Healthcare & Wellness

Verified against ChatGPT · 2026-08-12

Force loose visit notes into a proper SOAP structure without smuggling in a single unstated assessment

Restructures a clinician's free-text visit notes into Subjective, Objective, Assessment, and Plan sections strictly using only what was written — designed as an unsigned draft the clinician edits and finalizes themselves.

ChatGPT (GPT-5.1)2 fillable variables
Scope for this category: Organizational, educational and communication support only — no diagnosis, treatment recommendation, or medication guidance. Every prompt states plainly that a qualified clinician must review anything before it is acted on.

The prompt

Ready to copy — highlighted parts are example details you can swap.

Take the following free-text visit notes and restructure them into SOAP format. Nothing else — no added clinical content, no smoothing, no filling gaps.

FREE-TEXT NOTES
45yo here for annual physical, feels well overall, mentions occasional lower back stiffness in the mornings, no numbness or radiation. BP 128/82, HR 72, otherwise unremarkable exam. Discussed weight loss goals, wants to start walking 3x/week. Ordered routine labs, f/u in 3 months.

PATIENT CONTEXT (age, reason for visit only, no other clinical detail needed)
45-year-old, annual physical exam.

HOW TO SORT CONTENT INTO EACH SECTION
Subjective: anything reported in the patient's own words or paraphrased complaint, history, or symptom description — nothing the clinician observed or measured.
Objective: anything measured, observed, or tested — vitals, exam findings, results — nothing inferred from those findings.
Assessment: only if the notes explicitly state a diagnosis, impression, or differential — if the notes don't contain one, write "Not stated in source notes — clinician to complete" rather than inferring one from the subjective and objective content, even if an assessment seems obvious from context.
Plan: only actions explicitly stated in the notes — tests ordered, medications prescribed, follow-up scheduled, referrals made.

If a piece of the free-text notes doesn't clearly belong to one section (a comment that mixes observation and interpretation in the same sentence), split it at the natural boundary and place each half where it belongs, noting the split rather than guessing which single section it fits.

OUTPUT FORMAT
S:
O:
A:
P:
Below the four sections, a short list of anything from the source notes you couldn't confidently place, for the clinician to resolve.
Mark the entire output, top and bottom, as: "UNSIGNED DRAFT — for clinician review and completion, not a final note."

Customize

Optional — swap in your own details for the highlighted parts above.

Why this works

The rule that Assessment must default to "not stated — clinician to complete" rather than an inferred diagnosis is the single most important line in this prompt, because SOAP notes are exactly the kind of text a model has seen enormous volumes of during training, and it has a strong learned prior that a Subjective plus Objective section is almost always followed by a plausible Assessment — the model will readily generate one that sounds clinically reasonable given the symptoms described, and a fabricated assessment silently inserted into a note the clinician later signs without noticing is a direct path to a documentation error attributed to the wrong author. Splitting a sentence that mixes observation and interpretation at its natural boundary, rather than forcing the whole sentence into one section, respects the fact that clinicians genuinely do write notes this way in practice — a single sentence often contains both a measured finding and an interpretive comment — and silently choosing one section for the whole sentence would either lose the objective content or misrepresent an interpretation as a raw observation. Restricting patient context to age and visit reason only, rather than allowing broader clinical background into that field, keeps the reorganization task honestly scoped to what the free-text notes themselves contain, so nothing outside the clinician's own documentation can leak into the draft as if it had been part of the visit record. The unsigned-draft framing at both the top and bottom guards against exactly the failure mode of a note being copy-pasted into an EHR mid-review, before the clinician has completed the Assessment section the tool deliberately left blank.

What you get back

UNSIGNED DRAFT — for clinician review and completion, not a final note. S: 45-year-old presents for annual physical, feels well overall. Reports occasional lower back stiffness in the mornings, no numbness or radiation. O: BP 128/82, HR 72, exam otherwise unremarkable. A: Not stated in source notes — clinician to complete. P: Routine labs ordered, follow-up in 3 months. Couldn't confidently place: discussion of weight-loss goals and plan to walk 3x/week — placed under Subjective as reported intent, flag if it belongs in Plan instead. UNSIGNED DRAFT — for clinician review and completion, not a final note.

Verified against

ChatGPT GPT-5.1 · 2026-08-12

Changelog

  • 2026-08-12 Initial publish, verified against ChatGPT GPT-5.1.

Need this built into your business?

If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.

EXPLORE WHAT SCULT BUILDS
All Healthcare & Wellness prompts

Check your AI visibility

One URL in, a 0–100 score and the exact fixes out.

RUN THE CHECK

Browse all the tools

15 tools across six categories
13 of them never send your data anywhere

Free · No signup · No trial clock

SEE THE DIRECTORY