There is a difference between working denials and managing them.

The RealityWorking a denial pattern rather than a single claim

⚠️

The Problem

Denials are worked one at a time, forever.

Every rejection is fixed individually and nothing upstream changes, so the same codes return month after month at roughly the same volume.

🔧

The Solution

Read a year of remittance data as one dataset.

Group by reason code, sort by dollars, and trace the top five back to the person, screen or field that created the condition.

✅

The Resolution

A short list of upstream fixes, in dollar order.

Fewer claims touched twice, a measurable drop in the codes you targeted, and a team working exceptions rather than volume.

Start with the electronic remittance advice you already receive

Everything this analysis needs is already arriving at your practice. The electronic remittance advice is the structured record of what a payer did with each claim and why. As CMS describes in its guidance on how Medicare structures remittance data, all electronic remittance advice from Medicare contractors uses the X12 835 version 5010 format adopted as the national HIPAA standard. That adoption is not a convention, it is a regulation: the rule that adopts the standard names the ASC X12 835 version 005010X221 as the required remittance advice transaction. Every payer sending you one is sending the same shape of data.

Inside it are three things worth telling apart, because the whole analysis depends on the distinction.

Group codes say who owes the money

A group code assigns financial responsibility. CO means contractual obligation, which lands on the provider. PR means patient responsibility. This is not a technicality: only when the group code is PR can a Medicare beneficiary be billed for the balance. Practices that treat every unpaid balance as patient-billable are creating a compliance problem alongside a revenue one.

Why the distinction changes what you do next

A CO-coded adjustment is a conversation with the payer or a defect in your own submission. A PR-coded balance is a patient statement. Teams that do not separate them either chase the payer for money the patient owes, or bill patients for amounts they cannot legally be billed. Both are expensive, and only one of them is merely expensive.

Claim adjustment reason codes are the sorting key

Claim adjustment reason codes carry the overall financial explanation for an adjustment. They are the primary sorting key for any denial analysis because they are standardized across payers. Two practices in different states with different payer mixes can compare notes using them, which is not true of anything in the free-text remarks.

That standardization is what makes the exercise tractable. You are not reading thousands of individual rejections. You are counting a few dozen codes and asking which of your own processes produces each one.

Remittance advice remark codes carry the detail

Remittance advice remark codes supplement the reason code with more specific explanation, and they are usually where the actionable detail sits. Not simply that a service was not covered, but which element of the submission failed. Both code sets are maintained and updated centrally on a published cycle rather than invented by each payer, which means a code that meant one thing two years ago may since have been retired or refined.

Denial root cause analysis, step by step

The technique is simple. The discipline of finishing it is the hard part.

1. Pull twelve months of remittance data. Not ninety days. Seasonal patterns, annual benefit resets and payer policy changes need a full year to become visible, and a single quarter will point you at the wrong culprit.
2. Group by reason code, then sort by dollars. Volume is a distraction. Rank by what the code is actually costing you.
3. Trace the top five back to the step that produced them. This is the part practices skip. Do not stop at the code; ask which person, screen or field created the condition. An eligibility denial is not a billing problem, it is a front-desk workflow problem that shows up at billing.
4. Separate defects from disagreements. Some denials are your own preventable error. Others are payer policy you can appeal but not prevent. Mixing the two produces a task list nobody can act on.
5. Fix upstream, then re-measure at ninety days. One change at a time, so the result can be attributed. If the code’s volume has not moved, the change addressed the wrong step.

In our engagements, the top five reason codes usually account for most of the rework volume, and two of them almost always trace back to something that happens before the patient is seen. That is the useful finding, because registration and eligibility are cheaper to fix than anything downstream of them.

Where repeat rejections actually come from

Four sources cover most of what we see.

Registration. Insurance captured at the front desk incorrectly or incompletely, then propagated to every claim for that encounter. The cost is not one denial. It is every claim attached to that patient until someone corrects the record.

Eligibility checked too early, or not at all. A verification run at booking is stale by the time of a visit three weeks later. Coverage lapses, plans change at year end, and the check that passed in November fails in January.

Templates that do not prompt for what the code requires. If a payer requires specific documentation to support a service and your note template does not ask for it, the omission is structural. Every clinician using that template produces the same gap, which is why the denial recurs with such regularity. This is where the workflow actually breaks more often than practices expect.

Knowledge sitting with one person. One biller knows the payer’s quirk and handles those claims correctly. That person goes on leave and the denial rate moves. Undocumented knowledge looks like competence right up until it looks like a staffing risk.

Claim rework is the number that funds the project

Claim rework is the hidden cost in all of this. A denied claim does not simply get paid later. It gets touched again by someone paid to do something else, it ages, and a share of reworked claims are never collected at all. The labor cost is real money leaving the practice, and unlike a write-off it appears nowhere on a report as a loss.

We do not publish an industry figure for this, because the honest answer is that it varies enormously by specialty and payer mix, and a number borrowed from someone else’s practice will not describe yours. What we do instead is measure it in your own data: how many claims are touched more than once, by whom, and for how long. That number is specific to you, and it is usually the one that gets the work approved.

Appeal what is worth appealing

Not every denial should be appealed, and not every appealable denial is worth the labor. The Medicare appeals process runs through defined levels with defined timeframes, set out in the Medicare Claims Processing Manual. Knowing those timeframes matters, because an appeal filed late is not an appeal.

The rule we use: appeal where the denial is a disagreement about policy or medical necessity and the dollars justify the effort. Do not appeal where the denial is your own preventable defect. Fix that instead. A practice that appeals its own errors is paying twice for the same mistake and learning nothing either time.

Making the fix hold

Most of these projects fail within two quarters, not because the analysis was wrong but because the fix lived in one person’s head or in a process nobody owned.

Three things make it hold. A named owner for each of the top codes, reviewed monthly rather than annually. Changes built into the system as a required field, a scheduling stop or a template prompt, rather than into a reminder email. And targeted training for the specific step that failed, not a general refresher for everyone, which is how most practices respond and why most responses do not work.

There is a reporting dimension too. The same data quality that drives clean claims drives clean quality reporting, so practices that get this right usually find their quality reporting work gets easier at the same time. Both depend on structured data captured correctly the first time, by the person closest to the source. And because remittance files contain protected health information, how they are stored and shared belongs inside your HIPAA compliance practice rather than being treated as a billing convenience.

If the denials are coming from how the system is configured rather than how staff are working it, that is a different job, and it sits within our EHR consulting practice rather than with your billing team.

Denial management: common questions