Medical Billing Denials: 10 Most Common Reasons and How to Prevent Them

The true cost of a denial isn't the face value of the claim. It's the face value plus the labor cost to work it. Industry estimates for the cost of working a single denied claim range from $25 to over $100, depending on whether it requires a simple resubmission, an appeal with documentation, or escalation to a supervisor or coder. Multiply that by the number of denials your practice generates each month, and the operational load becomes significant fast.

There's also a recovery problem. Not every worked denial comes back. Some expire before the billing team gets to them. Others require documentation that's no longer obtainable. Some payers have hard timely filing limits on appeals as well as initial submissions. The industry average for claims that are eventually written off after denial varies widely, but 15–20% is a commonly cited figure for practices without a structured denial management program. That's real revenue that simply disappears.

The more useful way to think about denial rate isn't as a percentage of charges — it's as a percentage of your clean claim rate. A clean claim is one that passes the payer's adjudication on the first submission without going to manual review. Industry benchmark for a well-run billing operation is a first-pass acceptance rate above 95%. If you're below that, the denial types below are where to look first.

Denial code CO-4 means the service is not payable without an appropriate modifier. This is one of the most fixable denials in billing because the fix is known, codifiable, and catchable before submission.

The most common CO-4 scenario: a procedure that requires a laterality modifier (RT for right, LT for left) is billed without one. Knee arthroscopy billed without specifying which knee. Cataract removal without indicating which eye. Payers — particularly Medicare — expect the modifier when a code can logically apply to more than one anatomical site. When it's missing, the claim goes to manual review or denies outright.

The second common CO-4 scenario involves NCCI modifiers. A soft-bundled code pair (modifier indicator 1) is billed with both codes but without the modifier needed to override the bundle. The payer sees the pair, sees no modifier, and applies the bundling rule. The result looks like a CO-97 denial but often accompanies a CO-4 when the payer is telling you both that the modifier was missing and that the service was bundled.

Prevention is straightforward: build modifier rules into your charge capture process. For any CPT code that requires laterality — most unilateral procedure codes — make laterality modifier a required field in your billing software before the charge can be finalized. For code pairs that frequently trigger soft bundles, create an internal reference that reminds charge entry staff which pairs require modifier review. A pre-submission check against an NCCI tool will catch the bundle flags; your charge capture process needs to catch the laterality gaps.

CO-97 is the NCCI denial code: "Payment is included in the allowance for another service or procedure that has already been adjudicated." This means one of your billed codes was considered bundled into another code on the same claim.

There are two ways to get a CO-97 denial. The first is a hard bundle — the code pair has modifier indicator 0 in the NCCI table, meaning the column 2 service is considered inherent to column 1 under all circumstances. No modifier can fix this. The correct action is to remove the column 2 code and not bill it at all, because its work is already priced into the column 1 payment.

The second way is a soft bundle billed without a modifier — modifier indicator 1 in the NCCI table. The services may have been genuinely distinct, but without a modifier communicating that to the payer, the claim looks identical to an improper double-billing situation. The payer bundles it. The fix here is to determine whether the services were actually distinct (different encounter, different anatomical site, different clinical purpose), pull the documentation that supports that, and resubmit with modifier 59 or the appropriate X-modifier.

The prevention strategy is the same regardless of which scenario applies: check code pairs before you bill. For a claim with multiple procedure codes, run every code combination through the NCCI table. It takes seconds. A claim with five procedure codes has ten possible pairs to check; a tool that checks all of them simultaneously makes this a thirty-second task rather than a five-minute manual lookup per pair. The cost of catching a CO-97 before submission is essentially zero. The cost of working it after is far higher.

CO-50 is a medical necessity denial: "These services are not covered when performed with the procedure reported with this claim." In plain terms, the payer's adjudication system didn't find a match between the ICD-10 diagnosis on the claim and the procedure code billed. Either the diagnosis doesn't support the procedure under the payer's LCD or NCD, or the diagnosis is too unspecific to establish clinical necessity.

The most common cause isn't fraud or even poor documentation — it's a mismatch between what the provider documented and what the coder pulled for the claim. A physician documents "knee pain, right knee" and the coder pulls M79.621 (pain in right knee), which is a valid code. But if the procedure billed is CPT 27447 (total knee arthroplasty), M79.621 alone will likely fail medical necessity review because primary osteoarthritis — M17.11 — is the expected supporting diagnosis for a total joint replacement. The documentation may fully support the procedure; the selected diagnosis code just didn't communicate it.

Prevention requires two things working together. First, ICD-10 specificity: code to the highest level of specificity the documentation supports. A generic symptom code when a more specific condition code is available is a common trigger. Second, know your LCD and NCD landscape for your specialty's high-volume procedures. For procedures that are frequently scrutinized for medical necessity — imaging studies, DME, certain surgical procedures — build a quick-reference list of the ICD-10 codes that appear on the applicable coverage policies. When a billed procedure hits that list, verify the diagnosis assignment before the claim goes out.

