In short: AI assistants are good at explaining what a test measures and what to ask your doctor; they are not a diagnosis. Most bad answers come from bad questions: a photo without units, no reference range, no dates, and your name, date of birth and record number in the picture. This builder turns what you paste into a clear question that asks for plain explanations, comparisons with the range on your report, questions for your doctor and sources, and it removes personal details before you copy. For questions about your whole history, Anpheros Health connects your record to Claude read-only, and ChatGPT can send notes into it for you to accept.
Your identity stays out
E-mails, phone numbers, long ID numbers and dates of birth are removed before you copy.
A better question
Units, ranges, dates, context and what you want to understand, in one structured message.
Questions for your doctor
The question asks the assistant for explanations and doctor questions, not a diagnosis or doses.
How to use it
- Choose what the question is about: lab results, a report, symptoms you tracked, or a medicine.
- Paste the values or text exactly as printed, with units and the reference range. Leave out your name and record numbers.
- Add context if it helps: long-term conditions, medicines, what changed. Optionally your age range and sex, because some ranges depend on them.
- Write what you want to understand, or keep the default.
- Copy the question and paste it into ChatGPT, Claude, Gemini or the assistant you use.
What the question asks for
It asks the assistant to explain each value or term in plain language, to compare values only with the range printed on your report, to describe how values changed over several dates, to list three to five questions for your doctor, to say clearly when something needs urgent care, and to cite reliable sources such as MedlinePlus or the NHS. It also asks the assistant not to give a diagnosis and not to suggest or change doses. These are the same limits a careful clinician would put on an answer given without examining you.
What to keep out of any AI chat
Your full name, date of birth, address, phone, national or insurance number and hospital record number are not needed to explain a test. A photo of a report usually contains all of them. Typing the values, or using this builder, gives the assistant what it needs and nothing more. Read ChatGPT, Claude and your health data for the settings to check before you share anything.
Frequently asked questions
Is it safe to paste my lab results into ChatGPT?
What you paste is processed by the company that runs the assistant, under its terms, and may be used to improve its models unless you switch that off in the settings. Values without your identity carry far less risk than a photo of the whole report. This builder removes the identifiers it can recognise; check the text once more before you paste it.
Does this tool send my data to ChatGPT?
No. It builds the text in your browser. You copy it and paste it yourself. The buttons only open the assistant's website; your data is not put in the link.
Can the AI diagnose me from my results?
No assistant can diagnose you from a list of values: diagnosis needs an examination, your history and often other tests. The question asks for explanations and for questions to take to your doctor.
Which assistant should I use?
The question works in ChatGPT, Claude, Gemini, Copilot and others. Prefer one with web search or sources, and check the sources it cites.
Can an AI read my whole medical history instead?
Yes, through a connector you control. Anpheros Health connects your record to Claude read-only, section by section, and you can switch it off at any time. ChatGPT can send notes, symptoms and measurements to your Anpheros record, which you accept or reject, but cannot read it.
