Why claim denials are rising — and why most practices are fixing the wrong problem

Twenty years ago, I was the clinician justifying a treatment decision to someone who never examined the patient. Today I am the product person deciding what a front-desk screen should ask before that decision even gets made. Most writing on denials comes from people who saw the spreadsheet afterward. My view has mostly been the argument that produced it.

So let me walk through what the newest numbers are saying, and what to actually do about each one, because most practices are reading this data backward.

The number everyone quotes and the number they skip

Kodiak Solutions put out this report at the end of March, covering more than 2,300 hospitals and 375,000 physicians. The headline is that net revenue leakage grew 25%, and hospitals lost more than $48 billion to final denials and bad debt. Every trade publication ran that number, and fair enough — it's a big one.

The sentence after it matters more. Clinical denial rates, the ones tied to missing prior authorizations and disputed medical necessity, accounted for virtually all of that increase. Not coding errors. Not registration mistakes.

That distinction is where most practices lose their way. One kind of denial is a data problem: a wrong policy number, lapsed coverage, a typo in a name. Better training and a cleaner intake form fix that, reliably, almost every time. The other kind is a disagreement. A reviewer at a payer, sitting somewhere neither of us will ever see, decides after the fact that a service shouldn't have happened the way it did. No checklist resolves a disagreement, and that's the kind actually growing.

What to do: Split your denial tracker into two lanes at intake — clinical reason codes on one side, data or registration codes on the other — so you can see which one is actually rising before you spend a budget cycle guessing.

What practices say hurts versus what is actually hurting them

Splitting the tracker only helps if the two lanes are being read correctly, and that's exactly where a January poll from the Medical Group Management Association (MGMA) gets misread. MGMA ran this poll, asking 288 medical group leaders where their revenue cycle leaks the most. Denials and appeals came out on top at 48%. Front-end issues came in a distant second at 23%.

That looks like denials are the bigger problem and front-end issues are the smaller one. Sitting through enough of these reviews changes that impression, because denials are loud. Somebody has to open the letter, decode the reason and decide whether to fight it. A front-end mistake is quiet by comparison. A rushed eligibility check on a Tuesday doesn't turn into a denial until three weeks later, and by then nobody connects the two. The 23% figure isn't the real size of the front-end problem. It's the size of the problem people have learned to notice.

What to do: Put a name and a time stamp on every eligibility and authorization check, not just a checkbox. When a denial does land weeks later, that record lets you trace it back to the exact step that caused it, instead of guessing.

The trust problem hiding inside the workload problem

That same lag between an upstream mistake and its downstream consequence shows up again, at a much larger scale, in prior authorization. The American Medical Association (AMA) released this survey in May, based on 1,000 practicing physicians. It found physicians handle 40 prior authorization requests each week, with nearly a third of those requests often or always denied outright.

A number like that usually lands on a hospital board as a staffing request, and staffing requests miss what's actually going on, because a workload number invites a workload solution. Here's the part worth sitting with instead. Ninety-five percent of physicians say prior authorization delays care patients need. Only about one in four believe medical necessity denials are consistently reviewed by someone with the clinical training to make that call. A faster portal doesn't touch that second number. It just gets the same distrust confirmed a little quicker, on a nicer screen. This was never a speed problem. It's a credibility problem.

What to do: For service lines at high risk of denial, have a clinician or clinical liaison review the authorization request before it's submitted, not after it bounces back. That review is what actually moves the odds of a first-pass approval, not a faster submission tool.

Why hiring your way out of this stops working

If the fix for a credibility problem is a person who can speak the payer's clinical language rather than a faster tool, the next question independent practices usually ask is whether just adding staff gets there. It helps for a while, then stops holding up, for two reasons that show up the same way across very different practices. Payer rules change faster than any one new hire can track. One person can't keep up with dozens of payers, each quietly rewriting its own precertification requirements on its own calendar. And clinical denials need someone who can actually speak to medical necessity, which is a different skill than data accuracy and isn't something an afternoon of onboarding teaches.

What to do: Instead of adding a general front-desk hire, give one person real ownership of authorization and eligibility specifically, separate from the phones and check-in desk, with a standing rule sheet of which payers and service lines need prior authorization confirmed before scheduling, updated monthly as those rules change.

None of the four steps above needs new software first. What they need is a practice willing to admit that the fastest-growing category of denial isn't a data entry problem, even though that version of the problem would be so much easier to solve. Too many practices treat it like one anyway, because the comfortable answer is always more tempting than the correct one.

The practices holding up best right now aren't asking why they keep getting denied. They're asking which kind of denial is actually growing, and building their response around that answer.

Dr. Giriraj Tosh Purohit is product owner at OmniMD, where he leads development of clinical and revenue cycle technology, including EHR, RCM and AI-driven front-desk solutions. His background spans clinical workflows, HL7/FHIR interoperability standards and health care product management across two decades in the U.S. health care industry.