yada something like that, and it wasn't through sheer luck he saved the patient.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#172

Post by Sakos »

rpastuszak wrote: post #146
heliodor wrote: post #67

Earlier quoted context omitted.

The concept is right on point though: ChatGPT is excellent at writing "fluffy" pieces full of empathy, compassion, PR talk, politician speech, and any other wordy flavor you need. As an engineer who is unable to produce such writing, this tool is quite helpful!

> ChatGPT is excellent at writing "fluffy" pieces full of empathy, compassion, PR talk, politician speech, I think you're putting a bunch of completely different things in the same basket. Making the text more fluffy will not make it more empathetic, not only (but mainly) because there isn't anyone to empathise with the respondent. Our bullshit receptors are good spotting dishonesty, so it'll just sound cringe, putti…

> Our bullshit receptors are good spotting dishonesty

Is this a bad joke? What exactly leads you to that conclusion? People fall for dishonesty and straight up lies all the time. We're awful at recognizing it. Maybe you're the one in a billion who can spot a lie at a million miles, but the vast majority can't.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#173

Post by lukko »

hfo wrote: post #114

Physician here with ER experience (and IT too). I think it's quite realistic, except the relatives being in the ER room, giving orders to nurses. But that might be different in other hospitals and countries. The reason for the edema is not given, perhaps cardiac insufficience, perhaps she didn’t take her pills (it is mentioned that she had dementia). Anyway, it’s possible that treatment has already been done but woul…

I agree it's a realistic scenario. I always thought one of the most important things about being a junior doctor (FY1 / 2 in UK) was to be able to have these conversations with next of kin in a way that is empathic and easy for them to follow. If they are not processing the information then you have to allow for that, but I do think some relatives would react very badly if given a printout from an AI to explain a situation, patients always bring up if a doctor has googled their symptoms / rash in front of them, this is giving much more authority to the AI in a more sensitive and potentially difficult situation.

I think the way AI is used in medicine has to be thought about very deeply, I don't think it should just be a fallback in time-pressured or difficult scenarios.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#174

Post by lukko »

al_be_back wrote: post #167

>> I printed this response up, and read it to the concerned family. As I recited ChatGPT’s words, their agitated expressions immediately melted into calm agreeability. not bad for a corny $0.5 hardback novel - but wtf was that about? If a Doctor read that script about my mother in ER, I'd be more insulated than had they printed the source code of a Linux Kernel module and proceeded to read it out in full, verbatim! S…

Yep, I think fundamentally it feels wrong. I think many families would see it as a sign of incompetence if the doctor cannot explain a diagnosis or treatment plan - time pressure is part of the job unfortunately, and people may not process difficult information in stressful situations with as much clarity, you have to allow for those things. Reading a script from ChatGPT is insulting.

Patients hate when doctors just google their symptoms / rash in front of them, this is quite a step further. Printing out supporting information for families / patients is ok, but it needs to be validated – I think this is quite a dangerous use of generative AI for both the doctor and the patient.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#175

Post by f6v »

Sakos wrote: post #172
rpastuszak wrote: post #146

Earlier quoted context omitted.

> ChatGPT is excellent at writing "fluffy" pieces full of empathy, compassion, PR talk, politician speech, I think you're putting a bunch of completely different things in the same basket. Making the text more fluffy will not make it more empathetic, not only (but mainly) because there isn't anyone to empathise with the respondent. Our bullshit receptors are good spotting dishonesty, so it'll just sound cringe, putti…

> Our bullshit receptors are good spotting dishonesty Is this a bad joke? What exactly leads you to that conclusion? People fall for dishonesty and straight up lies all the time. We're awful at recognizing it. Maybe you're the one in a billion who can spot a lie at a million miles, but the vast majority can't.

> Our bullshit receptors are good spotting dishonesty

If I had to count with my fingers how many times people wondered whether someone was sincere or just polite, I’d have to be a thousand-finger person.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#176

Post by williamcotton »

vidarh wrote: post #166
tirpen wrote: post #155

Earlier quoted context omitted.

