Denial management is the work of finding out why a payer refused a claim, appealing or correcting it before the deadline, following it to payment, and fixing the cause so the same denial stops coming back. Done well, it recovers revenue on the claims worth fighting and shrinks the number of denials that arrive next month.
Denials are the most visible leak in the revenue cycle, and one of the most expensive to plug. Premier’s survey of its member hospitals and health systems put the average administrative cost of fighting a denied claim at $57.23 in 2023, up from $43.84 in 2022 (Premier, February 2025, vendor-published).
Payers deny at very different rates. Among insurers selling HealthCare.gov plans, in-network denial rates ranged from 3% to 36% in 2024, and fewer than 1% of denied in-network claims were internally appealed (KFF, March 2026). Patients rarely appeal on their own. If the provider doesn’t, nobody will.
Why claims get denied
Most denials fall into 4 groups. The group decides who fixes it.
| Denial group | Common reasons | Usually caused by | Who prevents it |
|---|---|---|---|
| Eligibility and registration | Coverage terminated, wrong member ID, wrong payer, coordination of benefits | Patient access | Registration and eligibility checks |
| Authorization | No prior authorization, authorization expired or didn’t match the service | Scheduling and patient access | Authorization team and ordering providers |
| Coding and billing | Invalid code combinations, missing modifiers, bundling edits, duplicate claims | Coding and charge entry | Coding, charge capture and revenue integrity |
| Clinical | Not medically necessary, wrong level of care, insufficient documentation | Documentation and case review | Physicians, clinical documentation integrity and case management |
Timely filing denials sit outside all 4. They mean the claim went out too late, and they’re almost never overturned.
The denial management process
- Pull every denial from the 835 remittance files and payer portals into one queue, with the reason codes intact.
- Group by payer and denial class, so each specialist works one type.
- Order the queue by dollar value, overturn likelihood and appeal deadline. Oldest-first queues let large claims expire while staff work small ones.
- Correct and resubmit what was a data error. Appeal what was a judgment call, with the documents the payer’s policy asks for.
- Track every appeal through the payer’s levels until it pays or the options run out.
- Tag each denial with a root cause and send it to the department that owns that step, then check next month that it stopped.
What a denied claim costs to work, and the claim value below which an appeal loses money, is on the denied claim cost page.
Denial prevention: stop the next one
Working denials recovers money. Preventing them is cheaper. The habits that make the biggest difference:
- Run real-time eligibility checks at scheduling and again before the visit.
- Keep a current list of which services each payer requires authorization for, and check it at scheduling.
- Fix the top 3 front-end denial reasons first. They’re usually eligibility, authorization and demographic errors.
- Load payer edits into the claim scrubber and update them when payers change policies.
- Report denials by root cause and department and by payer every month.
Medicare Advantage and Medicaid plans now have to give a specific reason for each prior authorization denial, and from Jan 1, 2027 they must run a prior authorization API that shows what documentation each service needs (CMS-0057-F). That’s a cleaner source of denial reasons than most teams have had. TSI’s summary of the 2027 rules covers the dates.
Where AI helps, and where it doesn’t
AI is useful in denial management where the work is reading, sorting and drafting:
- Reading denial letters and remittance codes, and sorting denials into classes.
- Ranking the queue by predicted overturn likelihood and value.
- Drafting appeal letters from the claim record, the denial reason and the payer’s policy.
- Spotting patterns: a payer that changed an edit, a department whose authorizations keep missing.
It’s weaker where the work is judgment: deciding whether a clinical record meets a payer’s medical-necessity criteria, talking to a payer’s medical director, or changing how a department works. Every appeal an AI tool drafts still needs a trained person to check it before it’s filed.
Ask any vendor that claims an AI lift for the baseline, the time period and the denial classes it measured. Results vary widely by payer mix.
How TSI works denials
TSI’s denial management service sorts denials by payer and class on arrival, ranks them by overturn likelihood, dollar value and deadline, and works the most winnable, highest-value claims first. PULSE, TSI’s denial and appeals platform, drafts appeals from the claim record and the payer’s rules; a specialist reviews and edits every appeal before it’s filed. TSI follows each claim to payment and reports the root causes behind repeat denials to your front-end teams.
On TSI’s own denial work, PULSE has measured an 86% denial overturn rate, 43% higher payment per appeal and 50% less time to appeal. Results vary by payer mix and client program, and outcomes depend on claim age and documentation quality.
Book a Denials Economics Sprint to see which of your denials are worth fighting, measured against your own baseline.
Frequently asked questions
What is denial management in healthcare?
The work of finding out why a payer denied a claim, correcting or appealing it before the deadline, following it to payment, and fixing the cause so the same denial doesn’t come back.
What does effective claims denial management include?
Sorting denials by reason and payer, ranking them by value, overturn likelihood and deadline, appealing with the right documents, tracking each appeal to payment, and sending every root cause back to the team that can prevent it.
What is the difference between a rejected claim and a denied claim?
A rejected claim never entered the payer’s system because of a formatting or data error, so you correct and resubmit it. A denied claim was processed and refused payment, so you appeal it or submit a corrected claim within the payer’s deadline.
What happens when a claim is denied as not medically necessary?
The payer decided the documentation doesn’t support the service under its coverage rules. The appeal needs the clinical record, the payer’s policy criteria and often a letter from the treating physician showing how the patient met them.
Can AI replace denial management staff?
No. AI is good at reading denial letters, sorting them, ranking them and drafting appeal letters from the record. People still need to check each appeal, handle clinical judgment calls, call payers and fix the process behind each denial.
What is the difference between reactive and proactive denial management?
Reactive denial management works denials after the payer sends them: appeal, correct and resubmit. Proactive denial management finds the causes, such as eligibility errors, missing authorizations and coding gaps, and fixes them before claims go out. Most programs need both: proactive work lowers the denial rate, and reactive work recovers what still gets denied.
How can denial management automation improve denial resolution rates?
Automation sorts denials by reason code and dollar value, routes each one to the right specialist, fills appeal letters with the records the payer needs and tracks deadlines, so fewer appeals expire. Staff spend their time on the denials worth winning. Track overturn rate and days to resolution to see whether it’s working.