Why your doctor only answers 1 of your 4 questions

Today's Tip: Prepping for a doctor's visit

Most people walk into the doctor's office and forget half of what they wanted to ask. The fix takes 10 minutes the night before… and almost nobody does it.

Let me explain…

My neighbor Doreen is 64. She just retired from teaching second grade in Tacoma after 38 years. And for the past eight months, she'd been carrying around four separate worries she wanted to ask her doctor about.

A clicking noise in her right knee when she climbs stairs. A patch of skin on her forearm that's changed in the last few months and isn't going away. Trouble falling asleep that started after she retired. And a question about whether she should still be on her cholesterol pill now that she's not eating teacher-lounge donuts every morning.

Four things. Eight months of stewing. And every single appointment, she'd walk out having only mentioned one of them… usually the wrong one.

Sound familiar?

Here's what almost nobody is told… your doctor has roughly 15 minutes with you. They're trained to chase the most urgent thing you mention and let the rest slide. So if you lead with the knee, you don't get to the skin patch. If you lead with sleep, the cholesterol question evaporates. The order you bring things up decides what gets answered.

And this is where AI quietly becomes one of the most useful tools you can possibly have before a checkup. Not to diagnose anything. Not to scare you with worst-case scenarios. Just to help you walk in with a one-page, prioritized agenda that the doctor will actually thank you for.

Doreen tried this two weeks ago. She typed out everything in her head, dumped it into Claude, and asked it to help her organize her thoughts. Ten minutes later she had a printed sheet with her four concerns ranked by urgency, the right way to describe each symptom, and three follow-up questions for the cholesterol issue.

She walked out of her appointment with answers on all four. Her doctor told her it was the most efficient visit they'd had in years.

Here's the exact prompt she used. Copy it. Save it somewhere. Use it before every appointment you have from now on.

★ Copy-Paste Prompt

Works in Claude, Gemini & Grok

I have a doctor's appointment coming up and I want to walk in prepared. Help me build a one-page agenda I can print and bring with me.

Here are the things on my mind, in no particular order:

[List everything — aches, questions about medications, sleep issues, anything weird your body has done, questions you've been meaning to ask. Don't filter. Just dump it all out.]

I want you to:

1. Group these into categories (new symptoms, ongoing issues, medication questions, preventive screening questions).

2. Rank them from most urgent to least urgent based on what a doctor would want to hear first.

3. For each symptom, suggest the specific details a doctor needs (when it started, how often, what makes it better or worse, pain level 1-10).

4. Suggest 2-3 follow-up questions I should ask for each item so I leave the appointment with a clear plan.

5. Add a final section called "Things to mention if there's time at the end" for the lower-priority stuff.

Format the whole thing so it fits on one page and is easy to read out loud. Do not give me medical advice or guess at diagnoses. Just help me organize what I already want to say.

Why this works: The prompt is doing three quiet jobs at once. It tells the AI what role to play (organizer, not diagnostician). It gives it a clear output format (one page, easy to read). And it sets a firm guardrail against the one thing AI shouldn't be doing here — playing doctor. You get a tool that helps you think, without crossing a line it shouldn't cross.

When Doreen ran her four worries through that prompt, the order surprised her. She thought the cholesterol question was the big one. The AI ranked the skin patch first… because anything on the skin that's been changing deserves a quick look, while a medication question can be answered without urgency.

It also coached her on how to describe the sleep trouble. She'd been planning to say "I can't sleep." The agenda told her to say "I've had trouble falling asleep for about three months. It usually takes me 90 minutes to drop off. I wake up around 4 a.m. and can't get back to sleep." Same problem. Completely different conversation with the doctor.

★ Today's "Aha!" Moment

The job of AI at the doctor's office isn't to replace your doctor. It's to make you a better patient… so your doctor can be a better doctor for you.

Now, I know what you're thinking. "I don't want to type my private health stuff into some computer that's going to remember it forever." That's a fair worry. And here's the honest answer.

You can do this without using your real name, your doctor's name, or any identifying details. Just describe the symptoms. "A 64-year-old woman with a clicking knee" works just as well as anything more specific. And inside Claude's settings, you can turn off the option that lets your conversations be used for training. Go to Settings, then Privacy, then toggle it off. Done.

★ One Thing to Watch

Do not ask the AI what's wrong with you. Ever. If you type "I have a clicking knee and trouble sleeping, what do I have?" you'll get a tidy-looking list of possible conditions that will keep you up all night. AI is not a diagnostic tool. It's an organizational one. The instant you feel yourself reaching for "what's wrong with me," close the tab and call the doctor's office instead.

