Denial Management in Healthcare: A Complete Guide

Denial management is the operational discipline of (1) identifying every denial as soon as it occurs, (2) categorizing it by root cause, (3) working it within the payer's appeal window, (4) tracking outcomes, and (5) feeding root-cause data back into prevention.

Most practices conflate denial management with denial response — handling denials reactively as they show up. Real denial management has both reactive (appeal/resubmit) and proactive (prevention) workflows. The proactive workflow is the one that stops the same root cause producing the same denial next quarter.

The terminology, with the line drawn where the payer draws it. Different sources use 'denials,' 'rejections,' 'adjustments,' and 'underpayments' loosely. These are the definitions this site uses on every page, and they are chosen so that each one names a single event with a single denominator:

• Rejection — the claim was returned BEFORE adjudication, by a clearinghouse or a payer front-end edit, and never became an adjudicated claim. Medicare states the consequence plainly: a claim returned as unprocessable for incomplete or invalid information "does not meet the criteria to be considered as a claim, is not denied, and, as such, is not afforded appeal rights," and contractors are instructed not to deny such claims and afford appeal rights (Medicare Claims Processing Manual, Chapter 1, §80.3.1 and §80.3.2, read 17 September 2026). The denominator is claims submitted. The fix is correction and resubmission, never an appeal.
• Initial denial — the claim REACHED adjudication and the payer made a payment determination against it, with a CARC attached and appeal rights. On this site the denominator is claims that reached adjudication, not claims submitted, so the metric is not silently inflated by claims the payer never looked at.

Watch the denominator whenever you compare yourself to a published figure, because the published ones do not agree with each other. The HFMA Claim Integrity Task Force defines the initial denial rate as total initial denial claims divided by total claims submitted, and publishes a dollar version — initial denial gross charges over submitted gross charges — beside it; it counts only the first denial on a claim, excludes rebills, and publishes definitions rather than target values. AAFP computes its 5% to 10% range differently again: the total dollar amount of claims denied in a period divided by the total dollar amount of claims submitted in that period (read 17 September 2026). Three respectable denominators, three different numbers off the same book of business. Pick one, write it down, and never measure a rate one way and compare it to a benchmark computed another.

• Adjustment — the payer reduced reimbursement against what was billed (contractual, multiple-procedure reduction, modifier reduction). No denial occurred.
• Underpayment — the payer paid, but less than the contract requires. Finding these takes a contract-rate audit, not a denial worklist.

Two consequences follow, and both are places published benchmarks go wrong. 100% − first-pass rate is not your denial rate: the complement of a first-pass rate also contains rejections and claims that paid only after a corrected claim or an appeal. And a clean claim rate measures acceptance into adjudication on first submission — the event is acceptance, not payment — so a clean claim can still be denied on the merits.

Denial management practices generally cover all four categories above, but the workflows differ, and a worklist that mixes rejections into the denial queue will route data errors to the appeals team.

CARC (Claim Adjustment Reason Code) is the standardized code a payer uses to say why a claim line was adjusted, denied, or reduced. Combined with RARC (Remittance Advice Remark Code), the pair tells you how each line was processed and why.

Read the list, not a cheat sheet. X12 maintains the CARC set and revises it several times a year, and it licenses the official descriptions — so the column below is our paraphrase of what each code points at, not X12's own text. Check the current entry on the X12 Claim Adjustment Reason Code list before you build a routing rule on any of them (list read 17 September 2026). We publish no claim about what share of denials these twenty-five represent: no free source publishes a code-level distribution for physician practices, and the share this page used to print had nothing behind it.

The Owner column is the point. A denial list sorted by CARC is a reading exercise; a denial list sorted by the team that owns the root cause is a worklist.

A working denial management workflow has six steps. Each is essential — skipping any step degrades the whole process.

Step 1: Daily ERA Review. Denials are identified within 24 hours of payer adjudication, not when someone gets to it. ERAs (835 transactions) reviewed daily — every denial flagged for next-stage workflow.

Step 2: Categorization by Root Cause. Each denial categorized by operational root cause:

• Eligibility-related (CARC 27, 31, 140)
• Authorization-related (CARC 197, 198, 39 — 39 is an authorization refusal, not an eligibility problem)
• Coding-related (CARC 50, 96, 97, 167, 181)
• Documentation-related (CARC 16, 226, 252)
• Provider-eligibility related (CARC 170, 183)
• Timely-filing (CARC 29)
• Payer policy (CARC 24, 109, 119, 256)
• Contract issue (CARC 45)
• Other

Step 3: Routing by Category. Different root causes need different teams to fix and prevent. Eligibility denials route to front desk for prevention training. Coding denials route to coding team. Authorization denials route to scheduling/auth team. Routing makes prevention possible — without it, the same root causes keep producing the same denials.

