Only 2% of U.S. adults turned to AI for healthcare information in 2024, and today the number is 61%, according to Salesforce's Connected Health Consumer report, a survey of 3,200 consumers worldwide aimed at better understanding how the rise of agentic AI is reshaping consumer expectations, attitudes, and demands within healthcare -- specifically patient experience.
Here are the four key findings of the 2026 Connected Health Consumer Report:
The report found that 60% of patients put off care because of scheduling friction. Patients are desperately seeking a better engagement experience with healthcare providers. The current multichannel engagement models fail to deliver a good experience, with 49% of patients noting abandoned calls after 10 minutes of holding.
The online experience is not better, with 46% of patients labeling websites as confusing and difficult to navigate. One in 6 now says ease of digital access is a deciding factor when choosing a provider. Record sharing is another major deficit for improved experiences. More than half of patients (60%) say poor record sharing between providers means repeating the same medical tests.
And 66% of patients have run out of medication while waiting for their prescriptions to be refilled.
Nearly 7 out of 10 patients would rather have access to 24/7 help via AI agents versus waiting to speak with a person during standard hours. Bad scheduling experiences are driving patients away from care, and AI agents are helping. Patients want proactive care, with 83% of patients interested in self-enrolled programs that can provide healthcare recommendations.
[...] Patients are ready to share health information for safer, more proactive care, with 73% saying they trust AI to flag potential drug interactions before picking up new prescriptions. The report found that 63% of patients want automatic reminders for medication use, 66% want AI agents that suggest prevention screenings, and 54% agree that AI agents can help them feel more secure in their provider's care.
Patients also look for AI agents to help create a smoother hospital-to-home handoff. More than 77% of patients would highly value an AI tool that can simplify the transition from hospital to home care. In fact, 72% would trust an AI agent to create a personalized follow-up schedule based on their complete health history.
The report found that the shift toward greater trust with AI in healthcare has grown significantly since 2024 -- 64% of patients would share their full medical history with AI for faster diagnosis, and only 15% would not share any data with AI agents. Patients are three times more likely to trust an AI agent integrated into their doctor's secure portal than one on a public chatbot or general website.
Patients do want human oversight in order to increase adoption of AI-driven support. The concern is if AI agents can handle sensitive health interactions, including concerns around accuracy and privacy of health data. Patients do not want AI agents to act alone, with 88% requiring evidence of human oversight before accepting AI for administrative support, and 90% expecting the same level of supervision for medical support. Patients also want the option to escalate to human support as an essential trust requirement. Patients also want proof behind AI-generated recommendations -- traceability and accountability.
The importance of AI agent traceability and accountability is one of the 12 rules for agentic AI successful transformations.
(Score: 5, Insightful) by SomeGuy on Sunday July 05, @03:18PM
The story sounds like complete made up bullshit.
Reads more like what people want than what people do. Hell, even I WANT talking paperclip heads sticking out of toilets to replace government leaders, high ranking CEOs, uber rich investors, and such. But that is not going to happen. Ever. No matter what.
Probably just more fake hype to try and make people accept this crap as it gets shoved down their throats.
(Score: 5, Insightful) by JoeMerchant on Sunday July 05, @03:46PM
Google goggles have been progressively using more and more AI to generate search results for the past 20 years. Recently they branded their LLM interface as Gemini and sometimes indicate when Gemini has been automatically engaged for a response. But, behind the scenes this stuff has been building since the earliest days of search vs SEO.
🌻🌻🌻🌻✌️ [google.com]
(Score: 5, Insightful) by looorg on Sunday July 05, @04:45PM (15 children)
Seems to be the current trend. People ask AI about everything, share everything with best digital friend. How to invest money, who they should vote for, whom they should date etc. So asking it for medical advice is probably a logical step in the same staircase of idiocy. God forbid you ask the medial professionals, friend AI knows better.
Right. So people are so stupid as they see healthcare as just any kind of business transaction. Why can't those surly nurses be just like the greeter over at Walmart or the cute girl in the clothing store ... Or why am I not a priority patient and the entire hospital revolves around me and my needs before all the other people that are there. They clearly are not as sick as I am. Me me me me me ...
