Slow collections in an independent practice are usually treated as a billing problem. The vendor gets called. The workflow gets reviewed. The staff gets reorganized. The denial rate gets a quick check and is reported as a single number for the month, something between 8 and 12 percent. That headline number, by itself, is the wrong metric. A 10 percent denial rate that hides which payers are rejecting, which CPT codes are triggering rejections, and which workflow step produced each error is not a tracked metric at all. It is a summary statistic. Small practices that report only the headline number rarely know where to intervene.

The headline number hides four separate problems

Denial rate, as commonly reported, is a single percentage. The number combines every payer, every service line, every denial reason, and every responsible workflow step into one summary. In 2025, the industry average for first-pass denial rate hovered around 10 to 12 percent across small practices. A practice tracking only that summary number has no actionable information.

The composite hides four separate, often unrelated, problems:

• Payer-level denial concentration: One payer may be denying 22 percent of submitted claims while another denies 4 percent.
• Code-level concentration: Two or three CPT codes may account for a third of all denials.
• Reason-code concentration: Denials tagged CO-16 (missing or invalid information) and CO-50 (services not deemed medically necessary) require completely different remediation paths and completely different owners inside the practice.
• Workflow-step concentration: Denials originating from front-desk eligibility errors are a different problem than denials originating from coder error or biller submission error.

A composite denial rate moving from 10 percent to 11 percent month over month does not signal which of those four problems got worse. The number reports the symptom. It does not name the source.

The dollar impact of the missing detail is not small. A practice running $1.2 million in annual collections with a 10 percent first-pass denial rate has roughly $360,000 in claims initially rejected each year, assuming average claim values consistent with the volume. If 60 percent of those denials are eventually recovered on rework, the practice still leaves approximately $144,000 in unrecovered revenue plus the staff hours spent on rework. A practice that cannot identify which payer, which code, and which workflow step generates the largest share of those denials cannot prioritize the work that closes the gap.

What usable denial tracking actually looks like

A denial tracking system that produces operator-actionable information has four dimensions, not one. It segments every denial by payer, by CPT code, by denial reason code, and by the workflow step where the error originated. Each dimension is aggregated separately and trended over time.

The output is not a single percentage. It is a small set of pivot tables the operator can scan in five minutes: top three payers by denial volume, top three CPT codes by denial volume, top three reason codes by denial volume, and top three workflow steps by denial volume. Each table answers a different operational question.

A practice that sees BCBS at the top of the payer table and CO-16 at the top of the reason-code table can read the answer in one minute. BCBS is rejecting because of missing or invalid information. The front-desk eligibility and intake process probably needs review. A practice that sees Medicare at the top of the payer table and CO-29 (timely filing limit exceeded) at the top of the reason-code table reads a different answer. The lag between encounter and claim submission is the problem. The biller workflow or biller staffing needs attention.

For practices evaluating in-house billing performance, this segmentation is the difference between knowing a problem exists and knowing where it lives. For practices comparing billing services or RCM vendors, a structured vendor evaluation framework that includes denial reporting requirements should require every candidate vendor to deliver sample denial reports segmented along these four dimensions before the contract is signed.

Reason codes are the most diagnostic dimension

Of the four dimensions, denial reason code is the most operationally diagnostic. The reason code names the cause. The other dimensions name the location. Practices that learn to read reason-code distribution can usually find the largest single fix in the system within an hour.

The most common high-volume reason codes for small practices in 2025 include CO-16 (missing or incorrect information), CO-50 (not medically necessary as billed), CO-97 (the service is included in another procedure’s allowance), CO-29 (timely filing limit exceeded), CO-22 (coordination of benefits not on file), and CO-11 (diagnosis inconsistent with procedure). Each has a different remediation path and a different owner inside the practice.

