AI In Healthcare

Krystal Shay

Krystal Shay

Well-known member
Has anyone else been asked to sign a release form to let AI record your conversation with doctor, when you go for a visit? My husband had his wellness visit this morning, which they force that upon him too. But he can’t get his blood pressure meds refilled unless he goes to the doctor, which consequently is about the same time a wellness visit seems to roll around. The young doctor got a little ticked off with my husband when he refused to sign the paper for AI recording his visit. The doctor tried to tell him that AI saves him 2 hours of paper work each day. My husband told him, “no” again, and he was not going to sign it! My husband said he thinks the doctor cut his visit short.

I had the same thing happen to me, when I went to my NP last year. I refused to sign the release form to have AI record my conversation. I told them I want to keep my conversations private; doctor patient confidentially. My NP tried to tell me that it is private, and she too, wasn’t happy about me not signing it. She has since left, taking another job with a private industry as their healthcare provider, and now I have to find another Primary.:rolleyes:

My husband and I go to two different hospital corporations. I guess all Healthcare companies are heading towards AI. We don’t have anything to hide, but it still aggravates me that they want to record me using AI, and having AI transcribe the conversations.
 
Last edited:
It's annoying but I suppose I don't care. All my medical records are electronic via "MyChart" so I'm sure those can and will be compromised at some point.

I have read that AI will provide much more accurate diagnostic tools such as reading scans and x-rays. No "human error" so that will be a good thing. Seems like the AI evaluated cancer screenings are significantly more accurate than human interpretation.
 
Since we changed over to the Devoted Health Primary Care, all of our doctor visits are video conference, so they are probably recorded as they are happening.
It is fine with me if they want to do that. Our doctor visits now last much longer, usually over a half hour, because they ask actual health questions and we can ask any health questions that we have.

Our PCP that we had who was a local doctor never spent much time or effort at all, and didn’t even seem to care about our actual health.
Personally, I think he was maybe one of those Hindee/Indian doctors who got the fake certificates and came to the US to pretend to be a doctor. He had his nurses do everything.
 
Last edited:
Yvonne Smith said:
Since we changed over to the Devoted Health Primary Care, all of our doctor visits are video conference, so they are probably recorded as they are happening.
It is fine with me if they want to do that. Our doctor visits now last much longer, usually over a half hour, because they ask actual health questions and we can ask any health questions that we have.

Our PCP that we had who was a local doctor never spent much time or effort at all, and didn’t even seem to care about our actual health.
Personally, I think he was maybe one of those Hindee/Indian doctors who got the fake certificates and came to the US to pretend to be a doctor. He had his nurses do everything.
Click to expand...
Yes, it is hard dealing with them especially when your in severe pain off and on. Mostly it is speaking with the doctors and nurses over the phone or portal which is better for us than riding up and down on the roads. We live so far from our doctors...they are close to an hour away.
 
Since we're unlikely to be able to make AI go away, I don't have a problem with it being used for some of the more mundane work, but I am very concerned that it will be relied upon too heavily, as we have found in several legal cases, even as far as in cases brought the to Supreme Court, where attorneys have allowed AI to prepare the briefs without double-checking its work, only to find that fictitious court cases were cited. As I have said before, at this point in its development, AI does what I jokingly once told my students I'd do, "If I don't know the answer, I'll make something up."

I use AI to do some of the research for me in my web directory job, but I often find that it is wrong about a lot of stuff. Rather than being restricted by accuracy, the objective seems to be to produce a result that looks like it could be right. That concerns me when it comes to reliance on AI in medical decisions.
 
For a while my urologist had an assistant in the room who took notes for him on a laptop (I called him "Scribe.") For some reason, that went away. I have a Welless visit coming up in a couple of months. I just went online and saw that the system my doctor is part of (University of Virginia Health Care) has developed AI tools for use in parts of patient care. The medical school claims to have developed an AI tool that has deciphered how medications work inside the cells, and claims to have already identified an existing drug that might prevent heart attacks in some patients.

I've heard nothing about AI being part of office visits, and am pretty sure that--based on the routine communications they send out--it would be mentioned in a notice or one of their newsletters and not announced during an appointment. I have no problem with it. All of our records are in computers and--as Beth said--are available online, anyway. HIPPA protections still apply, hackers notwithstanding.
 
Ken Anderson said:
Since we're unlikely to be able to make AI go away, I don't have a problem with it being used for some of the more mundane work, but I am very concerned that it will be relied upon too heavily, as we have found in several legal cases, even as far as in cases brought the to Supreme Court, where attorneys have allowed AI to prepare the briefs without double-checking its work, only to find that fictitious court cases were cited. As I have said before, at this point in its development, AI does what I jokingly once told my students I'd do, "If I don't know the answer, I'll make something up."