We can't even get the AI to create a proper and working staffing schedule. At least not one that takes into account what the staff wants. But if you didn't have to include staff wants or needs but just created a schedule you wouldn't need to involve the AI at all. Patients clearly have overestimated what friend AI can do for them, except in Google Fantasy Land (or whomever provides the AI service).
"A good experience". They want to keep you alive! Not sell you another massage. Or whatever product. I guess in the US the line might be a bit thinner then in some other countries. They are more into double dipping it seems.
Not sure what phone system they have where they did this survey. But over here if you don't get a human on the line in about a minute the machine kicks in and the voice asks if you want it to call you back. Then it estimates a time when it will call you back. You even get this option if you are in the queue. It has never failed to call back in the allotted time slot. This doesn't even involve AI. Cause that have been around for years. Years before the current AI explosion. I have not looked into ever but I assume it's just part of the phone exchange system.
What is friend AI going to do to you except talk to you? Come back when the Robot-overlords actually gives you healthcare and can do something beyond just talking to you. Come back when it can take a blood sample or run some basic testing on you. What they want then is some kind of AI-nanny to tell them what to do? Make it all better ...
That seems somewhat unrealistic. So not only do they want a state of the art AI system at their beck and call but they want humans to. The once in charge of the hospital do want AI so they can have less staff engage in this talking to patients part of the business transaction. Not to offer both. That won't save them any money. It's the double dip. Friend AI and a human in the loop. Just cut out friend AI and go for more humans instead.
Which is not really a hospital problem. It's a manufacturing and logistic problems. We are consuming drugs faster then they make them and it seems to be an issue having them at the right pharmacy right when you need them. One do wonder if patient A have a prescription that is going to be more or less forever and they always pick them up at the same pharmacy why don't they make sure there are a new batch at the drugstore for that person when she is supposed to pick it up cause the one she had will have run out. If AI can solve that logistical issue it would at least be good for something medical related.
That number is just horrifically low. As with all other things I wouldn't ask "friend AI" for advice about anything. Least of all I would share any kind of medical information with it and its owners. That are totally not keeping anything of what you have shared with it ... wink wink nudge nudge ...
(Score: 5, Insightful) by JoeMerchant on Sunday July 05, @06:16PM (3 children)
>God forbid you ask the medial professionals, friend AI knows better.
Look at you: with easy access to medical professionals. My earliest available appointment (from last week) with my new to practice GP is September 11, and they wonder why people "abuse" the Emergency Room system.
Then, when you get access to a medical professional, are their answers any better for your well-being (on average) than "friend AI"? How often is "friend AI" steering you toward an open slot on their surgery calendar to maximize their income? I literally had a "medical professional" make me sit in the waiting room while my insurance benefits were gathered for him, I heard through the thin waiting room wall "oh, yes, he's got great coverage, send him in" then I sat and the "medical professional" read me the top scary result for my condition from Google as incentive for me to get into his surgical suite tomorrow morning. Not all medical professionals are like this, but in my experience the rate is hovering somewhere above 50%.
I want my medical professionals to use "friend AI" to help diagnose my symptoms, but I trust them about as far as I can throw them, and that's getting to be a shorter and shorter distance as the years go by.
🌻🌻🌻🌻✌️ [google.com]
(Score: 2, Interesting) by Anonymous Coward on Monday July 06, @02:08AM
For me, trusting a medical professional is as trying to deal with a financial advisor. So far, all financial planners have done for me is couple me with sub-performing offerings, collected their commission, then left the area.
Four words would have saved me a lot of money and frustration:
Vanguard. Fidelity. Index funds.
Key tipoff word for losers: "managed".
They all use AI these days, and "the market" already prices that in. All you get is more fees to pay.
Oh, and you get lots of handshakes with men wearing suits, as well as lots of legal disclaimers to sign.
I figured each handshake costs several thousand dollars. But you do get a fine business handshake.
The biggest ripoffs were those portraying themselves as experts in fancy buildings and immaculately dressed.
(Score: 2) by looorg on Monday July 06, @06:28PM (1 child)
How are you being helped by the AI? Does it help to know what the AI think is wrong with you? It can't do anything about it. The medical staff isn't going to trust friend AI or Google or whomever. They are just going to run their own tests anyway. AI in that regard is selling you a false sense of security. One day it might be that you can come to the hospital and just hop onto the Probulator and it will tell all that is wrong and fix it (or assign you a job as a delivery boy that you don't want).