Prior authorization denials are unique in this list because by the time the billing team sees one, the service has already been performed. The patient is long gone, the provider's time has been spent, and now you're fighting with a payer over whether the service was authorized. In most cases, if prior auth was genuinely required and genuinely wasn't obtained, you have very limited recourse — and in some states, you're legally prohibited from balance-billing the patient for the resulting denial.

The cause is almost never that no one knew prior auth was required. It's usually a process breakdown: the scheduling team didn't check the payer's current auth requirements; the auth was obtained but for the wrong procedure code; the auth was in place but expired before the date of service; or the procedure changed between scheduling and the day of service and the auth wasn't updated.

Every one of those scenarios is preventable with a front-end process. The key elements of a functioning prior auth workflow are: a current, payer-specific reference for which procedures and services require auth (payers update their requirements frequently — checking once a year isn't enough); a verification step at the time of scheduling, not just at intake; and a confirmation that the auth on file matches the actual CPT code that will be billed. An auth obtained for CPT 27447 doesn't cover a revision done under CPT 27486, even if they're clinically related.

For practices with high prior auth volume, this process should be owned by a dedicated staff member with clear accountability for each auth from initial request through confirmation. The billing team shouldn't be learning about auth gaps at claim submission time; that's too late to do anything about it.

A duplicate claim denial occurs when the payer's system detects that a claim for the same service, same patient, same date of service, and same provider has already been adjudicated or is currently in process. This seems like a straightforward error to avoid — don't submit the same claim twice. In practice, duplicate denials often happen for less obvious reasons.

The most common scenario: a claim is submitted, the billing team doesn't see a response within the expected timeframe, and the claim gets resubmitted assuming the first submission was lost. But it wasn't lost — it was processing. The second submission hits as a duplicate. Now you have a denial to work and have to clarify to the payer that the second submission was an inadvertent duplicate of the first, which is still either processing or was already paid.

A related scenario: a corrected claim is submitted after a coding change, but the original claim hadn't been formally voided first. The payer sees two claims for the same date of service and denies the second as a duplicate, even though it was intended as a replacement. The correct process for submitting a corrected claim varies by payer and by claim type — generally it involves a claim frequency code indicating the correction (typically a "7" for Medicare replacement claims), not simply resubmitting the charge.

Prevention requires discipline in your follow-up workflow. Before resubmitting a claim that hasn't received a response, check the payer's online portal or call the EDI hotline to confirm the claim status. Most clearinghouses also provide acknowledgment reports (277CA transactions) that confirm whether a claim was accepted into the payer's system. Reviewing those reports before resubmitting eliminates most duplicate denials.

Timely filing denials are particularly frustrating because the service may have been completely appropriate, the coding may be correct, and the documentation may be perfect — but none of that matters once the filing window has closed. Timely filing denials are generally not appealable based on the merits of the claim. The only successful appeal path is demonstrating that the claim was submitted on time but the payer failed to process it correctly, which requires proof of timely submission.

Medicare requires initial claims to be submitted within one year (365 days) of the date of service. That's the most generous deadline in the industry. Commercial payers vary widely — 90 days is common, 180 days is generous, and some managed care contracts run as short as 30 or 60 days from date of service. If your practice sees a mix of payers, you could theoretically have claims that need to be submitted within 30 days of service sitting in a queue with claims that don't need to go out for a year.

The practical risk is in the work queue. A claim that's pending for whatever reason — missing documentation, a coding question, a patient eligibility issue — can age past its timely filing deadline while it sits unresolved. Billing teams that don't monitor claim aging by payer will miss this. The fix is operational: sort your work queue by payer-specific filing deadline rather than by submission date, and flag any claim that's approaching its deadline for priority resolution, even if the underlying issue hasn't been resolved. A clean-coded claim submitted on time beats a perfectly coded claim submitted one day late.

Patient demographic errors are boring to talk about but expensive in aggregate. A transposed digit in a date of birth, a misspelled name, an incorrect insurance ID, a missing group number — any of these can send a claim to manual review or cause an outright eligibility mismatch denial. These aren't coding problems; they're data entry and verification problems, which means they're fixable at a different point in the workflow than most of the other denial types on this list.

The most common demographic-related denial involves insurance eligibility: the patient's coverage wasn't active on the date of service, or the practice has the wrong payer on file. This often happens after a patient has changed jobs, changed plans during open enrollment, turned 65 and moved to Medicare, or had coverage terminated. If your intake process relies on an insurance card on file from a prior visit rather than verifying current eligibility before each encounter, you're billing blind on a meaningful percentage of claims.

Real-time eligibility verification — checking the patient's coverage through an EDI 270/271 transaction before the date of service — is the standard of care for preventing these denials. Most practice management systems support this natively or through a clearinghouse connection. For practices that see large volumes of established patients, an eligibility batch check the day before the schedule runs can surface coverage changes before the patient arrives, giving the front desk an opportunity to collect updated insurance information or discuss financial expectations with the patient.