CO-16 points to data capture at the front desk or to claim scrubber configuration. That is a workflow problem the biller usually cannot fix alone. CO-50 points to documentation quality or to coder diagnosis selection. That is a clinical and coding interface, not a billing system fix. CO-97 points to bundling errors at the coder level. The remediation lives in the coding review process, not the billing submission process. CO-29 points to the lag between encounter and submission, which is usually a biller capacity or assignment problem. CO-22 points to eligibility verification, which is again a front-desk and intake responsibility. CO-11 points to a mismatch between the diagnosis pointer and the CPT code, which typically lives in the encounter coding step.

The remediation owners are not the same person in most practices. The front-desk supervisor cannot fix coder bundling errors. The coder cannot fix front-desk eligibility errors. Assigning the same person to investigate every denial regardless of reason code is a common pattern in small practices and is also the pattern that produces the slowest improvement on denial rate.

A practice that ranks reason codes by volume each month, and notes which one moved most, has a focused remediation list with a named owner per category. A practice that tracks only the composite percentage has neither a list nor an owner.

Vendor denial reports often hide the segmentation

The same gap shows up when a practice asks its billing service or RCM vendor for a denial report. Many vendors deliver a one-page summary: composite denial rate, top reason codes by frequency, and a list of denials currently in rework. That report does not let the practice operate on the data. It does not segment by payer. It does not segment by workflow step. It identifies which codes appear most often without identifying which payer or which workflow is generating them.

The right question to ask a billing vendor before a contract is signed is not what is your denial rate. The right question is what does your monthly denial report look like, and can it be filtered along payer, CPT code, reason code, and workflow step. If the answer is the standard one-page summary, the vendor’s reporting will not produce operational decisions for the practice. At best it will produce a feeling that something is going on.

A vendor whose denial report is a single composite number and a top-codes list is a vendor whose product assumes the practice has no operational use for the data. That assumption is wrong, and it is the reason many small practices stay slow to identify the actual source of revenue loss month after month.

The rework list and the denial report are not the same thing. A list of claims currently being reworked tells the practice what the billing staff is working on this week. It does not tell the practice what pattern of errors is generating the rework queue in the first place. A vendor that delivers only the rework list, or that conflates the rework list with the denial report, has answered the question of what the staff is doing today without answering the question of how to stop generating the rework next month. Both reports matter. They are not interchangeable. The denial report names the source. The rework list names the response.

The operator question is segmentation, not percentage

Denial rate is a real metric. Headline denial rate is not the right metric. A practice that tracks denial rate as a single percentage has a number to report, not a number to operate on. The diagnostic information lives in the segmentation. Tracking denial volume by payer, by CPT code, by reason code, and by workflow step turns a summary statistic into a remediation list with named owners.

For practices evaluating internal billing performance or external billing vendors, the segmentation requirement is the difference between knowing there is a problem and knowing where the problem is. The fix is not better reporting language. It is operator-readable segmentation, every month, in five minutes.

The same segmentation produces a benchmarking surface over time. A practice that records the top three entries in each of the four pivot tables every month can read the trend in a single page. A payer that drops off the top-three list after a process change is evidence the change worked. A reason code that rises into the top three is an early signal that a workflow has slipped. None of that is visible in a single composite percentage.

GetPracticeHelp is an independent vendor evaluation and decision support resource for independent practice owners. The platform helps practice operators make informed operational decisions across EHR selection, revenue cycle and billing services, credentialing, compliance, vendor evaluation, and operational benchmarks for primary care, specialty medicine, dental, behavioral health, physical therapy, and chiropractic practices.

GetPracticeHelp publishes independently tested buyer’s guides, a comparison directory of verified service providers, and decision support tools that help practice owners evaluate build versus buy tradeoffs without vendor sales pressure. The platform does not accept paid placement. Affiliate revenue follows the ranking, not the other way around, and its methodology is fully disclosed.

Its writing covers vendor evaluation methodology, payer dynamics, regulatory and compliance shifts, AI-assisted operations for clinical workflows, and the structural challenges that limit how independent practices grow. Resources are available at GetPracticeHelp, with updates on LinkedIn.