This is why the queue is so long. People come there for everything. Even things that are not emergencies. Then the frustration rise as they get passed by patients that are actually sicker then they are. It's not called the waiting room for nothing.
One day friend AI might be incorporated to the hospital system and might actually offer something useful. Today? No. Not today. We are not there yet. It's just a talking layer acting as a barrier between humans and human interaction. Until the day comes that AI can actually do something beyond talking. Or there are idiotproof devices that the patients can just walk onto that does all the monitoring and testing. The machines won't replace the humans. They are just adding more layers. So far only layers of frustration and unrealistic ideas and hopes.
I'm waiting for the step where the diagnostic becomes so good that it will predicts future ailments. The problem then might be that we can't do anything about it. AI analyze blood sample and tells you with some kind of 99% certainty that you'll develop something. Which we can't cure or do anything about. What then? Enjoy worrying about things we can't do anything about for the next few decades. Perhaps we'll have a cure by then if your disease is common but if it's rare you are probably f*cked. But there are no guarantees. Will you be happy or feel better then? Not knowing all might be for the better in some cases.
(Score: 2) by JoeMerchant on Monday July 06, @07:05PM
>How are you being helped by the AI? Does it help to know what the AI think is wrong with you? It can't do anything about it.
Funny, I had sharp pains in my arm a couple of years ago, coupled with high blood pressure brought on by rather extreme stress. I went to the local "accessible" Care Spot (for-profit clinics run by one of our (redundant to say) corrupt conservative stage governors). Took nearly 1 hour to get face-to-face with anyone asking me anything about anything other than insurance and ability and willingness to pay, then they did perform an EKG and tell me: "We don't see anything here, but we can't see all signs of heart attack - for that you'd need to go to a cath lab and have them draw and analyze your blood" which they would then analyze on site and determine whether or not you "need" a high reimbursement procedure... see elsewhere on this thread regarding: "see a surgeon, get surgery."
So, I'm sorry, which is better? A) The AI which could answer my chest pain / BP questions in minutes, B) securing transport to a local Care Spot for $100 out of pocket shared cost to have them tell you the same things, with the same inability to intervene? or C) Driving an extra hour to get to the cath lab where they can do meaningful interventions and risking your life with their helpful procedure which you may or may not actually benefit from?
I will say: you can get a Bluetooth connected 6 lead EKG device to show you the 6 traces on your phone and auto-diagnose A-fib for less than the shared cost of a Care Spot visit - again: information available where you are, in minutes, rather than driving and waiting and hoping that somebody competent is going to be directing you to an appropriate medical professional when you get there?
>Until the day comes that AI can actually do something beyond talking. Or there are idiotproof devices that the patients can just walk onto that does all the monitoring and testing. The machines won't replace the humans. They are just adding more layers. So far only layers of frustration and unrealistic ideas and hopes.
Talking is the first step. Until the AMA starts releasing more M.D.s into the aging population, your access to "qualified people" will continue to be a multi-tiered system with the bottom tiers literally unable to even prescribe an aspirin. If there's any doubt, I consider billing, collections, benefits verification, and ability to pay admissions screeners to be scum of the earth, lowest of the low... and who do you see first and last in the Medical Care experience?
🌻🌻🌻🌻✌️ [google.com]
(Score: 1) by shrewdsheep on Sunday July 05, @06:49PM (4 children)
I have to say, often it does. Making a diagnosis involves putting together a pattern of symptoms and weighing differential diagnoses. This is basically an optimal application of LLMs. The problem is to describe symptoms properly, at best in medical terms. As with coding/product assessment/literature research, LLMs can be very helpful for those who know how to use them.
(Score: 5, Interesting) by JoeMerchant on Sunday July 05, @07:46PM (2 children)
It's an adjustment - medical doctors (as opposed to PhDs) are some of the most widely revered and respected "smart people" in our societies, had to be to keep up with all "the knowledge." But, even 50 years ago, the full breadth of "the knowledge" was beyond the capacity of the smartest human beings, so they specialized, but that only works to a limited extent, and when the generalist refers you to the wrong specialist, things can go less than optimally very easily.
