Inside the WoundScribe Compliance Engine: How Pre-Sign Checks Catch Audit Risk
product
A close look at the compliance engine that runs before a wound chart is signed — what it checks, how amber vs red flags work, and why citations matter.
Most documentation tools tell you what you wrote. The compliance engine in WoundScribe tells you what's missing — before the chart locks. Here's how it works under the hood.
What the engine evaluates
The pre-sign check runs across four layers:
- Measurement completeness — length, width, depth, and tunneling/undermining when applicable
- Code-pair integrity — E/M plus procedure modifiers, ICD-10 etiology plus location, NCCI edits
- Payer-specific rules — Medicare LCD criteria, offloading documentation for diabetic ulcers, frequency limits
- Narrative sufficiency — medical-necessity language in the Assessment, plan tied to diagnoses
Amber versus red
Not every gap blocks a signature. The engine classifies findings:
- Amber — non-blocking. A missing measurement for a 99213, for example, raises an amber flag. You can sign, but the gap is logged in the audit trail with a rationale prompt.
- Red — blocking. A Medicare-required modifier 25 absent from a same-day E/M plus debridement is a red flag. Sign is disabled until resolved or until you accept a one-click correction.
Citations, not just verdicts
Every flag includes a citation — CMS NCCI Policy Manual section, LCD ID, or ICD-10-CM guideline reference. The point isn't to override clinical judgment; it's to show the rule so you can decide whether to fix, document a rationale, or override deliberately.
One-click fixes
When the engine can resolve a flag deterministically — appending modifier 25, adding a missing ICD-10 location code that's already implied by the wound location field — it offers a one-click fix with the citation attached. Manual overrides are tracked in the revision history.
Why it matters for audits
The audit bundle exported at sign includes the compliance results — every flag raised, every resolution, every rationale. If a reviewer questions the chart later, the bundle shows the engine ran, what it found, and how it was handled. That's the difference between a defensible chart and a defensive one.
See the broader wound documentation system for how this fits with imaging, scribing, and coding.