Skip to main content
The denial queue is where every adjudicated claim that wasn’t paid in full goes. Working it well means two things: recovering the money, and — the part most practices skip — feeding root causes back upstream so the same denial stops happening.

Prerequisites

  • Posting configured so denials route to the queue rather than to write-off
  • A fixed root-cause taxonomy
  • Appeal deadlines recorded per payer
  • Fee schedules loaded, so underpayments surface

The daily loop

1

Triage new denials by CARC family

Group them, because the fix is per-family rather than per-claim:
2

Prioritize by dollars, not by count

Ten 40denialsmatterlessthanone40 denials matter less than one 900 denial. Sort the queue by amount, descending, within age bands.
3

Decide: correct or appeal

4

Tag the root cause from a fixed taxonomy

Eligibility · Authorization · Coding · Documentation · Credentialing · Timely filing · COB · Demographic error · Payer error · Benefit design.Free-text root causes are useless in aggregate, which defeats the entire point of tagging. Use a closed list.
5

Enforce the aging rule

No denial sits untouched longer than 14 days. Appeal windows are short, and a denial nobody has looked at is a denial you are about to lose.Make the age of the oldest untouched denial a tracked metric.
6

Close the prevention loop weekly

Rank root causes by dollars. Change one upstream control per week. Verify next week that the cause dropped.Prefer structural fixes, a scrubber edit, over checklist items. See Build the billing rhythm.

What not to write off

A CO-denial cannot be billed to the patient. Group code CO means contractual obligation — the patient did nothing wrong. CO-197 (missing authorization) is the common example. Billing the patient breaches your participation agreement.And do not write off a denial before assessing appeal rights. A write-off is a decision to abandon the money, and it should be a deliberate one with a root-cause tag, not a default.

Underpayments belong here too

An underpayment doesn’t announce itself — the money arrives and the claim closes. With fee schedules loaded, your posting flags allowed amounts below the contracted rate. Underpayment is usually systematic, not random: a payer that loaded your contract wrong underpays every instance of the affected codes. Aggregate before disputing — one 12shortfallisntworthacall;twohundredinstancesofthesame12 shortfall isn't worth a call; two hundred instances of the same 12 shortfall is a configuration error worth a formal dispute, and the correction is retroactive. See Underpayments and contracts.

Metrics for the queue

That last one is the real signal. If the same root cause tops the list for three months, your prevention loop isn’t working.

Verify it worked

  • Denials route to the queue, not to auto-write-off
  • Triaged by CARC family, prioritized by dollars
  • Fixed root-cause taxonomy in use
  • Nothing untouched over 14 days
  • Corrected claims use frequency codes and reference the original
  • CO-denials never billed to patients
  • Underpayments flagged and aggregated
  • One upstream control changed per week, and verified

Common failure modes