These new babbling idiot machines that can't get the number of fingers on a hand drawing correct, are, in fact, better at some aspects of accessing "the knowledge" and matching patterns than people ever will be; and at least in the US with our AMA training the MDs that they are Uber Alles Menchen in command and not to be questioned... it's going to be a hard cultural shift for many of them to humble their $400K starting salary selves to the fact that their patients' phones really do know better than them, particularly if the patient has spent hours refining the results with appropriately specific information about their condition into the pattern matchers at a fidelity of detail that's simply impossible to communicate in the 8 useful minutes of a 15 interview appointment slot.
🌻🌻🌻🌻✌️ [google.com]
(Score: 4, Informative) by darkfeline on Sunday July 05, @09:25PM (1 child)
Based on what I've heard from a family member going through med prerequisites and dealing with the abject mental... deficiency of the professor, I'd honestly trust AI more. e.g. a chem professor who can't do basic addition and grades students down using their own incorrect calculations.
The medical program is more about artificially limiting the supply of doctors, rote memorization, and gatekeeping based on how much bullshit you're willing to wade through, than intelligence
Join the SDF Public Access UNIX System today!
(Score: 2) by JoeMerchant on Tuesday July 07, @09:51PM
SSI sent us to a psychologist to evaluate our (profoundly autistic) sons for disability. Yes, our sons are disabled with intelectual and developmental differences, but this guy and his assistant doing the evaluation were barely literate / functional themselves. The official (paid for by your tax dollars and mine) report was a copy-pasta hash up of mostly other peoples' evaluations, outright inaccurate statements of facts, and deeply nonsensical where they tried to fill in the blanks between the copy pasta. That was for our more functional son, the one less verbal didn't get anything written on his report that didn't come from our interview questions or other not-applicable copy pasta and gibberish.
2 years later the genuises at SSI demanded a re-evaluation of our younger son (at your expense again), this time the lottery got us a pretty nice sharp psychologist - not an A+ speller, but what he wrote at least made sense and was applicable to the person being evaluated.
🌻🌻🌻🌻✌️ [google.com]
(Score: 2) by Bentonite on Tuesday July 07, @04:45AM
You want an expert system being used - not a useless LLM - using such chatbots, that are only practically suitable for entertainment purposes, to try and diagnose medical conditions, is extremely reckless.
Instead, an expert system is actually suited to the task, as after entering the symptoms, a list of likely conditions are output, based off the high quality inputs of previous diagnoses (importantly, if the output does not list the later determined condition, it is actually possible to work out why the expert system didn't include that in the output and then fix it).
(Score: 4, Interesting) by Anonymous Coward on Sunday July 05, @07:31PM (2 children)
> > And 66% of patients have run out of medication while waiting for their prescriptions to be refilled.
> Which is not really a hospital problem. It's a manufacturing and logistic problems. ...
Family members on meds have run out. There's no logistics problem (in the cases I'm familiar with), it's purely due to poor behavior of the doctors, their staff, and/or their poorly-managed offices.
The dance goes like this:
The patient (my family member) calls or goes on line to their pharmacy to request a refill by automated system (touch tone signaling for those without internet). If doctor authorization is required, the pharmacy contacts the doctor's office. This is usually a couple of weeks before they run out of pills.
Pharmacy doesn't get a response from the doctor's office, robo-calls the patient and tells them to contact their doctor.
Patient (now somewhat impatient!) talks to the doctor's office or leaves a message requesting a refill of the previously prescribed medication. Usually it's leaving a message, few medical offices answer their phones anymore. Rarely does the office call back.
The doctor (or doctor's office) doesn't approve the prescription renewal in a timely manner. This triggers the patient to call the doctor's office (many times in some cases) until they finally get through.
Of course the pharmacy isn't supposed to issue a refill until they have correct orders from an MD to do so...although sometimes the pharmacy can be convinced to take pity, bend the rules, and refill anyway.
Touch wood, I'm not on any meds (yet), so I've only experienced this incompetence at arms-length.
(Score: 5, Insightful) by JoeMerchant on Sunday July 05, @07:52PM
> poor behavior of the doctors, their staff, and/or their poorly-managed offices.
This nearly killed my wife and firstborn on the day he was extracted by emergency C-section.