I use AI to do some of the research for me in my web directory job, but I often find that it is wrong about a lot of stuff. Rather than being restricted by accuracy, the objective seems to be to produce a result that looks like it could be right. That concerns me when it comes to reliance on AI in medical decisions.
Click to expand...

I've read of others using AI and--because of their baseline knowledge--have identified outright BS in the results. When confronted, the AI system fesses up and "apologizes," as though it were intentional. Or maybe it's programmed to claim intent when it's just a case of Garbage In/Garbage Out and the user spots it. After all, AI is still programmed by humans, and the data it accesses is either entered by humans and/or created via systems that were programmed by humans. But as you said, it sure puts a pretty veneer over it.
 
Our Primary Doctors are the only ones that ask us to sign a release form for AI to record the visit. My specialty doctors; GYN and Dermatologist never does this, nor does my eye doctor or dentist. I have My Chart as well. I use it for echeck-in, making appointments, checking results, and to answer the questions and sign forms. They don’t always add the comments to My Chart what had took place in a visit. 🤨

I guess what aggravated me is the fact that my husband’s doctor badgered him for 5 minutes about why he wouldn’t sign the release. The doctor said it saves him 2 hours of work by letting AI record it. I don’t believe that for a minute. What is the doctor getting paid for? Isn't that his job to listen to a patient? Also, what the heck is all the admin people doing sitting around in the office then if AI is doing all the work?

I know we will be forced into it, but still. I'm crabby and old. 😀
 
Krystal Shay said:
Our Primary Doctors are the only ones that ask us to sign a release form for AI to record the visit. My specialty doctors; GYN and Dermatologist never does this, nor does my eye doctor or dentist. I have My Chart as well. I use it for echeck-in, making appointments, checking results, and to answer the questions and sign forms. They don’t always add the comments to My Chart what had took place in a visit. 🤨

I guess what aggravated me is the fact that my husband’s doctor badgered him for 5 minutes about why he wouldn’t sign the release. The doctor said it saves him 2 hours of work by letting AI record it. I don’t believe that for a minute. What is the doctor getting paid for? Isn't that his job to listen to a patient? Also, what the heck is all the admin people doing sitting around in the office then if AI is doing all the work?

I know we will be forced into it, but still. I'm crabby and old. 😀
Click to expand...
An eye doctor would be on the top of my list of practices to implement AI. My ophthalmologist told me that the scale for rating the stage of cataracts is subjective, and is unique to each doctor. Her "5" is not the same as another doctor's "5." So if I changed eye doctors, the new one would have no way of knowing if (or by how much) my cataracts changed since my last exam. I also pay extra to have a photograph of my eye blood vessels taken. I bet AI might detect slight abnormalities before a human could. Heck, I could see AI determine the specs for corrective lenses faster and more accurately than the "Better, Worse or About The Same?" process.

Regarding GPs: mine types in his computer as I am talking to him during the exam. He talks out loud as he is typing. I have to be careful to not talk to him as he does this, because no one can type and listen at the same time. I don't like that part of his process (it's probably required by the conglomerate that owns his practice), but I guess it's better than him trying to remember everything later... and I do like his ability to bring up past test results and have each one graphically displayed over time. Any method of recording things so he can focus on me would be of benefit.

I think using the label "AI" freaks us all out to a degree. If doctors said "electronically transcribe" no one would bat an eye. The only thing that makes me suspect is what is driving the need to ask permission to use AI? I have no idea what automated diagnostic tools any doctor might be using behind the scenes, and I would never deny his use of any of them. Nor would I ever think he needed my permission to use whatever he felt the need to. What in their eyes (or their lawyers' eyes) makes this one special?
 
Last edited:
I just skinned a few articles on this subject. Most of them address the ethics of Informed Consent regarding the degree of influence AI has in making medical care decisions. But one of them raised the subject of medical records serving a "secondary use" to train AI models. This means that AI might not fill a role in your doctor's health care decisions, but your data might be given (or sold) to third parties to fill out a database of real-life information used to improve the performance of AI in health care delivery and in medical research. The implications regarding our healthcare information now being in the hands of non-medical entities IT companies and AI start-ups who lack HIPPA processes, procedures and oversight are self-evident.

For those who like to wade through the weeds, here is a lengthy review of this topic by the International Institute Journal of Medical Informatics. [direct link to article] The gist I get out of it is that medical records data sharing is already happening in some places, and that adequate consent and safeguards are a "work in process." (Funny how ethical considerations always come after Execution.)

I don't know if data sharing is a component of those doctor visit agreements or not. The data-sharing issue should be something apart from using AI as a tool by the doctor, or it might all be wrapped up in one consent agreement.
 
