When should you use it?
- You have years of lab results as PDFs
- Your portal's charts are hard to read
- You want to discuss trends with your doctor
Not for: Diagnosis or treatment decisions; Uploading records to tools your provider has not cleared, if that worries you
What do you need, and what do you get?
You need
- Lab result PDFs
- Basic context such as age
You get
- Table of markers by date
- Interactive charts per marker
How does it work, step by step?
- Download resultsExport the PDFs from your patient portal.
- Ask for extraction firstHave it pull every marker, value, unit and date into a table.
- Check a few valuesCompare several rows against the PDFs.
- Build the viewerAsk for an interactive chart per marker with the reference range.
- Bring it to your doctorUse the charts as a conversation aid, not a diagnosis.
The prompt
The original prompt by Matt Stockton, word for word, with a link to where it was published. Below it is our adapted version with fill-in blanks, if you want it. Works in Claude and ChatGPT. Values in lime are placeholders.
Original prompt by Matt Stockton
I want you to build me an interactive tool for exploring my medical test result history. Here are some details on what I am looking for. I will attach several PDFs that contain various test results and diagnostic levels I want to be able to explore this data interactively in the browser I want the tool to highlight any abnormalities I want the tool to highlight any recommendations that I should take action on I want the tool to be creative in nature. The way I explore this data should be differentiated and unique compared to how one might explore this data normally. I want the tool to take into account any additional public data or relevant information for a 41 year old male. Feel free to ask me any questions you have first.
Exactly as published by Matt Stockton in Exploring My Unstructured Medical Data from PDFs with Claude.
Our adapted version (optional)
Here are my lab result PDFs. First extract every test marker into a table with date, marker, value, unit and reference range, converting units so each marker is consistent, and list anything you could not read. After I confirm the table, build a simple interactive page with one chart per marker over time showing the reference range. Do not interpret the results medically.
What can go wrong?
- Different labs use different units; make sure they are normalized.
- Interpretation belongs with a clinician; the author stresses human oversight.
Our take
A clear personal example of structuring messy documents. Ask for the table first and check it, since a pretty chart of a misread value is worse than none. Matt's starting point is common: a patient portal with basic trend charts and years of results trapped in PDFs. Asking for a table of every marker, value, unit and date before building any chart gives you a checkpoint, so compare several rows against the reports. Different labs use different units, so make sure values are normalized before they are plotted side by side, and show the reference range on each chart. The viewer is most useful as something to bring to an appointment; interpretation belongs with a clinician, and the author stresses that human oversight. Keep the files on your own machine where possible, and think about which tools your provider would be comfortable with before uploading records. This is for understanding trends, never for diagnosis or treatment decisions.
Where the time goes back: Years of scattered results become one view you can actually read.
FAQ
Did formats matter?
Is this medical advice?
Source and credit
The original prompt on this page is by Matt Stockton, shown word for word with a link to where it was published. The summary, notes and adapted version are ours.
- Original
- Exploring My Unstructured Medical Data from PDFs with Claude
- Creator
- Matt Stockton
- Published on
- Matt Stockton
- Original date
- Jul 18, 2024
- License
- Not openly licensed; the original prompt is shown word for word with credit and a link to the original
- Originally for
- Claude
- AI dry run
- Oct 23, 2026: passed. Dry-run tested by our AI test agent on sample input. PASS: Gates held on desk check; no live sends or purchases.
AI dry run on sample data; not run with live connected accounts. The playbooks.bot test agent filled the prompt with synthetic inputs, produced the output and checked it against the page's stated outputs and guardrails. - Verified
- Oct 24, 2026: source link resolves to the original, the credited creator published it, our excerpt or prompt matches it, and the license allows our use.
Are you the creator? Claim or correct this page.