A primary-care appointment is short almost everywhere. The largest international review of consultation length found an average under ten minutes in 40 of the 67 countries it covered, and five minutes or less in 18 countries that between them hold about half the world’s population (Irving et al., BMJ Open 2017). Most people spend the first few minutes of that time re-remembering what they meant to say, and the last few realizing they forgot to ask the one question that actually mattered. The problem is rarely the clinician rushing you - it is walking in with a worry instead of a brief.
This is a 20-minute method for turning scattered notes, dates, and questions into a one-page appointment brief, using AI only to organize what you already know. Nothing here asks a model to guess what is wrong with you, how urgent it is, or what to do about it - those are exactly the calls your clinician is there to make, with information you did not have to fight to hand over.
If you are preparing for an appointment because something feels urgent right now, stop preparing and contact your clinician’s office, a pharmacist, or your local emergency number instead. This method is for a scheduled visit, not a substitute for judging whether something needs care sooner than that.
Step 1: Write down the facts, not the worry (5 minutes)
Before opening any AI tool, spend five minutes writing down what is actually true - dates, what happened, what changed, what you have already tried. Resist the urge to write “I think it’s X” or “it’s probably nothing.” Facts only, in whatever order they come to you. This raw list does not need to be organized yet; that is what the next step is for.
- Started noticing [thing] on [approximate date]
- It happens when [context], not when [other context]
- I tried [thing already done], and [what happened after]
- It has changed since I last saw a doctor about this: [how]
- I am currently taking: [list, or "see attached medication list"]
Step 2: Let AI organize, not interpret (5 minutes)
Paste your raw notes into a general-purpose assistant and ask it to structure them into a short timeline - explicitly telling it not to add interpretation. This is the one place AI genuinely saves time: turning a messy paragraph into an ordered, scannable list is exactly the kind of restructuring task a model does reliably.
Here are my rough notes about a health concern I want to discuss with
my doctor: [paste notes]. Organize this into a short, dated timeline
in plain language. Do not suggest a diagnosis, likely cause, or
urgency level - I only want my own information organized clearly, not
interpreted.
If the model adds a sentence like “this could suggest…” anyway, delete it before you save the output. Models are tuned to be broadly helpful, and “helpful” sometimes means volunteering an opinion nobody asked for. Your brief should contain only your own facts and questions - stripping out anything else takes ten seconds and matters more than any other step here.
Step 3: Rank what matters most (3 minutes)
Short appointments force priority. If you have three concerns and time for one, decide which one goes first - because the honest failure mode of most visits is spending all the available time on the easiest thing to describe, not the most important one.
I have three things to cover in a [N]-minute appointment: [list them].
Help me put them in the order I should raise them, assuming I might
only get through the first one or two - not based on medical
seriousness, which I'm not asking you to judge, but based on which
ones I said matter most to me.
Notice the instruction again: the ranking is about your own stated priorities, not a model-generated urgency score. That distinction matters enough to repeat at every step of this method.
Step 4: Bring a verified medication list, not a remembered one
If medications are part of the visit, do not rely on memory or a model’s guess at what you take. Use your actual prescription labels, pharmacy printout, or medication app export as the source, and ask AI only to format what you already have.
Here is my medication list exactly as printed on my labels/pharmacy
records: [paste, exactly as written]. Format this into a clean table:
medication name, dose as printed, and how I actually take it. Do not
change, add, or infer anything I did not give you.
If you manage several medications or multiple prescribers, the fuller method in reconciling a medication list before an appointment is worth the extra time before a visit that involves prescription changes.
Step 5: Write your questions as questions, not statements
A brief full of statements (“my sleep has been bad”) gives the clinician nothing to act on quickly. A brief full of questions (“what, if anything, should I be tracking about my sleep before the next visit?”) gives them something to answer directly.
Turn this list of concerns into direct questions I can ask my doctor:
[paste concerns]. Each question should be answerable in one or two
sentences during a short appointment - not open-ended essay
questions.
Assemble the one-page brief
Combine the outputs from steps 2 through 5 into a single page, in this order: today’s date and what you want out of the visit in one sentence, the ranked timeline, the medication table if relevant, and your questions. Leave white space to write down the answers - a brief you cannot annotate during the visit is only half as useful as one you can.
The full medical appointment brief template gives you this structure ready to fill in, plus a short section for note-taking during the appointment itself.
A rough symptom timeline and a full medication list are sensitive personal data. Use a chat tool with training turned off where that option exists, and see what ChatGPT remembers, sees, and shares before you paste anything you would not want retained in an account’s history indefinitely.
What this method does not do
It does not make your appointment longer, guarantee a diagnosis, or replace the clinician’s judgment about what matters most once they hear you out. It also will not tell you whether something is urgent - if you are unsure whether to wait for a scheduled visit at all, that question goes to a pharmacist, an after-hours line, or your local emergency service, not to this preparation method or to a chatbot. See AI is not a doctor for the fuller boundary on what a general-purpose assistant should never be asked to judge.
Common pitfalls
- Letting the model editorialize. Reread every organized output before printing it, and strip out any sentence that reads as a guess about cause or severity.
- Skipping the verified medication source. A model can format a list beautifully and still be wrong if you typed the dose from memory instead of the label.
- Ranking by what is easiest to say. The most important concern is often the one you feel most awkward bringing up - put it first anyway.
- Treating the brief as the whole appointment. It is a starting point for a conversation, not a script to read verbatim while the clinician waits.
- Bringing pages of AI-generated background reading. One page a clinician can scan in fifteen seconds is worth more than five pages they cannot.
Try it today
Before your next appointment, spend the twenty minutes: write the raw facts, organize them with the prompts above, rank them, verify your medication list against the actual labels, and turn your concerns into direct questions. Use the medical appointment brief template to hold it all on one page, and bring a pen.
