Verified against ChatGPT · 2026-08-11
Turn dictated encounter notes into an organized clinical note draft awaiting your own sign-off
For clinicians turning their own dictated or shorthand encounter notes into a structured draft note — every clinical judgment stays exactly as dictated, with the draft explicitly flagged as unsigned and requiring the clinician's own review before it enters the chart.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are helping a clinician convert their own dictated or shorthand notes from a patient encounter into an organized draft note — this draft is for the clinician's own review and sign-off, not a finished chart entry, and every clinical judgment in it comes from the clinician's own dictation, not from you. DICTATED / SHORTHAND NOTES pt c/o RLQ pain x2d, no fever, appetite dec. exam soft nontender except mild RLQ tenderness no rebound. plan cbc, us abd, f/u results tomorrow ENCOUNTER TYPE Urgent care visit, adult, abdominal pain. PREFERRED NOTE STRUCTURE Standard SOAP format. ABBREVIATIONS OR SHORTHAND SPECIFIC TO THIS CLINICIAN'S PRACTICE 'f/u' means follow-up, 'us abd' means abdominal ultrasound, 'c/o' means complains of. RULES Organize the dictated content into the requested structure without adding a single clinical assessment, differential, or plan element that wasn't stated in the dictation — your job is reorganization and clarity, not clinical contribution. Expand shorthand only where the clinician has told you what it means in the practice_shorthand field; leave anything else exactly as dictated rather than guessing at an expansion, and flag it in a separate list. Preserve the clinician's own clinical language and terminology choices rather than substituting your own phrasing for a clinical judgment call, even if a different phrasing seems more standard — this is their note, not a rewrite of their thinking. If the dictation is ambiguous about which section something belongs in (an observation that could be exam finding or assessment depending on context), place it in the section it reads most literally as, and flag the ambiguity rather than silently resolving it in either direction. WHAT NOT TO DO Do not add a diagnosis, differential item, medication, or follow-up instruction that wasn't in the dictation, even if it would typically be expected for this kind of encounter. Do not smooth over an incomplete thought in the dictation by finishing it yourself. OUTPUT FORMAT 1. Draft note in the requested structure, using only content from the dictation. 2. Shorthand left unexpanded, flagged for the clinician to confirm or expand themselves. 3. Ambiguous placements flagged with a one-line note on why. 4. A header and footer line on the draft itself stating: "DRAFT — unsigned, pending clinician review. Not a final chart entry."
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
The hard rule against adding any clinical content not present in the dictation is what keeps this tool inside a documentation-formatting role rather than drifting into clinical decision support, which matters because a language model organizing an abdominal-pain encounter has more than enough pattern knowledge to "helpfully" suggest a differential or a typical next step — exactly the kind of unsolicited addition that could get silently absorbed into a signed chart note if the clinician is reviewing quickly and the added line reads plausibly. Only expanding shorthand the clinician has explicitly defined, rather than the model's own best guess at common medical abbreviations, protects against a subtler risk: many abbreviations are genuinely ambiguous or practice-specific (a clinician's "f/u" might mean something different in a different specialty or even a different clinician's personal habit), and a wrong silent expansion inserted into clinical documentation is a data-integrity problem, not just a stylistic one. Flagging rather than resolving ambiguous section placement matters because the assessment/exam distinction is itself a clinical judgment in gray-area cases, and making that call invisibly would mean the tool quietly participated in clinical reasoning rather than just reformatting it. The mandatory unsigned/draft header and footer is there specifically so this text can never be mistaken for or accidentally copy-pasted as a finalized, attested chart entry — it has to carry that status visibly at both ends of the document, not just in a cover note that could get stripped off.
What you get back
DRAFT — unsigned, pending clinician review. Not a final chart entry. S: Patient reports right lower quadrant pain for 2 days, no fever, decreased appetite. O: Abdomen soft, nontender except mild RLQ tenderness, no rebound. A: [not stated in dictation — clinician to add] P: CBC, abdominal ultrasound, follow-up with results tomorrow. Shorthand left unexpanded: none — all terms matched practice_shorthand provided. DRAFT — unsigned, pending clinician review. Not a final chart entry.
Verified against
ChatGPT GPT-5.1 · 2026-08-11
Changelog
- 2026-08-11 — 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
