The Patient in Front of you has Already done their Reserach

There's a particular moment every GP/NP knows well. A patient sits down and before you've finished reading their notes, they've already told you what they think is wrong, what test they want and sometimes what supplement they should be taking for it. A few years ago that patient had been on Google the night before. Now they've been talking to Claude or ChatGPT, sometimes for weeks, working through symptoms, reading about hormone pathways and arriving with a level of detail that didn't exist in general practice consultations a decade ago.

This isn't a problem to be managed. It's the new shape of general practice and it's worth pausing on properly, because the modern patient simply has different needs than the patient of ten years or even six months ago. They've often already gathered information, formed early theories and sat with their questions for longer before ever walking into the room. Meeting that well takes real thought, and it means considering, deliberately, how you'll provide balance and education to a patient who has arrived this way, rather than relying on the same consultation style that worked for a patient who arrived with nothing but a list of symptoms.

Anthropic and OpenAI both launched dedicated consumer healthcare products within days of each other this year, Claude for Healthcare and ChatGPT Health, both built specifically to let people connect their medical records and wellness data and ask questions in plain language. These aren't fringe tools used by a handful of tech enthusiasts. ChatGPT alone is fielding roughly two hundred and thirty million health and wellness questions globally every week, and the people asking them are trying to make sense of lab results, understand what their three minutes with a doctor actually meant, and prepare better questions for next time. What that means in your consulting room is straightforward. The patient in front of you has likely already had a longer, more detailed conversation about their symptoms with an AI assistant than they're about to have with you. The opportunity in that isn't to compete with the research they've done, it's to be the person in the room who can actually make sense of it for them, with the training and judgement an AI assistant doesn't have.

This shift is showing up most clearly around functional and specialty testing. Patients are increasingly asking for MTHFR genetic testing, methylation panels, DUTCH hormone testing, gut microbiome panels and a wider suite of tests that sit outside standard general practice ordering. The functional medicine lab testing market itself reflects this, with industry analysis projecting it will grow from roughly eight billion dollars in 2026 to over fifteen billion by 2033. This isn't a fringe niche anymore, it's a structural shift in how a meaningful portion of patients want their health understood and it's worth taking seriously as exactly that. It also means the modern GP needs working fluency in tests that may sit outside their usual ordering pathway. The evidence base for several of these tests is genuinely mixed. MTHFR testing in particular has drawn sustained scrutiny from mainstream genetics and primary care research, with reviews concluding there are very limited clinical indications for testing the common MTHFR polymorphisms in asymptomatic patients. Knowing that, and being able to explain it clearly, is part of what it now means to practise well. The patient asking for it isn't being unreasonable, they've simply been guided toward something popular without always being able to tell the difference between content that's popular and evidence that's strong, and that distinction is exactly what your training is for and it's a genuine point of value you can offer that no AI assistant can replace.

The practitioners who will do well through this shift aren't the ones who hold the line against patient research and they're not the ones who simply approve every test a patient asks for either. They're the ones who pause and genuinely consider what this particular patient needs from the conversation. Often it isn't a yes or a no. It's to feel heard about the research they've done, to understand clearly why a test is or isn't appropriate for their specific picture and to leave with something more useful than they arrived with. That kind of balance starts with curiosity about the patient rather than a fixed script. Some patients want reassurance that they're not missing something serious, others want to understand the reasoning behind a clinical decision well enough to feel like a partner in it rather than someone being managed, and thinking through what this specific patient is actually looking for, before deciding how to respond to the test they've asked for, tends to produce a far better consultation than treating every research led patient the same way.

A few things are worth thinking through deliberately rather than leaving to instinct. How much time a consultation actually needs when a patient arrives already informed, rather than assuming the standard appointment length still fits. How you'll explain a clinical decision in a way that respects the effort a patient put into their own research, rather than simply stating a verdict. And as a practice, what your collective position is on which functional tests you'll order, which you'll refer out, and which you'll decline with a clear explanation, so the patient gets a consistent, thoughtful answer regardless of which clinician they see. There's a genuine opportunity in approaching it this way, because a patient who feels properly considered, rather than processed, tends to build a far stronger relationship with their practice than one who's simply told what to do. The practitioners who take the time to think about what this patient in front of them actually needs, rather than applying the same response to everyone who arrives with a printout or a screenshot from an AI chat, are the ones who'll earn that trust.

General practice doesn't get to choose whether patients use AI tools to research their health before walking through the door. That's already happened, at scale, and it isn't reversing. What's within your control is how thoughtfully you respond to it, by genuinely considering what this patient needs, what balance looks like in their specific case, and what kind of education actually helps them rather than simply manages them. The modern practitioner's job hasn't changed in its purpose, it's still to bring judgement, training and care to a patient who needs it. What's changed is the patient, and the practices who pause to properly think through what that patient now needs are the ones who'll navigate this well.