Why Claims Get Denied: The Patterns Behind Most Rejections

Ask a practice administrator why claims get denied and you will usually hear a shrug and the word "payers." Ask the same question of a worked denial log and you get something much less mysterious: a short list of repeat offenders, most of which were created inside the practice before the claim ever left the building. Denials feel random because they arrive scattered across weeks and payers. Grouped by cause, they stop looking like weather and start looking like a process problem you can name.

That reframing matters because it changes what you do next. If denials are weather, the only response is to work them harder. If denials are output, the response is to change the input. The practices that get their cash flow under control almost always make that shift, and they make it by insisting on one discipline: every denial gets a cause, and causes get counted.

The five families of denial

Reason codes are numerous; underlying causes are not. Nearly every denial we work belongs to one of five families, and each family has a different owner inside the practice.

1. Coverage and identity problems

The patient was not active with that plan on that date, the plan is secondary rather than primary, the member ID has a digit wrong, or the patient's name in your system does not match the payer's record. These are the cheapest problems in medicine to prevent and among the most expensive to discover after adjudication, because by then the visit has happened, the patient has left, and someone has to chase both the payer and the patient to reconstruct the truth. Coverage denials are a front-desk output, which is why we treat eligibility verification as a revenue function rather than a clerical one.

2. Authorization and referral problems

The service required prior authorization and did not have one; the authorization existed but covered a different date range, a different site of service, or a different number of visits than were delivered; a referral was required and expired. Authorization denials are particularly punishing because many are not appealable on the merits — the service may have been perfectly appropriate, and the answer is still no. The fix is procedural: know which of your payers require authorization for which service families, and never let scheduling outrun that check.

3. Documentation and coding mismatches

The note does not support the level or the service billed, the diagnosis does not support medical necessity for the procedure, laterality or units contradict the record, or a modifier is missing, wrong, or applied out of habit. This family is where clinical and administrative worlds collide, and where the honest answer is often that the care was right and the record was incomplete. Coding-driven denials cluster hard by specialty, which is the entire argument of why specialty matters in billing.

4. Timing problems

The claim was filed after the payer's filing limit, the appeal was filed after the appeal window, or a corrected claim was submitted so late that the correction no longer mattered. Timing denials are the only family that is unambiguously self-inflicted, and they are the reason a billing backlog is more dangerous than it looks. A claim sitting in a work queue is not neutral; it is depreciating.

5. Duplicate, bundling and coordination problems

The service was included in another service already paid, the payer sees the claim as a duplicate of an earlier submission, or two payers each believe the other is primary. Coordination-of-benefits stalls in particular can hold a balance for months while nothing appears to be wrong — no one denied the care, no one is paying for it either.

• Eligible
• Coded
• Scrubbed
• Submitted
• Adjudicated
• Posted

A denial you work is revenue recovered once. A denial cause you eliminate is revenue recovered every month afterwards.

Why practices keep re-solving the same denial

Most billing operations are structurally biased toward the single claim. The work queue shows claims, the productivity metric counts claims touched, and the satisfying feeling of closing one out is immediate. Root-cause work has none of that: it is slower, it requires talking to clinicians and front-desk staff, and its payoff shows up as an absence — claims that never denied and therefore never appeared anywhere.

The practical countermeasure is boring and effective. Categorise every denial into a small fixed taxonomy — the five families above are enough to start — and review the counts monthly with the people who create the inputs, not only with the people who work the output. When the same category tops the list twice in a row, the assignment is not "appeal harder." It is a workflow change with a named owner and a date.

What good denial handling looks like day to day

• Every denial is opened, categorised and dispositioned — worked, appealed, billed to the patient, or written off with a reason. Nothing closes silently.
• Corrections and appeals are distinguished on purpose. A claim that never reached adjudication does not need an appeal letter; see denials vs rejections.
• Filing and appeal deadlines are tracked per payer, and anything approaching a limit is escalated ahead of the rest of the queue regardless of dollar value.
• Appeals cite the record, not indignation. The strongest appeal is a short letter with the relevant documentation attached and the payer's own criteria addressed line by line.
• Category counts get reported to the practice in plain language, alongside the aging picture described in days in A/R.

Where the money actually leaks

The most damaging denials are not the loud ones. A large surgical claim denied for authorization gets attention within the day. The quiet leak is the small-dollar denial that is expensive to work relative to its value, so it gets deferred, deferred again, and eventually written off in a batch nobody reviews. Multiply a modest per-claim value by a steady weekly count and the annual figure is rarely modest.

This is where an explicit write-off policy earns its keep. Someone should decide, in advance and in writing, which balances are genuinely uneconomic to pursue and which are simply annoying. Without that policy the decision still gets made — just informally, by whoever is clearing a queue at the end of a long week.

Prevention beats recovery, but recovery still has to happen

None of this argues for neglecting the existing pile. Prevention changes next quarter; the balances already sitting in your aging buckets need working now, in filing-limit order. Both halves belong in the same operation, which is how we structure denial management and A/R recovery alongside day-to-day claim submission.

If you want a concrete starting point: pull your last ninety days of denials, sort them into the five families, and look at which family is largest. That single afternoon of sorting tells you more about your revenue cycle than any dashboard you have been shown, and it is the same exercise we run before we quote a practice.