Step 4: Action Within the Appeal Window — and know which window you are in. For Medicare fee-for-service the deadlines are federal and fixed: a redetermination must be filed within 120 calendar days of the date the party receives the initial determination, with receipt presumed five calendar days after the notice date (42 CFR 405.942); a reconsideration by a Qualified Independent Contractor must be filed within 180 calendar days of receiving the redetermination (42 CFR 405.962). Commercial appeal windows are set by your contract and by state law, and they are frequently shorter than the federal ones — pull them from each contract rather than assuming a common number. Set an internal SLA under whichever window is tightest, and work high-dollar denials and denials late in their window first.

Step 5: Outcome Tracking. What was the decision? How much was recovered? How long did it take? Tracked by payer, by reason code, by provider. Patterns reveal which payers are systematically wrong (worth appealing aggressively) and which root causes need front-end fixes.

Step 6: Prevention Loop. Root-cause data flows back to front-end processes weekly. If eligibility denials are rising, front desk gets retraining. If authorization denials are rising, the auth team's process gets reviewed. Without the prevention loop, denial 'management' is just rework — same denials forever.

Steps 5 and 6 are the ones that get dropped when the queue is deep, and they are the only two that change next quarter's number. Without outcome tracking you cannot tell which payers are worth appealing; without the prevention loop, denial 'management' is rework in perpetuity.

Working denials is reactive. Preventing them is the real opportunity — but the size of that opportunity is a number you have to measure in your own data, not one you can take from a page like this.

What the categories look like, without invented shares. Eligibility denials (CARC 27, 31, 140) are largely preventable with real-time eligibility verification before every visit, with the 271 response captured and acted on rather than glanced at. Authorization denials (CARC 197, 198, 39) are preventable with authorization tracking wired into scheduling and a hard stop before the service. Timely-filing denials (CARC 29) are preventable with charge-lag discipline and daily monitoring of unbilled and unsubmitted claims. Coding denials (CARC 50, 96, 97, 167) respond to modifier discipline, coverage-policy review and pre-submission scrubbing against payer-specific edits. What none of us can honestly tell you is what percentage of your denials each category represents: the only free cause-share breakdown we could source is the Optum denials index, which is built on hospital inpatient and outpatient remits and does not transfer to a physician practice.

The arithmetic, as an illustrative model rather than a benchmark. The inputs below are the two sourced figures available, and the multiplication is shown so you can redo it with your own numbers:

• A practice submits 2,000 claims a quarter.
• Apply MGMA's published 8% of claims denied on first submission for physician practices — a claim-count figure from a single-specialty aggregate, with no formula published behind it (MGMA). That is 160 denied claims.
• Apply MGMA's $25.20 cost to rework a denied claim (MGMA). MGMA attaches no population, denominator, sample or data year to that figure — the article's only methodology note belongs to a separate Change Healthcare denials index built on hospital inpatient and outpatient claims — so it is a rough unit cost, not a measured one for your practice. 160 × $25.20 = $4,032 of rework labour per quarter, before any claim is written off.

That is a model with two stated assumptions, not a projection, and it deliberately stops short of a revenue-recovery figure. Halving the denial count halves the rework line; what it does to collections depends on how many of those denials would eventually have been paid anyway, which your own overturn data answers and no published benchmark does.

Prevention still beats rework on first principles, because a denial that never happens costs nothing to work, ages nothing in A/R and consumes no appeal window. We do not publish a multiplier for how much better it is. Anyone who does is guessing.

Appeal success depends on three things: speed, documentation, and payer-specific format.

Speed. File as early in the window as you can. For Medicare fee-for-service the first-level window is 120 calendar days from receipt of the initial determination, and receipt is presumed five days after the notice date (42 CFR 405.942); the second level adds 180 calendar days from receipt of the redetermination (42 CFR 405.962). Commercial windows come from the contract. Waiting also costs you evidence: documentation memory fades and the people who can explain the encounter move on.