Office staff that won't communicate messages to the MDs. MDs that overbook themselves beyond reasonable expectations of human endurance. Office managers only focused on billing and collections with zero awareness of patient care. And these were the most highly sought in-demand ObGyns in Miami-Dade county at the time. During our second pregnancy I made a simple demand: your body, your choice, you want to go back to the four ring circus of women who nearly killed you, you can, but I will not be setting foot in those offices ever again. We got another ObGyn in the same building, he was astonished "women are always leaving me to go to them, why are you coming this direction?" Quite simply: because they're too popular (and greedy) to deliver adequate care.
🌻🌻🌻🌻✌️ [google.com]
(Score: 5, Interesting) by Common Joe on Monday July 06, @07:12PM
This.
I'm marking you informative because this describes pretty well what my best friend goes through occasionally. He's paralyzed. I didn't realize until he had his accident, but being paralyzed can be very, very painful -- all the time. So, he requires specialized medication that he must take at specific times every day. When the doctor's office fucks up or the insurance company gets snippity, he doesn't get his medication in time. This spirals him quickly into withdrawal and pain. Once he gets his meds, it can take him two weeks before the meds work well enough again that the pain is back under control enough to work.
Imagine trying to hold down a job when this happens two or three times a year.
(Score: 3, Interesting) by Common Joe on Monday July 06, @03:13AM (2 children)
There are so many things wrong in your write up, I don't even know where where to begin, so I'll just generally address a couple of things.
I've used AI because doctors are incompetent, they give out bad information, they have no desire to engage in searching for my health problems, and it's damn near impossible to get an appointment. Over the years, I've had them berate me, yell at me, lie to me, leave open door wide while I'm undressed, walk out in the middle of our appointment and not come back because they didn't want to search for the reason I was in pain, tell me they have no capacity to search for my ailments, and making appointments is sometimes literally impossible on the computer, over the phone, or in person. Currently, I'm in the middle of a six month wait just so I can get a skin screening.
My father just died. He was in the middle of a near year-long wait for an appointment with a Parkinson's doctor just so he could have his medication adjusted. There was no capacity for him.
My best friend is a paraplegic. You would not believe how painful that condition can be on a daily basis. For that reason, he consumes a lot of prescription pain meds. However, there is an extremely narrow window for him to get pain meds which is allowed by the pharmacy / insurance company. And sometimes the paperwork gets screwed up by others. So... no meds for him. At that point, he has had to go several days without meds which means he goes through some ugly withdrawals (because he shouldn't just stop some of these medications suddenly) and once he's back on them, it takes about two weeks before the meds take full effect and his daily pain is manageable again. When someone else fucks up, that means he's unable to work for about three works. Hard for him to hold a steady job. That happens way too often.
So, yeah, when there is no where else to go, people are going to AI. Not much of a choice anymore because our health care is bonkers.
(Score: 2) by looorg on Monday July 06, @06:07PM (1 child)
That is your experience with the interaction, that is clearly different then mine. I'm sorry that you find the system broken, or that your system is broken. It's a bit weird, if you live in the USA, that you are supposed to have the best system. The one with the most resources. But judging from most of the post here it seems to be utterly broken and nobody appears to be happy about it. Yet nothing much changes or is done to change things. That isn't my perspective since I live in Europe. I think a lot of the problems with your system is that you have turned it into a business instead of a national service. Not that our system is perfect by any stretch. It's just if I wasn't ultra-rich, and I'm not, I would prefer our system to yours.
I'll disagree on the incompetent doctor part. The area is now so complex you can't know everything. But I doubt they are incompetent. AI doctor in that regard isn't from my experience better. It's bad that your system sucks. I know this won't make you feel better but the system could be different. If I leave a message with my doctor he'll call back the same day or the day after. If I have to wait they start sending me to places further away at their expense. If I have to wait more then 90 days I can go to another country in Europe and they'll still pay for that. They'll renew all prescriptions in a day if it's an ongoing one. Just leave a message (phone, email, app or even write a note on paper and drop off at their office) and they are renewed the same or the next day. The problem here is getting the pharmacy to keep stock. I wish they would use "AI" to keep better stock, so for me it's a logistical problem.
Working at a research hospital we are testing all kinds of weird AI systems these days. For everything from staff and patient scheduling, they are all crap so far. Not better then before and they didn't unlock resources as staff have to manually go in and do things and monitor it. It's just more layers that are currently causing more problems then they are solving. We are all guinea pigs in their AI-fantasies. Testing to see what works and fails. So far it's more or less only failures and frustration. $ we could have spent on more humans and more interaction with the people that need help.
