AI protocol comparison: your protocols vs. the national model

✔ Reviewed by A. Keith Wesley, MD, FACEP, FAEMS, board-certified EMS physician and former State EMS Medical Director (WI, MN) · Updated 2026-10-16 · How we review

Every medical director eventually asks, "How far have our protocols drifted from the national model?" Answering that by hand takes weeks. With a grounded AI tool it takes an afternoon, followed by the part AI can't do: physician judgment. This page shows the workflow I recommend, with the prompts and a sample output format.

Reference standard: the NASEMSO National Model EMS Clinical Guidelines, published free by the National Association of State EMS Officials (current edition: Version 3, March 2022, PDF; a fourth edition is in development). This page reproduces no NASEMSO text and makes no clinical recommendations.

Step 1. Get the two documents

• Your current protocol (one guideline to start, e.g., cardiac arrest or stroke) as a text-searchable PDF.
• The NASEMSO guideline PDF linked above.
• Your state scope-of-practice document, if you want a scope check.

Use a tool that answers only from uploaded sources and cites them: NotebookLM, or a Projects/workspace feature in an enterprise Claude, ChatGPT or Grok plan. See our ranked research AI tools.

Step 2. Map the structure first

Before comparing content, have the AI map both documents. Structural gaps (no care goals, no pediatric column, broken cross-references, unclear medical control triggers) are the fastest wins and carry no clinical risk.

List the sections of [OUR PROTOCOL] and of the matching NASEMSO guideline.
For each, note whether it contains: patient care goals, inclusion/exclusion
criteria, care tiered by EMT/AEMT/paramedic, a pediatric section, medical
control triggers, cross-references, and QA/documentation measures.
Output a two-column table with page numbers for both documents.

Step 3. Run the comparison prompt

Compare [OUR PROTOCOL] with the NASEMSO National Model EMS Clinical
Guideline for the same condition (both uploaded). Make a table:
Topic | Our protocol (quote + page) | NASEMSO (quote + page) |
Type of difference (structural / sequence / scope / medication / other) |
Clinical significance for physician review (low / medium / high).
Rules: quote only from the uploaded documents. Do not recommend changes,
doses or drugs. If something is absent from one document, say "not addressed".
If you are not certain, say so.

What the output should look like

This is the table format only. The rows below are placeholders, not results from any real protocol.

Step 4. Verify and decide (the human part)

1. Open both PDFs and check every quote and page number. AI tools can misquote or blend sections.
2. Have a pharmacist or second clinician review every medication row.
3. Check scope against your state rules.
4. For each difference, decide: adopt, keep with a documented local rationale, or study further.
5. Log the decision, rationale, evidence and approver in a change log, and note that AI assisted.

Common things this process turns up are structural: documents on different versions, cross-references pointing to sections that were renumbered, the same policy existing in two versions, pediatric definitions that differ between documents, and equipment brand names that are no longer stocked.

Go further

Want a physician to do it? See protocol review consulting. Free version: the protocol comparison checklist.

FAQ

AI assists, the medical director decides. This page is educational and is not medical advice or a substitute for your local protocols and medical direction.