Documentation. A won appeals package includes:

• Original claim (with line item being appealed)
• Denial notice (EOB or denial letter)
• Operative note, progress note, or other clinical documentation supporting medical necessity
• Coding citation if relevant (CPT Assistant, AAPC guidance, NCCI edits)
• Cover letter explaining clearly: what was billed, why it's correct, why the denial is wrong, what action you're requesting
• Any prior authorization documentation if applicable
• For complex appeals: peer-reviewed literature supporting medical necessity

Payer-specific format. Each major payer has specific appeal forms and submission methods:

• Medicare fee-for-service: Redetermination Request (form CMS-20027) at the first level, filed with the MAC; Reconsideration (form CMS-20033) at the second level, filed with the Qualified Independent Contractor
• BCBS: Plan-specific appeal forms; varies by state
• UnitedHealthcare: Provider Appeal Form available via Provider Portal
• Aetna: Practitioner and Provider Complaint and Appeal Request form
• Cigna: Provider Reconsideration Form

Using the payer's own form and channel matters, because an appeal filed the wrong way can be dismissed without reaching the merits. We do not publish an uplift figure for it, and we do not publish appeal success rates by level: no free source measures them for physician practices. The nearest published figure is from a different population — Premier's February 2025 survey report states that approximately 70% of denials are overturned and paid — 68.6% in its own methodology section — and that the administrative cost of fighting a denial rose from $43.84 per claim in 2022 to $57.23 in 2023 (read 17 September 2026). That is a voluntary survey of 280 hospitals across 23 states, weighted by acute bed capacity and covering calendar-year 2023 claims, so it is hospital and post-acute data; read it as evidence that appealing is often worth it, not as a rate your practice should expect.

The number that actually governs your appeal strategy is your own. Track overturn rate and days-to-resolution by payer and by reason code, and you will find within a quarter which payers are systematically wrong and worth appealing hard — and which denials are correct and should be prevented instead.

Use this structure for first-level commercial appeals:

---

[Practice Letterhead]

[Date]

[Payer Appeals Department Address]

Re: Appeal of Claim Denial

• Patient name: [Patient]
• Member ID: [#]
• Date of service: [Date]
• Claim number: [#]
• Denial reason code: [CARC#] [description]

[Payer Name],

This letter requests reconsideration of the above-referenced claim denied on [date] for [denial reason]. We believe this denial is in error for the following reasons:

1. Medical necessity established. The patient presented with [chief complaint and clinical findings]. The performed service [CPT code] was medically necessary because [clinical justification with reference to documentation].

2. Documentation supports billed service. Attached operative note / progress note documents [specific elements that support the billed code]. The documentation meets [relevant CPT or LCD/NCD criteria].

3. Coding is correct per [reference]. Per [CPT Assistant, AAPC guidance, NCCI edit, etc.], the appropriate code for the described service is [CPT]. Modifier [X] was applied because [specific reason].

Action requested: Please reverse the denial and process the claim for payment per the contracted fee schedule.

Attachments:

• Original claim form
• EOB / denial notice
• Operative note / progress note
• [Other supporting documents]

If this appeal is not granted, we request the matter be escalated to second-level review per our provider agreement.

Sincerely,

[Signature] [Provider name, credentials] [Practice name and contact info]

---

Success factors: specific (not generic), referenced (cites guidance), professional (no emotional language), and complete (all attachments).

Real denial management requires:

1. ERA-based denial worklist that surfaces denials within 24 hours of adjudication. Most modern PM systems support this; verify yours does and that it's being used daily.

2. Denial reason coding that maps CARC/RARC codes to operational categories. Out-of-the-box CARC reporting is too granular for action — you need an operational categorization layer (eligibility, auth, coding, documentation, timely filing, payer policy, other).

3. Aging by appeal window — visibility into which denials have how many days left to appeal. Without this, denials age out of appeal eligibility silently.

4. Outcome tracking by payer and reason. This is what surfaces a pattern such as one payer denying modifier 25 far more often than its peers, or one reason code dominating a single provider's denials. Track overturn rate and days-to-resolution, not just counts.

5. Root-cause feedback to front-end processes. Weekly summary to scheduling, eligibility, auth teams about which root causes are driving denials. Without this, prevention doesn't happen.

6. Dollar-weighted denial reporting. The same denial rate on $200 claims and on $2,000 claims are different problems. Track denial dollars alongside denial counts — and keep the two denominators straight, since a dollar-based rate and a count-based rate are not comparable to each other or to a published figure computed the other way.

Most practice management systems have rudimentary denial worklists but lack the categorization, prevention loop, and dollar-weighting. Either invest in a denial management module or use an outsourced billing service whose denial workflow is built around these requirements.