I'm sure someone wants to sell us a system that will do AI doctor where you can just talk to a machine. The problem is that a lot of patients don't know what is wrong with them or they don't know to how to describe things. So it will be vague. If people find it frustrating to just be stuck in the phone-maze of "press 1, press 2, press # ..." being stuck in the AI loop of questions will suck just as much. As AI in that regard can, and will, be as clueless as you appear to find the human staff. And eventually they'll have to send you to a human instead, cause robotics just are not there yet nor is the AI. That interaction will also just increase the levels of frustration as they'll have to explain things again. Or they get frustrated that they are kept waiting. So from my perspective AI adds nothing but more trouble so far. In all aspects of care.
The best use so far have been the systems that analyze results of different kinds -- radiology have a lot of use for it. But the results are not great. Then everything have to be rechecked by humans anyway. But it's a good thing to have running in the back. But it's not a replacement. It's not really solving anything groundbreaking, yet. I assume we are just going to get more and more AI fueled medical systems, monitoring equipment etc. But so far? It's just a layer between human to human interaction and it's in the way. It's creating unrealistic expectations for all involved.
If you chose AI over nothing or a broken system that is sad. I just don't see how in the end AI will help you. It might give you advice, but it might not be correct and as of yet it can't help you beyond words.
(Score: 2) by Common Joe on Monday July 06, @07:04PM
Have I got a surprise for you. I live in Europe too, but I lived in the US for a very long time. I've been in Europe for over 10 years now, but I'm still American. My dad always lived in the US, though. So, I've got stories from both sides. And the stories are same between the two places. Mostly crap doctors.
As for incompetent doctors in the US: I had a doctor diagnose me with IBS just by looking at me for three seconds. He was correct, but it was a dangerous diagnosis because IBS is a diagnosis of exclusion. If he had been wrong, I could have continued eating things that would have killed me. Not an exaggeration. In the US, I had a doctor who promised surgery for a minor procedure which is sensitive, so I won't get into it. When I arrived to the doctor's office on the day of the procedure, they said they didn't do it. Waste of time and they charged me for the visitation. Not worth a fight in court. In the US, I had a colonoscopy. The anesthesia left me loopy. The doctor told us that as long as my wife was with me, it was ok to call a taxi to bring me home. (My wife doesn't drive.) After the procedure, they didn't allow her to call a taxi due to liability. Resolving that while I was in no condition to make decisions was not fun. Incompetence.
If you think those stories are bad, you should listen to the ones from my paralyzed friend who sees them more often than I do. (He lives in the US.)
As for Europe, there are incompetent doctors here too. I've had an asthma doctor diagnose me both as having asthma and not having asthma -- same doctor, same appointment. Also in Europe, I had another doctor insist I had go through a full multi-round diagnosis for asthma every six months. (My asthma is not bad at all.) In Europe, I had a doctor yell at me because he couldn't find a reason behind my back pain. Incompetence.
No capacity for new patients? Yeah, I faced that here in Europe for me while at the same time, my dad faced the same thing in the US.
Don't get me wrong. I've found some good doctors in both countries, but they are damn hard to find. And the waiting time is long in both countries.
So, a few months ago, I had symptoms which I couldn't figure out, so I had a dialog with an AI. I gave it symptoms and I told it to diagnose me. After some back and forth with a lot of questions, it guessed milk protein intolerance. Sounds reasonable based on the symptoms. (Lactase did not help. I tried.) Is it right? Who knows? But was better than going back and forth to various doctors for several months. In fact, I feel better about going to a doctor now because I can now say, "Test me for milk protein intolerance." Very specific. Very quick. Then I'll know. (I'm overly simplifying. Don't think that I believe AI was my savior. it wasn't that easy to figure out. I spent a lot of years figuring out my triggers. I have several. Milk products are just one trigger. And not all triggers are ingested. Some are triggered by things in the air. I'm not going to get into details.)
(Score: 5, Touché) by epitaxial on Sunday July 05, @06:50PM
I listen to RFK Jr. You can’t go wrong with raw milk, colloidal silver, and ivermectin.