A molecular or genetic test can be ordered by the right physician, for the right patient, performed flawlessly in a CLIA-certified lab, and still come back denied. For diagnostics and genetic labs, denials are not an unlucky edge case. They are a structural feature of how molecular testing gets reimbursed, the rate is high, and it is rising. The encouraging part, and the reason this article exists, is that most of those denials are preventable, because most of them were never really about the medicine.
Disclaimer: This is educational, not billing, legal, or medical advice. Payer policies and coverage criteria change frequently and your situation may differ from the examples here. Always verify current requirements with the payer's most recent published policy and consult qualified billing or compliance professionals. Use this information at your own risk.
The denial numbers are worse than most teams assume
Start with the most rigorous evidence available. A 2025 cohort study in JAMA Network Open, conducted by Georgetown University researchers and funded by the National Cancer Institute, analyzed 29,919 cancer-related next-generation sequencing (NGS) claims among 24,443 unique Medicare beneficiaries. It found that 23.3% of those claims were denied between 2016 and 2021, and the denial rate climbed across the study period.
The trend underneath that number matters more than the headline. The claim denial rate was 16.8% before the National Coverage Determination for NGS (NCD 90.2), 20.3% after the NCD took effect in 2018, and 27.4% after the NCD was amended in 2020 to include certain hereditary breast and ovarian cancer mutations. Read that sequence again: coverage expanded twice, and denials went up both times. That is the central paradox of molecular reimbursement, and we will come back to why it happens, because it explains almost everything about how to prevent denials.
Industry data points the same direction. XiFin's 2024 Payor Denial Impact Report, based on more than 20 million laboratory claims, reports that roughly 35.3% of billed molecular CPT codes are denied (industry data rather than peer-reviewed research, but directionally consistent with the Medicare-specific findings above). And the exposure per denial is not trivial: in the JAMA Network Open study, the median charge on a denied NGS claim was $3,800, which the authors describe as the upper limit of provider or patient liability for each denied claim. Put a denial rate in the twenties or thirties against a four-figure median charge across your annual volume, and denials stop being a billing nuisance and start being a line on the income statement.
Where you sit changes your odds
Denials are not distributed evenly, and the JAMA Network Open study is unusually specific about who absorbs the most. Claims for NGS testing were more than twice as likely to be denied when performed by an independent laboratory (odds ratio 2.76) or another non-hospital site (odds ratio 2.55) than when performed by a hospital. For independent labs specifically, the NGS denial rate reached 37.3% in the most recent period the study examined.
Panel size compounds it. The study found that claims for 50 or more genes were more likely to be denied than smaller panels for solid tumors. So the profile most exposed to denials, an independent molecular lab running large, comprehensive panels, is precisely the profile that defines modern oncology diagnostics. If that describes your lab, upstream prevention is not a back-office tidiness project. It is a core financial control.
Most denials are operational, not medical necessity
Here is the reframe that changes how a team responds to a denial. The instinct, after a rejection lands, is to assume the payer weighed the clinical case and disagreed. Usually they did no such thing. The claim failed a structural check before any reviewer ever evaluated medical necessity, and the denial code is a symptom of a clerical or coding gap, not a clinical verdict.
The most common operational causes are worth understanding individually, because each has a different fix:
- A missing or invalid MolDX Z-code. The single most preventable hard denial, and now an outright gate (more below). A claim can be perfect in every other respect and still be rejected on this alone.
- Diagnosis-code positioning. The ICD-10 code that actually supports coverage is on the claim, but it is buried below a non-covered indication, or it is not linked to the right line item. The information is present; the structure is wrong, and the payer's system reads "not covered."
- A missing test name or descriptor on unlisted codes. Many molecular tests bill under unlisted codes such as CPT 81479 (and 81599, 84999). Those codes describe a category, not a specific assay, so the payer needs the test name spelled out to adjudicate. Leave it off and there is nothing for the payer to approve.
- Documentation mismatch. The encounter note, the order, and the claim disagree about the test performed, the indication, or the date of service. Each document is fine in isolation; together they tell three slightly different stories, and the payer resolves the conflict by denying.
- Prior authorization not obtained where the payer required it, often because the requirement changed and no one caught it.
- Experimental or investigational classification, where the payer's medical policy has not caught up to the test's evidence base, or the claim did not cite the policy's own covered indication.
Every item on that list is correctable before the claim goes out. That is the whole argument of this article: the denial rate is high, but a large share of it is addressable at the rule-and-scrub stage, where a fix costs minutes, rather than at the appeal stage, where it costs weeks.