This section used to carry a table of "healthy" denial rates by specialty. We removed it, because we could not source a single one of those numbers.

Here is the state of what is actually published. MGMA puts 8% of claims denied on first submission for physician practices — one figure, one claim-count denominator, a single-specialty aggregate, with no specialty breakdown behind it. AAFP publishes practice-management guidance that a 5% to 10% denial rate is the industry average and that below 5% is more desirable, with no population, sample or data year attached. HFMA's Claim Integrity Task Force publishes the definitions for initial denial rate and overturn rate, not target values. MGMA's fuller medians sit inside licensed DataDive products that are not free to quote. The 11% to 12% initial-denial figures that circulate come from vendor indices built on hospital or hospital-weighted remittance data — a different metric on a different population, and not a physician-practice rate. Nobody freely publishes a denial-rate median for cardiology, or for orthopaedics, or for behavioural health.

So a specialty table with a tidy range in every row is invented, wherever you find it, including in the version of this page that shipped before 17 September 2026.

What to do instead — build the only benchmark that binds you. Pull twelve months of your own remits and compute, separately per payer and per specialty line: claims submitted; claims rejected before adjudication; claims that reached adjudication; claims denied at initial adjudication; and denied claims overturned and paid. That gives you a rejection rate (rejections ÷ submitted), a denial rate (initial denials ÷ claims that reached adjudication) and an overturn rate, each with a denominator you can defend. Your baseline quarter is then the benchmark, and the question stops being "are we normal?" and becomes "are we better than we were, and which root cause moved?"

Payer mix, prior-authorisation intensity, modifier complexity and case mix genuinely do move denial rates between specialties. That is a reason your internal baseline is the right comparator — not a reason to trust someone's table.

A practical 90-day plan that produces a measured baseline, a prioritised prevention plan with a named owner per root cause, and a working denial worklist. It does not promise a denial rate, because the achievable rate depends on your payer mix and on which root causes you actually own:

Days 1–14: Baseline and categorization.

• Pull last 90 days of denial data
• Categorize by operational root cause
• Identify the top 5 root causes (typically eligibility, prior auth, modifier 25, timely filing, medical necessity)
• Document the baseline: rejection rate, denial rate (against claims that reached adjudication), denial dollars, overturn rate, and the category distribution — this baseline, not a published table, is what the next two months are measured against

Days 15–30: Quick-win front-end fixes.

• Implement real-time eligibility verification 24–48 hours before every visit
• Build hard stop in scheduling: no service without confirmed auth when required
• Charge entry SLA: 24 hours from date of service
• Train front desk on eligibility response interpretation

Days 31–60: Coding and modifier discipline.

• Audit modifier 25 use; train providers on documentation requirements
• Review LCD/NCD requirements for top 20 procedures
• Implement multi-stage scrubbing with payer-specific rules
• Begin daily ERA review with same-week denial action

Days 61–90: Prevention loop and KPI tracking.

• Weekly root-cause feedback to front-end teams, with a named owner per root cause
• Monthly denial trend review by payer and reason, against the day-14 baseline
• Update scrubbing rules based on emerging payer-policy patterns
• Document a playbook for each of the top five denial reasons

What you should have at day 90 is a measured before-and-after on your own baseline, a worklist that routes by owner, and a playbook per root cause. We publish no expected improvement figure, because we would be making it up: the achievable change depends on which root causes are yours to fix and which belong to payer policy. Re-run the loop annually and new denial patterns surface before they compound.

Most practices outsource billing fully or keep it fully in-house. A growing third option: outsource just denial management while keeping primary billing in-house.

When this makes sense:

• In-house team handles current claims well but never has time for denial work
• There is an aged-A/R backlog of unworked denials large enough to matter against your own collections — size it from your own aging report before anyone quotes you on it
• Specialty has high prior-auth complexity that one team can't keep up with — surgical lines such as orthopedic billing services, where prior auth and modifier rules stack up
• Provider audits or RAC reviews require denial-defense specialization

Pricing models vary by vendor and by scope: contingency on recovered dollars for a backlog clean-up, a per-worked-denial fee for ongoing work, or a percentage of denied dollars worked. We do not publish rate ranges for denial-only engagements, because no free source measures them — ask two or three vendors to price the same defined scope and compare those quotes against the internal labour cost the work displaces.

Risks of split workflows: Two teams working the same revenue cycle creates handoff issues. Worth considering only when there's clearly more denial recovery work than the in-house team can handle, AND the in-house team is doing primary billing well enough to keep.