Each diagnosis you submitted raised a member's risk score, and with it, what you were paid. In a RADV audit, CMS selects a sample of members and asks for the medical records behind each of those diagnoses.
Where the chart supports the diagnosis, the payment stands. Where it does not, the money goes back. The whole question comes down to one thing: can you show what the provider documented, and when?
By the time the letter comes, the encounter is years old. The coder who submitted the diagnosis has moved on. The provider's office takes weeks to send records. Your team pulls charts against the clock and hopes the right note is in there.
Most plans meet RADV the same way: reactively, under deadline, one chart at a time.
AAVAE scores the documentation behind every risk-adjusting diagnosis continuously, in shadow, inside your own data platform. You see your exposure member by member while the chart is still easy to find and the provider still remembers the visit.
Which diagnoses the record supports, which it does not, and what the difference is worth. Updated as new claims arrive.
Evidence for each diagnosis is sealed when it is reviewed, not reconstructed years later. Your reviewers decide; nothing changes automatically.
Sampled members are matched to packages sealed earlier. The response starts finished, not from scratch.
This is what bi-directional payment integrity means in risk adjustment. Every diagnosis gets an answer, and every answer gets a sealed file written for the person who will read it.
Every rule in our platform is a parameter, not code. So when the HCC model changes, your scoring changes with it, and the evidence you sealed under the old model stays exactly as it was.
A recent RADV sample or a slice of your risk-adjusted population. Within two business days, a written view of what is supported, what is not, and what it is worth.