A Country Doctor Writes:

My news sources (you can’t just say “the news” anymore, since everyone can choose what angle/bias/viewpoint/political spin they want to believe) are predicting that AI will very quickly replace midlevel professionals. Even in healthcare, areas like radiology seem to be pretty vulnerable to the advancing non-human capabilities.

In my practice I see many elderly people who are unable to use technology to interact with professionals outside their own homes. House calls are appreciated by patients and caregivers and give me as a physician a clearer picture of what my patients lives are like. And I hear every day how people have qualified for in-home help but the agencies don’t have enough staff to provide it. So I believe home health nurses and home health aides can feel fairly secure for quite a while.

And even if a doctor sees a patient in an office setting, getting a clear medical history is something not easily automated because of the way older people communicate and describe things. And when it comes to communicating a diagnosis and treatment options while also exploring each patient’s ultimate life goals seems nearly impossible to entrust to AI in my view. And the more diagnoses a patient has, the more judgment calls need to be made about where to set the priority, because what the patient primarily wants may not be exactly what the treatment algorithms would suggest.

There are many interesting examples in the literature of how two different doctors can prescribe the same medication and get totally different outcomes. This boils down to the personality and style of the doctor. Maybe you could teach AI what usually works, but patients need to see the face and know the name of their provider and it would be dishonest and immoral to make up fake names for AI “doctors“ communicating with patients.

As a side issue, liability in medicine has been a big concern, and I just read in my newsfeed today that in some places manufacturers of driverless cars will be held liable for traffic violations and presumably personal injury cases. I don’t believe the issue of liability with AI involvement in patient care is anywhere near settled.

When I look at my WordPress blog stats I often see what people call evergreen posts popping up, pieces that always seem to find relevance. But after work this afternoon I saw a title I almost didn’t recognize. It turned out to be from 2019 and it brought me back to a time of change in my life. But it also reminded me of how a lot of things have gotten tougher in life lately in so many more ways than just what I wrote about almost seven years ago:

Despair and Happiness in America and in Medicine

I have many patients who google their symptoms in order to come up with a diagnosis. I also have many who google tests and treatments. Right now I have very few who use ChatGPT, but one in particular seems to have an ongoing relationship or dialogue where they exchange medical history, prior treatments, new treatment ideas and tests to request that I, the primary cre physician, order.

What I put in a patient’s chart is protected by HIPAA, the privacy law in this country. That law doesn’t apply to ChatGPT from what I understand and that doesn’t seem right to me.

ChatGPT knows the identity of my patient and their medical history, their test results and medications. I have only used the app for writing purposes just to see what it can do, but I don’t see any consents or privacy warnings when I do that. I actually suspect ChatGPT could “learn” and then share my ideas or concepts that I asked it to flesh out for me.

I think the use of AI for medical advice when health information is entered into its vast data base needs to be regulated or at least should require informed consent about what could possibly happen to individual patients’ otherwise protected health information.

I happened to read an article about the use of pilocarpine for age related difficulties reading up close. As a primary care physician with an interest in ophthalmology and a lifelong amateur photographer, I knew darn well that these ancient eye drops constrict your pupils and therefore give you a better depth of field, just like when you choose a smaller aperture on your camera. The same phenomenon is involved when I have to pull out my reading glasses instead of my progressives, trying to read the menu in a dimly lit restaurant that forces my pupils to dilate to pretty much see anything. It never occurred to me that this ancient drug could actually be used in lieu of reading glasses.

Old drug with new application, and obviously this particular formulation is patented by Abbvie.

https://en.wikipedia.org/wiki/Pilocarpine#:~:text=Pilocarpine%20was%20isolated%20in%201874,the%20South%20American%20plant%20Pilocarpus.

https://eyewiki.org/Vuity