A denial report is one of the most revealing documents in a medical practice. Not because of any single denial — claims get denied for all kinds of reasons — but because of patterns. The same code appearing week after week, concentrated on the same dates, originating from the same providers or the same front-desk staff: that's not a billing error. That's an operational signal.
The data on where denials come from is unambiguous. 60–90% of denied claims result from front-end errors, not clinical ones. Rivet Health's analysis of denial drivers found that 49.7% of claim denials are caused by front-end problems alone — registration and eligibility failures accounting for 26.6%, authorization issues for 11.6%, and non-covered services for 10.6%. The most common single contributor: missing or inaccurate claim data drives 50% of all denials (Experian Health, 2025).
This means the majority of revenue lost to denials is a front-office process problem. The five codes below are the ones that, when they appear with any regularity, confirm that conclusion.
CO-16: The Intake Failure Code
CO-16 means the claim reached the payer but is missing required data fields, contains invalid information, or has errors that prevent adjudication. It is consistently among the top three denial codes by volume across all practice types and payer combinations.
A single CO-16 denial means a claim had an error. A recurring CO-16 pattern means the intake and charge-entry workflow is broken. The most common contributors: incomplete patient demographic capture at registration, missing or invalid NPI or taxonomy codes, incorrect place-of-service codes, and failure to capture required secondary diagnosis codes.
What makes CO-16 particularly important to watch: it is a precursor to CO-29. A CO-16 denial that isn't caught and corrected quickly enough becomes a timely filing denial once the submission window expires. The missing-information denial ages into an unrecoverable one. A practice that lets CO-16 denials sit in a queue is, on a delayed timer, creating permanent revenue loss from what started as a correctable data entry problem.
Operational signal: CO-16 volume concentrated on specific date ranges almost always traces to a staff change, a system update, or a period when intake protocols were bypassed. Pull the denial dates against your scheduling calendar. The correlation will be visible.
CO-29: The Most Expensive Code in the Report
CO-29 means the claim was submitted after the payer's filing deadline. Medicare allows 12 months from the date of service. Commercial payers range from 90 days to 18 months, and some behavioral health carve-outs run tighter than their parent payer contracts. Once that window closes, there is no appeal. CO-29 denials represent 100% unrecoverable revenue loss.
A practice with recurring CO-29 denials has no denial management follow-up process whatsoever. Claims are submitted, denied for any reason — eligibility failure, missing information, bundling edit — and then sit in a worklist untouched until the filing clock expires. The denial wasn't the problem. The absence of follow-up was.
The revenue math is merciless: a practice that loses 1% of $1 million in annual allowable charges to timely filing denials is forfeiting $10,000 per year that cannot be recovered under any circumstance. There is no appeal, no secondary billing option, no exception. It is simply gone.
CO-4: The Modifier Education Code
CO-4 means the CPT code submitted does not support the modifier attached to it — the modifier either doesn't apply to that code or conflicts with its definition under payer editing rules. Unlike CO-16 or CO-29, which surface broad workflow failures, CO-4 is a high-signal diagnostic code that points to a specific education gap.
CO-4 clusters appear in predictable circumstances: after a new provider joins the practice and imports modifier habits from a previous employer, after CPT annual updates change modifier applicability rules, or after a billing system change alters how modifiers are applied automatically. In each case, the fix is targeted — not a wholesale billing overhaul but a coder-provider feedback session focused on the specific modifier-code combinations generating denials.
In behavioral health practices specifically, CO-4 is frequently associated with telehealth billing (GT/95 modifiers), staff qualification modifiers (HO/HN), and place-of-service code conflicts. A CO-4 cluster in a behavioral health setting usually means telehealth billing protocols were never properly implemented when the practice added virtual services — a gap that often started accumulating months before the denial pattern becomes visible in reports.
CO-97: The Bundling Code That Can't Always Be Appealed
CO-97 means the service billed has been bundled into a previously paid service under CCI editing rules or payer-specific bundling logic. The payer considers it already reimbursed as part of an earlier adjudicated claim. Industry studies show 65–70% of CO-97 denials are preventable with proper CCI edit checking before submission.
What distinguishes CO-97 from most other denial codes is that it is often not appealable. If the bundling edit is correct — if your add-on code truly is included in the base procedure payment — the payer is right, and the only recourse is to understand the rules before billing. The correct response to CO-97 denials is upstream prevention through claim scrubbing, not downstream appeals.
The cases where CO-97 can be overturned involve modifier usage: modifier 59 (or the more specific X-modifiers: XE, XS, XP, XU) can unbundle procedures when the services are genuinely distinct and separately documented. But applying modifier 59 incorrectly to force payment is a compliance risk, not a billing strategy. The right answer is pre-submission CCI edit checking that flags these conflicts before the claim goes out.
CO-109: The Eligibility Verification Code
CO-109 means the claim was submitted to the wrong payer, the patient's insurance was inactive on the date of service, or the service falls outside the patient's benefit plan. It is one of the cleanest diagnostic codes in the entire denial taxonomy because it has essentially one cause: eligibility was not verified before the visit.
CO-109 denial clusters that appear on the same date range almost always trace to a specific eligibility verification failure event — a system outage, a front-desk staff change, a period when verification was performed by a new employee who hadn't yet learned the protocol, or a stretch where the practice was busy and verification was skipped to keep the schedule moving. The pattern is not random. It has a cause date, and that date is findable.
The standard that prevents CO-109 is verification of eligibility 72 hours before each scheduled appointment for established patients, not just at initial registration. Insurance changes at the start of the calendar year, when employers change plans, and whenever a patient changes jobs — all events that occur after the initial intake and that a one-time verification at registration will never catch.
The 78% rule: More than 78% of insured patients begin their provider search inside their payer's online directory (CredEx Healthcare, 2025). A provider whose enrollment lapses with a payer doesn't just lose those claims — they disappear from that payer's directory entirely, routing new referrals to competitors. CO-109 patterns can signal not just a verification failure but an enrollment lapse that's actively costing new patient volume.
Reading the Pattern, Not Just the Code
Individual denial codes tell you what happened on a specific claim. Patterns tell you what's wrong with the operation. The diagnostic framework:
- Same code, concentrated date range → workflow failure at a specific moment in time. Find the event on that date.
- CO-16 and CO-29 together → no denial follow-up process; missing-info denials are aging to timely filing failures.
- CO-4 from specific providers → modifier training gap; targeted education will resolve it.
- CO-97 at high volume → no pre-submission claim scrubber; CCI edits aren't being checked.
- CO-109 spikes → eligibility verification broke down on a specific date; find the cause.
MGMA's 2024 Stat Poll found that 60% of medical groups reported higher claim denial rates in 2024 compared to 2023. The problem is accelerating, not self-correcting, and the primary driver is not payer behavior — it is the operational discipline (or lack of it) on the provider side.
A practice that reviews its denial report not for the codes but for the patterns behind the codes turns a problem document into a management tool. The patterns are there. They just require someone to look for them rather than process each denial in isolation and move on.