Name and date of birth mismatches are usually caught by payers during the eligibility match and return as claim-level errors before adjudication. They're fixable quickly — correct the data and resubmit — but they add to the rework volume unnecessarily. A verification step during patient registration that cross-references the insurance ID against the insurance card takes sixty seconds and prevents the downstream work.

CMS Medically Unlikely Edits set maximum unit limits for every CPT and HCPCS code per date of service. When a claim exceeds those limits, it denies — and depending on the MAI adjudication indicator for that code, the denial may not be correctable even with documentation.

MUE violations tend to cluster in a few service categories: therapy codes, evaluation codes billed in time-based units, injection codes, and supply codes. Physical therapy practices billing high-unit claims for therapeutic exercise (97110) or manual therapy (97140) frequently encounter MUE limits. Infusion and injection practices billing multiple units of a drug administration code hit limits that don't match the clinical reality of how those services are performed.

The critical distinction is the MAI level. An MAI 1 limit is a claim-line edit — it applies per claim line, and in some circumstances, the same code can be billed on a second claim line to accommodate additional units. This is payer-specific and requires documentation support, but it's a path that sometimes exists for legitimate high-unit services. An MAI 2 or MAI 3 limit is an absolute date-of-service maximum. There is no workaround, no documentation override, no modifier that helps. The units billed genuinely cannot exceed that number on a single date of service for a single patient under Medicare guidelines.

Prevention requires knowing the MUE limit and MAI level for every high-volume code your practice bills before billing it. For practices with therapy or infusion services, this means building MUE limits into your charge capture logic so that claims exceeding the limit are flagged before submission rather than denied after.

When a patient has more than one insurance policy, claims must be submitted in the correct sequence. The primary payer adjudicates first; once their EOB is returned, the secondary payer can be billed with the primary's payment and adjustment information attached. Submitting to the wrong payer first, or submitting to the secondary payer without the primary's EOB, generates a COB denial.

The most common COB error is straightforward: the practice has the payer order wrong on file. A patient added a spouse's employer plan but the practice still has the patient's individual plan listed as primary. Or a Medicare-eligible patient has an employer group plan that should be primary under Medicare Secondary Payer rules, but the claim goes to Medicare first. Medicare has specific MSP rules that determine when Medicare is secondary — active employment coverage, workers' compensation, auto insurance, and several other scenarios all have their own sequencing rules under federal law.

A less common but more complex COB issue involves children covered under both parents' plans. The "birthday rule" — the parent whose birthday falls earlier in the calendar year has the primary plan for dependent children — is widely used but not universal. Some payers use different rules, and some states have laws that modify the birthday rule in certain circumstances. If your front desk team doesn't know the birthday rule or doesn't apply it consistently, you'll generate COB errors on a portion of your pediatric or family medicine claims.

Prevention comes down to eligibility verification that captures COB information, not just coverage status. When a patient has multiple insurances on file, verify the sequencing with the primary payer directly. Most payers can confirm COB order through the eligibility transaction. Getting the sequence right before submission prevents the follow-up work of correcting it after a denial.

A pre-submission checklist only works if it's embedded in the workflow rather than posted on a wall. The most common failure mode for checklists in billing is that they exist as a reference document but aren't actually integrated into the steps a billing team member takes before a claim goes out. When the workflow is busy, reference documents get skipped.

The most effective approach is to build the checks into your billing software or clearinghouse as automated flags rather than manual steps. Most modern billing platforms support claim scrubbing rules that can be configured to flag missing modifiers, exceed-MUE-limit claims, and common code pair issues before submission. For the items that can't be automated — prior auth status, COB verification, documentation completeness — a structured queue with a required checkbox before a claim can be moved to "ready to submit" status creates accountability without relying entirely on individual memory.

The checklist items that address the ten denial types above, in rough workflow order:

• Eligibility verified for date of service (prevents demographic and COB denials)
• Prior auth confirmed and matches scheduled CPT (prevents auth denials)
• All procedure codes checked against NCCI table (prevents CO-97)
• Units checked against MUE limits for high-volume codes (prevents MUE denials)
• Laterality and required modifiers present on all applicable codes (prevents CO-4)
• Primary ICD-10 code supports the procedure under LCD/NCD (prevents CO-50)
• No duplicate claims in queue for same patient/date (prevents duplicate denials)
• Claim date checked against payer timely filing deadline (prevents TF denials)

Not every item applies to every claim type or every practice. A primary care office has different risk exposures than a surgical practice. The point is to identify which denial types your practice actually sees most often — pull your denial report from the last 90 days and rank by volume and by denial code — and build your checklist around those specifically. Targeted is better than comprehensive. A checklist with eight items that address your top denial types will be used. A checklist with twenty items will be ignored.