Skip to content
Guides · Risk adjustment

Risk adjustment audit trail: making HCC coding defensible

The question is not only whether a code was captured. It is whether someone can later show why it belonged on that date of service.

A defensible code has a story

The story is simple: this patient had this condition on this date, the clinician assessed it or managed its implications, the documentation supports the code, and the submitted code matches the record.

Weak workflows skip parts of that story. They find a diagnosis in an old note, leave it on a list, or capture it without showing current relevance. That creates audit exposure and erodes clinician trust.

What the trail should capture

  • Candidate condition and the reason it was suggested.
  • Source evidence: prior note, claim, lab, imaging, medication, specialist report, or hospital record.
  • Current-year assessment, plan, monitoring, evaluation, or treatment.
  • Physician action: accepted, rejected, changed, resolved, or marked historical.
  • Coder or QA review, including corrections and rationale.

The trail should be easy to read

An audit trail buried in logs is technically available but operationally useless. The people who need it are physicians, coders, compliance, and practice leaders. They need to see the code, the evidence, the note text, and the decision path without reconstructing the whole chart.

Where Layrd fits

Layrd ties candidate HCCs to source documents and keeps physician review explicit. The output is not a mystery suggestion. It is a supported finding that can be accepted, edited, or dismissed.

Keep reading
Common questions
What is a risk adjustment audit trail?

It is the record of evidence showing why a risk-adjusting diagnosis was assessed, supported, coded, reviewed, and submitted for a specific encounter.

What should it include?

The source document, current-year clinician assessment, treatment or monitoring evidence, coder action, date of service, and any correction history.

Is a problem-list diagnosis enough?

Usually no. The problem list can prompt review, but the encounter note needs current support for the diagnosis.

How should AI be used?

AI should surface candidate conditions and evidence, then let qualified humans review, accept, reject, or correct with a visible trail.

See it on your own workflow.

Free for the first 4 weeks, and live in under a week.

HIPAA compliant, inside your EMR, at one flat monthly rate