Here's what to do next. The night before your next appointment, sit down with a cup of tea and a notepad. Write down every single thing you've been wondering about your health for the past six months. The list will be longer than you think.

Then go to claude.ai, paste in the prompt above, and print the result. Bring it to the office. Hand a copy to the doctor at the start of the visit if you want — most will be delighted.

Ten minutes of prep. A completely different appointment.

★ ★ ★

Tool Spotlight ★

Claude

Site: claude.ai

Made by: Anthropic

Cost: Free tier available

Claude is the careful one. Of the three big AI tools, it's the most thoughtful with medical, legal, and financial topics — the most likely to say "I'm not the right tool for that" instead of making something up. That's exactly what you want standing next to you when you're preparing for a doctor's visit. It will organize and clarify… but it won't pretend to diagnose.

It's also the best writer of the bunch. If you want a one-page agenda that reads cleanly out loud, or a follow-up note to the doctor's office after the appointment, Claude produces tighter, more careful prose than the others. You can upload printed test results too — the free version handles up to 20 files per chat — and ask Claude to explain what each line means in plain English before your appointment.

Best for: writing letters and emails, summarizing long documents, organizing thoughts, explaining medical or legal language, reviewing contracts before you sign

★ ★ ★

★ Quick Wins

The "Read it back at sixth grade" trick — Any time you get a confusing document — a lab report, an HOA notice, a bank letter — paste it in and type: "Rewrite this at a sixth-grade reading level. Keep every fact. Cut every piece of jargon. Tell me what it actually means for me." Works in all three (Claude, Gemini, Grok). Claude gives the cleanest rewrite. You'll be shocked how much hidden meaning was buried in the original.

Argue against your own draft — Before sending a sensitive email — to a sibling, a contractor, your bank — paste your draft in and type: "Pretend you're the person receiving this. What would offend you, confuse you, or make you want to say no? Be honest." Works in Claude and Gemini. You'll find weaknesses you'd never have caught reading it yourself. Then revise.

The "is this still true today?" check — Anything time-sensitive — policies, deadlines, prices, government rules — should go through Grok, which pulls live information from the web. Type: "As of this week, what is the current rule on [X]? Cite your sources with links." Use it before you trust anything you read in an older article or hear from a friend. Especially useful for Medicare, IRS, and Social Security questions.

Skill Builder: The first answer is never the best one

Here's a mistake almost every beginner makes with AI. They type one question, get one answer, and accept it. Done. Move on.

That's like calling a friend who's an expert in something, asking one question, hearing the first sentence of their reply, and hanging up before they can finish. You'd never do that with a person. But people do it with AI constantly.

The first answer from any AI tool is almost always a starting point, not a finished product. It's a draft. A general response. A best guess based on what you said. The fantastic answers come from the second, third, and fourth message — when you push back, narrow down, and ask it to refine.

Watch what Doreen did with her agenda. The first version Claude gave her had her sleep issue ranked second. She typed back: "Actually, the sleep trouble has been getting worse the past two weeks. Does that change the ranking?" Claude moved it up. Then she said: "What questions should I ask if the doctor suggests a sleeping pill? I'd rather not take one." A new list appeared. By the fourth message, she had something nobody could have written for her in one shot.

Think of it as a conversation, not a search engine. Search engines give you one shot. AI gives you a back-and-forth. The trick is to keep going.

Here's a simple prompt to keep handy. Use it right after the AI gives you a first response — any time, on any topic — to squeeze a much better answer out of it.

★ The "Push back harder" prompt

Works in Claude, Gemini & Grok

That answer was a good starting point, but it's too general. Now do three things:

Point out the two or three places in your last response where you were vague or used hedging language like "it depends" or "consider consulting an expert."

For each of those vague spots, ask me the specific question you'd need answered to give me a sharper, more useful response.

After I answer your questions, give me the improved version.

Why this works: You're flipping the conversation. Instead of you struggling to know what to ask next, the AI tells you what it needs to do better. It turns a one-way answer into a two-way conversation… and the second answer is almost always two or three times more useful than the first.

Try this on your next AI conversation, whatever the topic. Recipe ideas. A draft letter. A summary of a long article you couldn't finish. Get the first answer… then push back. You'll be amazed at the difference.

The people who get the most out of AI aren't the ones who write the cleverest first prompts. They're the ones who don't stop at the first reply.

Why your doctor only answers 1 of your 4 questions

Read More from the AI Advantage