And thus we will have invented a lossy rarefaction scheme for communication? https://en.wikipedia.org/wiki/Rarefaction

We already do that during normal communication - people rarely say what they mean in the simplest way possible. We couch the message in all kinds of extraneous details. The tediousness of which is part of why some people want to use LLMs to generate messages for them in the first place. Stripping that out might ironically reduce the need: Of people come to expect their messages to be automatically stripped of fluff,…

It is possible to learn how to write concisely.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#177

Post by dkarras »

rpastuszak wrote: post #146
heliodor wrote: post #67

Earlier quoted context omitted.

The concept is right on point though: ChatGPT is excellent at writing "fluffy" pieces full of empathy, compassion, PR talk, politician speech, and any other wordy flavor you need. As an engineer who is unable to produce such writing, this tool is quite helpful!

> ChatGPT is excellent at writing "fluffy" pieces full of empathy, compassion, PR talk, politician speech, I think you're putting a bunch of completely different things in the same basket. Making the text more fluffy will not make it more empathetic, not only (but mainly) because there isn't anyone to empathise with the respondent. Our bullshit receptors are good spotting dishonesty, so it'll just sound cringe, putti…

>Making the text more fluffy will not make it more empathetic,

Of course not, but empathy in communication (not in action) is full of fluff. It almost requires it.

>Our bullshit receptors are good spotting dishonesty, so it'll just sound cringe

All "genuine" affirmation for the sake of empathy sounds cringe to me. I'd rather the doctor devoted their time to doing their job well instead of trying to make someone feel heard and validated - especially in an ER scenario where they are juggling critical patients. This tech can help with that.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#178

Post by mattkrause »

williamcotton wrote: post #176
vidarh wrote: post #166

Earlier quoted context omitted.

We already do that during normal communication - people rarely say what they mean in the simplest way possible. We couch the message in all kinds of extraneous details. The tediousness of which is part of why some people want to use LLMs to generate messages for them in the first place. Stripping that out might ironically reduce the need: Of people come to expect their messages to be automatically stripped of fluff,…

It is possible to learn how to write concisely.

You can learn to write concisely.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#179

Post by suoduandao2 »

firebirdn99 wrote: post #101
heliodor wrote: post #67

Earlier quoted context omitted.

The concept is right on point though: ChatGPT is excellent at writing "fluffy" pieces full of empathy, compassion, PR talk, politician speech, and any other wordy flavor you need. As an engineer who is unable to produce such writing, this tool is quite helpful!

Many comments are saying it's helpful. But what happens when it's widespread, everyone crafting Hallmark type empathic messages to everyone. Or once everyone has bots, then most likely they will all be sent straight to spam or something. We will be inundated with such messaging.

"Chatgpt, write an email based on these bullet points!"

"Chatgpt, summarize this email into a few bullet points!"

Somehow, I doubt we'll notice. The reason we don't send the bullet points email today is that the receiver will think we're angry with them. In University the engineers all needed alcohol as a social lubricant, if chatgpt replaces it I expect we'll trade hangovers for higher electricity bills and carry on.

Re: I’m an ER doctor. Here’s how I’m already using ChatGPT to help treat patients

#180

Post by elliotpage »

seabombs wrote: post #52
firebirdn99 wrote: post #44

This really feels farcical, and kind of like vc bros hyping up a product. Considering this was a serious issue of life and death of a patient, it was surprising how it read like a prank or like the scene from Harold and Kumar, where Kumar plays a fake doctor and saves an ER patient through sheer luck. I know some doctors are full of baloney but c'mon. I really hope maybe that we create better doctors that have a lot…

The author is a doctor, but also an executive at a health-focused VC fund, with several AI companies in it's portfolio. Helpful context to have before reading his article I think.

Gotta love a classic Conflict of Interest!
Post reply on HN