Last edited:
Beth Gallagher said:
It's annoying but I suppose I don't care. All my medical records are electronic via "MyChart" so I'm sure those can and will be compromised at some point.

I have read that AI will provide much more accurate diagnostic tools such as reading scans and x-rays. No "human error" so that will be a good thing. Seems like the AI evaluated cancer screenings are significantly more accurate than human interpretation.
Click to expand...
I have mentioned that the specialty of Radiology (conventional) may be eliminated by AI in an earlier post. It has been shown many times that AI-read scan and x-rays are far superior to those read by humans. I, myself, have been the victim several times of misread x-rays and scans. I think many times it is carelessness, lack of training, or some other mix up. I don't think I would agree to having conversations recorded in any form, much less by AI. Written records are not particularly secure, but recordings are even less so. Another step in the decline of American healthcare.
 
Last edited:
April Vee said:
Any and everyone should refuse anything to do with AI!!!

Im nervous emailing friends of mine who use GMAIL! (totally AI run)
Click to expand...
I heard a security expert" talk about Gmail being the least secure email system. All free web-based email systems are not secure. That is why they are free. There is something gathered and sold that finances the system. This expert said that Gmail copies and stores every attachment sent by Gmail. She said nobody outside the company knows why this is done or what is done with the data, but apparently it has been documented that it occurs. Anyone, therefore, who sends or receives medical data by Gmail is building some kind of file with Google. Disquieting to me.
 
AI transcription wouldn't bother me a whole lot, except that I know it wouldn't stop there. I can see how it would take a lot of the tedium out of the job, especially for EMS and police, who are also using AI in many areas. As a paramedic, I would have to jot notes down because after each call, I would have to fill out a report on the call that included patient progression or regression, what we did for them, and when. Since some of our transport times were 40 minutes or more, there was a lot to record. In some calls, the more hectic ones, I might have notes jotted down on a few different scraps of paper, EKG strips, bed sheets, or even my elastic gloves, and my handwriting isn't great to begin with. On many calls, I used a recorder in the event that I got stuck or couldn't read my own writing. With AI, a paramedic could simply say everything that he was doing and have the AI write the report at the end, reviewing it for accuracy. Of course, there would be medics who, after experiencing a few calls where the AI report seemed accurate, would neglect to review, and an inaccurate report would be filed, which is not to say that everyone's reports were accurate in my day.

Several years ago, a class action suit was being filed against the manufacturers of a mesh that was used in hernia surgeries. Having had a hernia surgery in which the physician had said that he had used a mesh, I asked for the records to see if it qualified to be part of the suit, having had doctors comment on what a mess my stomach was after a botched surgery for a strangulated hernia, another hernia surgery, and a gallbladder removal. The surgeon had retired, and when I looked at the report he had filed, it was clear that he had mixed my report up with someone else's. It said nothing about using a mesh and even included some text about burns on my chest, which had nothing to do with me. So, even without AI, there are medical report errors.
 
Last edited:
Don Alaska said:
I heard a security expert" talk about Gmail being the least secure email system. All free web-based email systems are not secure. That is why they are free. There is something gathered and sold that finances the system. This expert said that Gmail copies and stores every attachment sent by Gmail. She said nobody outside the company knows why this is done or what is done with the data, but apparently it has been documented that it occurs. Anyone, therefore, who sends or receives medical data by Gmail is building some kind of file with Google. Disquieting to me.
Click to expand...
Whether it's email, cell phones, texts, etc. we are the product. You can no longer buy a Smart TV that does not spy on (and report back) your viewing habits. That is why they are so cheap...we offset the cost of installing their spying device in our homes so they can monetize our information. I wouldn't trust one to be on my network and not be reporting other activity. There are all sorts of goofy work-arounds (like getting a commercial video display with your own external sound system and streaming with a legacy X-Box), but that is cumbersome...and still imperfect security.

I've thought of getting SmartLink as a second internet service dedicated to video streaming. At least is would not be on my regular Wi-Fi network. For now, I just have antenna tv.

Gee, where is congressional oversight?
 
"On Dec. 4,2025 we detected unauthorized activity on certain activity within our networks.

This information may have included your name, social security number, state identification number, treatment information, date of birth, and medical record number." This from a medical provider.

This makes twice in the last year that I have received a letter like this. They say remain vigilant, blah, blah, blah. They offer 12 months credit monitoring and identity protection services.

A few years ago, the computers at NCDMV were hacked. I received a letter from them saying about the same thing.

Everything is already out there so it can hardly be worse. Maybe AI will not be as bad. 🤷‍♀️
 
You must log in or register to reply here.
Back
Top