Medicare’s own numbers tell the story quickly. In 2024, Part B paid $8.4 billion for lab tests. Genetic tests were about five percent of the tests performed, yet they took 43 percent of the dollars, roughly $3.6 billion, and spending on them grew about 20 percent in a single year while the rest of the fee schedule barely moved. If you run a molecular lab, you already know what that curve feels like from the inside. More volume, more payer attention, and a billing process that somehow gets harder every quarter.
Most lab teams start with some version of the same question: why are our claims denying? It’s a fair question, but it’s not quite the right one. A molecular claim asks the payer to solve three different problems at once, and the denial letter almost never says which one failed. Until you know which problem you’re actually looking at, every fix is a guess.
The spending curve explains the scrutiny
When one category of testing grows from a rounding error to nearly half of Medicare’s lab spend, payers respond the only way they know how: they add friction. More front-end edits, more prior authorization, more requests for records. The Office of Inspector General publishes these spending analyses year after year, and year after year genetic testing is the line that draws the attention.
None of this is an accusation. It’s arithmetic. But it creates a problem that honest labs inherit, because friction built for bad actors lands on everyone. To an automated edit, a legitimate pharmacogenomic panel and a fraudulent one look about the same. So the burden of looking different falls on the lab, and looking different means the order is clean, the indication is clear, and the documentation matches the policy before anyone asks for it. That isn’t fair, exactly. It’s just the environment now.
Three problems hide inside the word “billing”
Start with identity: does the payer even know which test this is? Broad CPT codes were never built for proprietary panels, which is why PLA codes and DEX Z-Code identifiers exist at all. In MolDX jurisdictions the question is formalized, and Palmetto GBA runs an entire program for identifying molecular tests and deciding how to cover them. We wrote separately about test identity, because it fails in its own particular ways.
Coverage is a different question: is this test, for this patient, with this indication, something the policy actually pays for? That’s where local coverage determinations, technical assessments, and prior authorization live. And price is different again. What a code pays is set by the Clinical Laboratory Fee Schedule, shaped by crosswalk and gapfill decisions, and increasingly governed by PAMA.
The distinction matters because each problem has a different fix. Identity problems get solved with registration, code mapping, and report language. Coverage problems get solved with documentation, and sometimes with a fight. Price problems get solved with contracting and menu strategy. A team that treats every molecular denial as an appeal-writing exercise is using one tool on three different jobs, and it shows in the recovery rate.
The price of a test is a moving target
PAMA sets Medicare’s lab rates from reported private-payer data, and that machinery is running right now. The current reporting window opened May 1 and closes July 31, 2026, covering payments from the first half of 2025. The statutory rate cuts, capped at 15 percent a year, are paused through the end of 2026 and scheduled to resume in January 2027, when rates built on this summer’s data take effect.
Sit with what that means for planning. A test that anchors your menu today may pay meaningfully less in eighteen months, and the size of the cut depends on numbers being assembled across the industry as you read this. So what’s a test worth next January? For a lot of codes, the honest answer is that nobody knows yet. A lab modeling its menu against a frozen fee schedule is modeling a world that no longer exists.
Coverage moves slower than the science
A molecular menu can change monthly. Coverage policy can’t, or at least it doesn’t. New assays and revised intended-use language take time to move through payer review, and in that gap the lab keeps performing tests while the claims pend, develop for records, or quietly deny. The uncomfortable question for leadership isn’t whether that gap exists. It’s who is watching it, and whether anyone can name the tests being billed into it right now.
Regulation is starting to push on response times, though unevenly. Under the CMS Interoperability and Prior Authorization Final Rule, the payers it reaches, including Medicare Advantage plans, have to decide expedited prior authorization requests within 72 hours and standard ones within seven calendar days, with electronic infrastructure behind it. That’s real progress. It’s also narrower than it sounds, because commercial plans outside the rule keep their own clocks, and a faster deadline doesn’t lower the documentation bar. It just moves the date.
Questions worth asking inside the lab
- Can billing say, for every assay on the menu, which payer policies apply and what documentation they expect?
- Does the requisition capture the indication the policy will ask about, or does that information first show up after the denial?
- Who confirms the code, the identifier, and the report language before a new or revised assay starts billing?
- Which denials get routed to a correction or an information response instead of a reflex appeal?
- When a claim develops for records, how many handoffs does the request survive before it leaves the building?
- Does anyone own the repeat-denial pattern by payer, test, and ordering account, or is every denial worked like it’s the first?
- If the PAMA cuts resume in 2027, which tests on the menu stop covering their own cost?
What a lab can actually do
The labs that get paid well for molecular work aren’t doing anything exotic. They’re sequencing the work differently. Reimbursement happens as part of the test rather than after it: identity is registered and mapped before launch, ordering documents are built to capture what policies actually ask, and someone reviews orders for readiness before the claim goes out the door. Downstream, denials get routed by cause. Information requests get answered with the record instead of a form letter. Appeals are saved for the claims where the documentation can win.
Then there’s the part most labs skip: reporting the pattern. A monthly view of denials by payer, test, and cause turns billing from a cost center into an intelligence function. It tells you which payer changed behavior, which ordering account keeps sending incomplete indications, and which fights are worth escalating. Without it, the same denial gets rediscovered every month by whoever happens to touch it.
Some labs build this muscle internally. Others borrow senior capacity for the pieces that need it. Either way, the test of a working process is simple. When a molecular claim fails, can someone tell you quickly whether it failed on identity, coverage, or price, and what the next move is? If that answer takes a meeting to produce, the process is the problem.
Where this is heading
The direction of travel isn’t mysterious. Genetic testing keeps growing as a share of lab medicine, so the scrutiny grows with it. Payers keep automating review. Regulators keep tightening timelines for the payers they can reach. And PAMA’s 2027 reset will reprice a meaningful part of the fee schedule using data being reported this summer.
The science is not slowing down, and the payment system is not speeding up much. The distance between those two facts is where molecular labs either build a discipline or bleed. The ones that treat reimbursement as part of the product, with the same version control and ownership they apply to the assay itself, are the ones that will get paid for the medicine they’re actually practicing.
Important note
This article is for general educational purposes only and is not legal, billing, coding, coverage, or reimbursement advice. Coverage, documentation, appeal, and payment requirements vary by payer, plan, jurisdiction, contract terms, and claim facts.