Back when I was born, my doctor would come to the house. He (always a white “he”) was the knower of all things medical and, no matter what it was, cold, asthma, etc, he was the trusted resource and the person we counted on. If you had a question, he had an answer. I was born in the 1960s for context.
Let’s jump to the mid-late 1990s. The interwebs became a thing. Now, most information (and some disinformation but not tackling that here) was accessible to anybody who knew how to look for it. Using WebMD to check symptoms, Google to answer questions, Quora, Yahoo Answers, it was all there. If you had a question, there was now more than one way to get an answer.
Now, let's leap to the 2020s. In 2022, ChatGPT was released to the public. I can ask it how to interpret my reports, generate graphics, and translate any text into anyone's voice and personality (or language). It’s extremely capable in 2024, and it’s still so early. I have uploaded my symptom tracker data (I track and score my symptoms, diet, and activity every day), my imaging reports, my diagnoses, and (anything other than medical imaging) it can read it for me, analyze it, and make recommendations. I can upload multiple data sources, and it will pull them together for me. And no human being can ingest all of the data that ChatGPT (there are other AI engines, but I am just using ChatGPT for this discussion) can and make actionable sense out of it. So now, I just ask a question and get an answer, much like our species has been doing since we got here. (FYI, it can’t read imaging yet, but that can’t be far off. So I upload my written reports of my scans, not the images.)
If you agree with my semi-timeline, it begs the question, how has (and will) your relationship change with your doctor in the future? And, how will a doctor's education have to change to meet the shift in the patient-doctor relationship?
Let's go back to my Dr. in the '60s. As I said above, he came to the house or we went to his office for anything. He was THE guy. If he said take chicken soup, you did it. Everyone deferred to him because he had a medical license (and a whole bunch of other patriarchal reasons that we don’t need to go into here) and we didn’t. His word was pretty much gospel. He felt respected and revered. It’s called a God Complex for a reason, and there was no reason for a doctor not to feel like a god.
Your question, his answer, case closed.
In the 90s, things started to change, and the change went even faster in the 2000s. Dr. Yahoo and Dr. Google graduated from medical school. People realized they had greater access than ever to information and “ePatients” were born, people who viewed themselves as partners in their own care, not just recipients of a doctor's care.
In my appointments, my doctor would tell me I had an issue, put me on a med or whatever, and I’d run home and look it up to see what I could learn. I would then discuss and understand. If needed, I would find an online community to understand what the doctor had recommended AND to find out what else I could be doing, or how to deal with the side effects of whatever treatment was recommended.
Your question, his answer, your research.
Now we get to the roaring 2020s. ChatGPT went public in 2022 and will be the most radical catalyst for change in the doctor-patient relationship ever (that we have seen). Now, in addition to your doctor making recommendations, you can add a lot of layers. Do you want to see how your symptoms are trending? Upload the data and ask for an analysis. Do you want to map your symptoms to certain behaviors? Do you want to understand your imaging reports without waiting for a doctor's appointment? Upload the results. I did and it was great. I will always take what it tells me as a starting point (I question most things), but ChatGPT can tell me things and synthesize my data in a way that my doctors can’t. There are just some things that machines are better at than humans and synthesizing huge, multiple, and disparate data sets is one of them. This is a big generalization, but most doctors treat according to the averages and the studies, not the individual. AI can change that.
So now, when I see my doctor, I already have answers to most of my pressing questions such as what my symptoms mean, how my imaging looks, what I need next, what else I can do or be thinking about, etc. My doctor's appointments have evolved from a lecture from the doctor to a participatory discussion involving me, my doctor, my AI sources, community input, etc. But the conversations are much better, deeper, and more focused. They will ultimately help me live my best life.
Your question, your doctor is one of a few inputs.
(Note: AI is only as good as the data it ingests. So, if you are looking for data on marginalized populations, there have to be sources that will provide the data. So I wouldn’t call AI in healthcare “diverse” yet, but it will be over time as it reads and ingests more data and more data is collected.)
So if I were to guess what the relationship will look like in 25 years I’d say that a lot of the questions that used to be asked in an appointment will now be asked, answered, explained in English, and made actionable before you ever have a doctor's appointment. I imagine that my doctor's appointments evolve from “lectures” to discussions. I can ask questions or share inputs that my doctor might not know or be able to know. After all, a doctor is a human being and there is only so much they can know and learn.
Data is complicated, data from multiple sources is even more complicated.
An example. I am a huge believer in integrative oncology (IO). My doctor doesn’t understand it. In fairness, he never learned about IO in med school and probably doesn’t have a lot of patients who talk to him about it because the reception they get is “Well, I don’t know about it but if it works for you why not?”. For the record, IO is all about evidence-based treatments that don’t involve pharmacological interventions (think acupuncture, supplements, mindfulness, yoga, exercise, diet, etc).
So if I ask my doctor, he will say “it’s not in the data zone”, “I don’t know but if it doesn’t hurt why not try it”, etc. Noncommittal types of answers.
I asked ChatGPT to design an IO and Western treatment plan for me based on my diagnosis. I uploaded all of my details and asked for a plan. For fun, I asked it for the plan in Jerry Seinfelds voice. Here is what I got (this is a cut and paste, I didn’t edit it at all).
