Note honesty guardrails

Medipyxis progress notes must document what the clinician actually observed and did — nothing more. Two guardrails landed the week of August 25, 2026 that remove long-standing places where the renderer was quietly filling clinical fields the clinician had not documented. If a Medicare reviewer reads one of your notes, everything on the page is now traceable back to something you entered.

Before you start


Guardrail 1 — No synthesized post-debridement measurement

What used to happen. On any visit with debridement, the renderer set the post-debridement wound area equal to the pre-debridement area and printed both, labeled as if they were separate measurements taken at separate times:

Wound area — pre  (auto: baseline)    12 cm²
Wound area — post (auto: this visit)  12 cm²

The wizard has never captured a post-debridement area anywhere in the flow, so the "post" row was a measurement that never happened, appearing on every debridement, on every note. Identical pre/post pairs on repeat visits are exactly the pattern a reviewer reads as fabricated measurement data — and identical by construction meant no clinician error could ever produce anything different.

What the note does now. The renderer emits the pre-debridement area only when it was documented, and prints "Not measured" for the post-debridement row rather than echoing a number that was never taken. The misleading (auto: baseline) / (auto: this visit) qualifiers are gone. The "post > pre" clinical safety check (which could never fire while the two values were forced equal) can now fire for real once a real post-debridement measurement is ever captured.

What this means for your workflow. No change to how you document. If you did not take a post-debridement measurement, the note will not claim you did.

Guardrail 2 — No invented comorbidity type or stage

What used to happen. Where a clinician had documented that a patient carried a condition but had not specified its subtype, the note renderer filled the blank with a concrete clinical value:

Each of those subtypes is ICD-10-determinative: it picks E11 vs E10, N18.1 through N18.6, and E43 vs E44 respectively. A coder reading the note had no way to tell whether the subtype came from the record or from a default.

What the note does now. When the qualifier is not documented, the renderer emits the condition alone — "diabetes mellitus", "Chronic kidney disease", "Malnutrition" — with no invented type or stage. The signed note now says only what you actually said.

What this means for your workflow. If the subtype matters for coding, document it. When you leave it blank, the note will not silently pick one. The diagnosis-code validation drawer flags the corresponding ICD-10 gap so the coder is not left guessing.

A signed note that carries an undocumented "Type 2" or "Stage 3" gives a payer a reason to deny — and, on audit, gives the reviewer a fabrication finding. If a subtype is clinically true, put it in the record; if it is unknown, leave it blank and the note will report it as unspecified.


What "note honesty" means in Medipyxis

Both guardrails share one rule: the printed note may never state a clinical fact that was not entered by the clinician. That includes:


What still needs your attention

Guardrails remove silent fabrication. They do not replace clinician review:

Auto-rendered from related: in frontmatter.

ESC