5 Documentation Mistakes That Get Biologics Prior Auths Denied
About 30% of biologic prior authorization denials have nothing to do with whether the treatment is clinically appropriate. The patient needs the drug. The prescriber knows it. The payer's own criteria might even support it. But the submission gets denied anyway — because the documentation didn't make the case clearly enough.
That's an expensive problem. Biologics run anywhere from $15,000 to over $100,000 per year. A denial means an appeals process, a delay in care, and hours of staff time that most practices don't have to spare.
After working with dozens of rheumatology, gastroenterology, and dermatology practices on their PA workflows, we've seen the same five mistakes come up over and over. Here's what they are, why they kill submissions, and what to do instead.
1. Missing Step Therapy Documentation
Most biologics require proof that the patient tried cheaper alternatives first and didn't respond — or couldn't tolerate them. Payers call this step therapy, and their reviewers are looking for a specific paper trail.
The problem isn't usually that the patient skipped step therapy. It's that the documentation doesn't clearly show the failure timeline. You might have notes scattered across three different visit records, a lab result buried in an attachment, and a medication history that doesn't explicitly say why the prior drug was stopped.
A payer reviewer spending four minutes on your submission won't reconstruct that story for you. If the failure isn't obvious and explicit, the box doesn't get checked.
Fix it: Create a clear, sequential narrative: drug started, dose, duration, documented inadequate response or adverse event, drug discontinued. Each step should be explicitly stated — not implied. If the patient tried methotrexate for six months and developed hepatotoxicity, that needs to say exactly that, not just "patient did not tolerate prior therapy."
2. Incomplete Disease Activity Scores
Payers reviewing biologics for rheumatoid arthritis, Crohn's disease, or psoriatic arthritis aren't just looking for clinical impressions. They want numbers. Specifically, they want the validated scoring tools that their LCD criteria reference by name.
For RA, that means DAS28 or CDAI scores. For Crohn's, CDAI or Harvey-Bradshaw. For PsA, HAQ-DI or DAPSA. A note that says "patient has moderate-to-severe disease with significant functional impairment" reads as subjective to a utilization reviewer — even when it's accurate.
Here's a real scenario: a rheumatology practice submits a PA for adalimumab with thorough clinical documentation and a clear step therapy history. Denial comes back: "insufficient evidence of moderate-to-severe disease activity." The chart has detailed exam findings but no documented DAS28 score. The reviewer had no quantifiable threshold to approve against.
Fix it: Know which scores your target payer's policy references for each indication, and make sure those exact scores are documented and included in your submission. If the score wasn't captured at the last visit, it needs to be — before you submit.
3. Wrong Diagnosis Code Specificity
ICD-10 was designed to capture clinical specificity, and payers have built their coverage policies around that specificity. Using an unspecified code when a specific one exists is one of the fastest paths to a denial.
Take rheumatoid arthritis. M06.9 — "rheumatoid arthritis, unspecified" — exists in the codebook. But a payer's biologic coverage policy may specifically require M05.xx or M06.0x codes that indicate seropositive or seronegative RA with specific joint involvement. Submit with M06.9 and you may fail their automated eligibility check before a human even looks at it.
Same issue in dermatology. L40.9 (psoriasis, unspecified) versus L40.0 (plaque psoriasis) — these are not interchangeable on a biologic PA, even if clinically the diagnosis is obvious.
Fix it: Pull the actual payer policy before you code the submission, not after. Most payers publish their LCD/NCD criteria online and list the specific ICD-10 codes they recognize for biologic coverage. Use those codes. This takes ten minutes of upfront research and prevents days of appeals work.
4. Generic Documentation That Ignores Payer-Specific Criteria
This one is subtle but it accounts for a huge share of preventable denials. Every commercial payer — and Medicare Advantage plan — has its own coverage criteria for biologics. Aetna's requirements for ustekinumab are not the same as Cigna's. United's step therapy requirements for TNF inhibitors differ from BCBS in ways that matter.
Most PA submissions are written generically. The same letter template gets sent to every payer. It covers the clinical bases but doesn't specifically address what that payer's policy actually asks for.
When a reviewer is checking your submission against their internal checklist, missing criteria get flagged as missing — even if the clinical record would support approval if it were framed differently. Generic documentation forces reviewers to interpret and infer. Payer-specific documentation gives them what they need to check the box.
Fix it: Before writing the PA letter, look up the specific payer's current policy for the drug and indication. CMS publishes Medicare coverage policies in their online database. Commercial payers publish LCDs on their provider portals. Your documentation should address each criterion in that policy, in language the reviewer can map directly to their checklist.
This is time-consuming to do manually, which is why most practices don't do it — and why AI tools that can research payer policies on demand are changing the math here.
5. Stale Clinical Data
Payer policies often include recency requirements for the clinical data supporting a biologic request. Most want lab values, disease activity assessments, and imaging within the past 30 to 90 days. Submit documentation with a DAS28 score from eight months ago and you'll likely get a denial citing "insufficient current clinical information."
This seems obvious in theory. In practice, it gets missed constantly — especially for patients who've been on a biologic that's being renewed, or when a submission gets delayed in the queue and the supporting data ages out before it's actually sent.
CRP levels, inflammatory markers, functional assessments, imaging results — all of these have expiration dates from a payer's perspective. A liver function panel from before the step therapy trial doesn't satisfy a requirement for current labs.
Fix it: Build a checklist that includes data recency requirements for your most common payers and indications. Before submitting, verify that every data point in the submission falls within the required window. If something is expiring soon, coordinate with the patient's care team to get updated values before you submit — not after you get denied.
The Pattern Behind All Five Mistakes
Look at this list and one thing stands out: none of these are clinical failures. They're documentation failures — gaps between what's in the chart, what the payer needs, and what actually makes it into the submission.
The clinical case is often completely solid. The patient needs the medication. The prescriber has the evidence. But the submission doesn't translate that clinical reality into the specific language, scores, codes, and criteria the payer is evaluating against.
This is the problem that AI-assisted PA documentation is actually well-suited to solve. Not replacing clinical judgment — that stays with the physician. But automating the research-heavy, criteria-specific work of translating a clinical case into a payer-ready submission.
Luma does this by pulling the current policy for the specific payer and drug combination, identifying what criteria need to be addressed, and generating documentation that maps directly to those requirements. Step therapy timelines get structured clearly. Disease activity scores get included. Diagnosis codes get verified against the payer's coverage criteria. Data recency gets flagged before submission.
The administrative burden of prior authorization is well-documented — practices spend an average of 13 hours per physician per week on PA-related work, according to AMA data. Most of that time is spent on exactly these kinds of documentation issues.
Getting the documentation right the first time doesn't just improve approval rates. It eliminates the appeals cycle, reduces the administrative load, and — most importantly — gets patients on treatment faster.
If you're seeing biologic PA denials that feel like they shouldn't be happening, these five areas are the right place to start. Fix the documentation, and the clinical case you've already built will do its job.
Sources:
1. CMS Medicare Coverage Database — cms.gov/medicare-coverage-database
2. AMA 2023 Prior Authorization Survey — ama-assn.org
3. Health Affairs — Administrative Burden of Prior Authorization — healthaffairs.org
4. ACR — DAS28 and Disease Activity Scoring in RA — rheumatology.org
5. AMCP — Biologics Prior Authorization Policy Analysis — amcp.org