Category | Western Medical Treatments | Integrative Oncology Treatments |
Pancreatic NETs | "Alright, we’re sticking with Lanreotide shots. You know, every 28 days—it’s like paying rent for your pancreas." Targeted therapy? Everolimus or sunitinib—fancy names, but they just want to slow things down." | "You ever try a ketogenic diet? It’s basically no carbs. Feels like a punishment, but hey, it’s for your health. Oh, and throw in some turmeric and ashwagandha—I don’t know what they are either, but apparently, they help!" |
Liver Metastasis | "Your liver’s like, ‘What did I do to deserve this?!’ So, we’re gonna throw in hepatic artery embolization. Not a party move, but it might help." | We’re giving it some milk thistle. Yeah, milk and thistle—like a bad coffee flavor, but apparently, it’s good for you. Oh, and eat some plants. Gotta throw in some antioxidants—because everyone loves a good antioxidant!" |
Pain & Fatigue Management | "Here comes the meds! Pain? We’ve got opioids, NSAIDs—fancy talk for ‘your pain is annoying, let’s get rid of it.’ Fatigue? Pop a stimulant—it’s like your body's saying, ‘Let’s go!'" | "Acupuncture—yep, needles. But you’re not a pincushion; you’re healing! And if that’s not enough, there’s always lavender oil and Reiki—because nothing says pain relief like a little ‘woo-woo.’" |
Gut Health & Inflammation | "Gut health—who knew we’d care so much about our intestines?! We’ve got anti-inflammatories, steroids—you know, the usual suspects." | "Get on the probiotics train—because, apparently, your gut needs friends! And eat leafy greens—nobody loves them, but they’re good for you. Oh, and how about some intermittent fasting? No food for a while—makes you appreciate it more!" |
Emotional & Mental Health | "You’re gonna need a counselor—because, let’s be honest, cancer is stressful! It’s like talking to your therapist: ‘So, how’s life?’ ‘Well, I’ve got TWO cancers…'" | "You ever try meditation? It’s like taking a nap for your brain. And don’t forget support groups—misery loves company, right? Oh, and keep a journal—because everyone’s dying to know what you’re thinking!" |
Nutritional Support | "Meal replacement shakes—sounds gross, but sometimes you just don’t want to eat. Your dietitian’s got your back." | "Eat your cruciferous veggies—what are they? I don’t know, but you should eat them. Add a green smoothie too, because that’s what the healthy people do!" |
Spiritual Support | "Need some spiritual guidance? The hospital chaplain’s there for you!" | "Meditation, prayer, gratitude journaling—write it all down. And if that’s not enough, go to one of those healing retreats where they make everything seem possible!" |
Is it perfect? No. Is it way better than I can get from my oncologist or any other individual? Yes. You see, the data and training are from multiple places, and THAT’s the kind of thing that AI can help with and people (and doctors) struggle with.
So what happens in the future? Honestly, the sky is the limit (and actually the limit is way beyond the sky - the universe maybe?). Not only can AI pull together multiple sources and make sense of them, but it should be able to predict outcomes and recommend things based on Burt the individual, not just the majority of other cases who have tried a treatment. Treating Burt is the goal, not treating a generic person with my diagnosis. Individuals truly matter in healthcare and most treatments aren’t at the individual level yet.
Maybe the days of PFS (progression-free survival) and OS (overall survival) (two measures used to judge the effectiveness of cancer treatments based on trials and the mean/average) will go away, and data about Burt will help Burt feel his best and heal. Imagine, a medical system based on YOU, not the average of everyone. AI can and will help precision and personalized medicine (across all modalities) in so many ways, and that's the goal.
So am I excited? YES. If you are a clinician or a healthcare worker, do you need to understand AI and embrace its potential? 100%. If you don’t, you will be left behind, for sure. I can’t stress this enough. If you aren’t on the train, you will get left behind.
It’s so early, but never underestimate patients. We always want answers and want to help ourselves, so we will be (and some are) all over this. We always look for the latest and greatest answers that are relevant to us (and again, a big generalization but over time more patients will start adopting new tools and methods). We aren't going to wait for doctors or med schools to catch up. Some of us already are using AI for answers.
The doctor-patient relationship is going to change. A lot. Not to be controversial, but the role of the doctor will change, too. No longer will the doctor be the “god” but they will become another input. Bedside manner and data interpretation that’s specific to you will matter more from a doctor (although there are studies that show that AI can be more empathetic than a doctor. That's a topic for another day). A doctor will become someone on your team, but not the only part of your team that matters.
Let me know what you think. It’s a very exciting time but can also be overwhelming. Hopefully, med schools start teaching how to use and partner with AI. If doctors aren’t thinking about it, patients will pass them by. There, I said it!
I took the image at the beginning of this post from the following article: Allen, Matthew & Webb, Sophie & Mandvi, Ammar & Frieden, Marshall & Tai-Seale, Ming & Kallenberg, Gene. (2024). Navigating the doctor-patient-AI relationship - a mixed-methods study of physician attitudes toward artificial intelligence in primary care. BMC Primary Care. 25. 10.1186/s12875-024-02282-y. I am not sure what all that means but its the citation for the article!
Comentarios