Joshua Liu, MD

Patients Choose Risky Care When Health Systems Fail

I don’t think patients want ChatGPT or an AI doctor startup to replace their health system - but when you’re desperate, you’ll take what you can get - the safety and efficacy be damned. So if health systems don’t fill their own gaps, someone else will. It’s not like AI represents the first time we’ve seen patients willing to make trade offs of this sort. Medical tourism has been around for decades. No, it’s not safer for a patient to get surgery in a less developed country, with a less developed healthcare system and lower standards for healthcare quality/safety - not to mention, a lack of continuity of care from the surgical team abroad. But if the choice is between getting a surgery they can afford and no surgery at all, are we surprised patients made that trade off? Of course, those inside the healthcare system are often disgruntled with this: “we’ve seen patients get low quality surgery, develop complications, and our healthcare system at home has to treat the downstream problem created by medical tourism”. We’re going to see this play out with AI as well. Patients can’t wait months to see a doctor - so they’ll turn to ChatGPT to try and deal with it on their own - or as we’ve seen in some cases, they’ll turn to a purpose-built AI doctor for care. And yes, despite the few miracle stories that circulate on social media, there are also horror stories that will come out with patients experiencing real harm because of some of these AI tools - especially those that weren’t developed with the expertise of clinicians who deeply understand the complexity of medicine. But here’s the thing we’ve already seen with medical tourism: demand hasn’t slowed - in fact, it continues to grow 13% - 18% per year. If you think that’s a lot… well, ChatGPT and AI doctors are only going to increase because the demand for any healthcare is much higher than medical tourism specifically. So what does this mean for health systems? You have to lead - you can’t follow. You need to be in the trenches tinkering with AI, launching your own products, and leading from the front. Now it doesn’t mean being reckless like some of those companies who shall not be named. It means being nimble, fast, progressive AND safe/appropriate. It’s why at @SeamlessMD we’re thrilled our health system customers have been progressive with us on using RAG-based Conversational AI for patients using safe, health system-approved clinical content. Our customer partners are leading from the trenches AND doing it responsibly. We need health systems doing more AI, not less - because if health systems don’t fill the gap, then folks who don’t care as much about quality/safety will - and there needs to be a thoughtful, counterbalance to that.