The MolDX Z-code is now a hard gate
The Z-code deserves its own section because its enforcement changed. For labs billing in MolDX jurisdictions, the DEX Z-Code is no longer a soft expectation. Effective May 1, 2025, MolDX claims submitted without the Z-Code identifier in loop 2400 (SV101-7) deny as unprocessable, per Noridian guidance. There is no clinical review and no partial adjudication; the claim is rejected on sight. Labs have to register the test, obtain the Z-Code through the DEX Diagnostics Exchange, and submit exactly one Z-Code per MolDX CPT line, correctly matched to the assay the code represents.
This is the cleanest possible illustration of a denial that has nothing to do with whether the test was appropriate. A clinically flawless order with a missing or mismatched Z-Code is an automatic loss, decided by a field on a claim form rather than by a reviewer's judgment.
Why the denial rate keeps climbing
Return to the paradox from the top: as Medicare broadened NGS coverage in 2018 and again in 2020, the denial rate rose from 16.8% to 27.4%. If more was covered, why were more claims denied?
Because coverage and payability are not the same thing. A test can be covered in principle and still denied in practice if the claim does not satisfy every specific condition the policy attaches to that coverage. Each expansion of NCD 90.2 arrived with more criteria, the right cancer stage, the right prior-testing history, the right ordering context, the right documentation, and every additional requirement is one more edge a claim can fall off. Broader coverage, written as more detailed rules, produces more denials unless the lab keeps perfect pace with the rules.
That is the deeper reason most molecular denials are "operational." The dominant cause is not payers refusing to cover good tests. It is the gap between what current policy requires and what the claim actually reflects, and that gap widens every time a policy changes, which is constantly. A coverage fact that was accurate when a billing protocol was written quietly goes stale, and the first sign of the drift is a denial pattern no one can immediately explain.
Appeals work far more often than teams expect
If prevention is the first lever, appeals are the second, and the data here is genuinely surprising. A 2025 Health Affairs study from UCSF researchers found that 46% of appealed genetic-testing denials were overturned through independent medical review (IMR) between 2019 and 2023. Seventy percent of the claims examined were cancer-related. Overturn rates varied by state (roughly 60% in California, under 30% in Washington) and by indication (54.8% for ovarian cancer testing, 44.3% for non-small-cell lung cancer).
Now the other half of the finding: fewer than 1% of denials are appealed through IMR. Hold those two numbers next to each other. Nearly half of the genetic-testing denials that get challenged are reversed, and almost nobody challenges them. The reversible revenue is sitting on the table, and the reason it stays there is rarely that the cases are weak. It is capacity. Appeals take skilled staff time, so labs triage, chase the cleanest and largest cases, and write off the rest, which means the true cost of a high denial rate is not only the claims you lose but the winnable ones you never reach.
Building even a modest, repeatable appeals process changes that math. The appeals that succeed tend to share a structure: they identify the exact denial reason rather than arguing in general terms, they quote the payer's own coverage criteria and walk through how the patient meets each one, and they attach the specific documentation, pathology, prior results, the ordering rationale, that proves it. None of that requires heroics. It requires knowing what the governing policy actually says and matching the appeal to it, which is the same capability that prevents the denial in the first place.
The prevention checklist
The data points to a short list of upstream controls that head off the most common losses:
- A valid MolDX Z-Code on every applicable molecular claim, every time, matched to the right assay.
- Diagnosis-code positioning that maps the claim to the payer's covered indication, in the right order.
- The test name and descriptor populated on every unlisted molecular code.
- Documentation that agrees with itself, order, encounter note, and claim aligned on test, indication, and date.
- Prior authorization confirmed before the service wherever it is required, against the payer's current list.
- A standing process to appeal medical-necessity denials, given the roughly 46% overturn rate.
The throughline connecting every point above is this: molecular and genetic denials are decided overwhelmingly by whether the claim satisfies the payer's current, specific rules, not by whether the test was the right call clinically. The labs that get paid are the ones that treat those rules as living data, tracked, encoded, and kept current per payer, per test, per indication, and applied before the claim is ever submitted. That is what turns a 23% denial rate into a recoverable one, and it is the problem Converus was built to solve.
Sources
- Claim Denials for Cancer-Related Next-Generation Sequencing in Medicare — JAMA Network Open, April 18, 2025 (doi:10.1001/jamanetworkopen.2025.5785; PubMed 40249617)
- Use of Independent Medical Review: Almost One-Half of Coverage Denials Overturned — Health Affairs, December 2025 (doi:10.1377/hlthaff.2025.00716)
- 2024 Payor Denial Impact Report — XiFin, Inc. (analysis of 20M+ laboratory claims)
- Proper Submission of DEX Z-Code for Molecular Diagnostic Services (MolDX) Claims — Noridian Healthcare Solutions (effective May 1, 2025)
- National Coverage Determination (NCD 90.2): Next Generation Sequencing — Centers for Medicare & Medicaid Services