Most denied claims were decided before anyone touched a billing code

A denied claim feels like a billing problem. It arrives from billing, it is worked by billing, and it is discussed at the billing meeting.

Quite often it was decided at check-in, three weeks earlier, when somebody typed a policy number.

What is reasonably solid

Initial denial rates have been rising. Analysis of hospital and physician claims data by Kodiak Solutions, reported through HFMA, put initial denials at 11.8% in 2024 — up from single digits a few years earlier. That is roughly one claim in eight rejected on first submission.

That number comes from claims data rather than a survey, which is why we lead with it.

What is quoted everywhere and is much softer

You will also see, repeatedly and confidently:

• "Registration errors cause 30% of denials"
• "Insurance verification errors cause another 20%"
• "Front-office errors cause 65% of all denials"
• "86% of denials are preventable"

We could not trace any of these to a primary study. They circulate between revenue cycle blogs, and where they carry an attribution at all it is usually to a vendor selling eligibility-verification or registration software — that is, to a company whose product looks better the larger the number is.

That does not make them wrong. Eligibility and registration errors are obviously a real and large category of denial; anyone who has worked a denial queue knows it. It makes them unverified, and we are not going to repeat a number we cannot source just because it supports what we sell.

How to get your own number, which is better than any of these

Your clearinghouse already has this. Pull three months of denials and group them by denial code. You are looking for the front-end family:

| Denial category | Typical codes | | --- | --- | | Coverage terminated / not in effect on date of service | CO-27 | | Patient cannot be identified as our insured | CO-31 | | Non-covered under the patient's plan | CO-96, CO-204 | | Prior authorization missing | CO-197 | | Duplicate claim | CO-18 |

Every one of those was determined before the encounter was coded. Add them up as a share of total denials and you have a real figure for your practice, which is worth more than anyone's industry average.

Then do the arithmetic that matters: multiply by your average claim value and by how many of those you actually rework rather than write off.

Why this sits at the front desk and not in billing

Because the fix is upstream. Verifying eligibility before the visit, catching a plan that terminated on the first of the month, noticing that the authorization on file expired — none of that is billing work. It is front-office work that happens to determine billing outcomes weeks later.

It is also the clearest example of something that has to be done for the practice but does not have to be done at the practice, by the person who is also greeting patients and answering the phone.

---

Denial rate: Kodiak Solutions analysis reported via HFMA, 2024. The registration and preventability percentages above are widely circulated but, as far as we can establish, vendor-sourced and not traceable to a primary study — treat them accordingly.