NEW YORK – When Gabe Bien-Willner came into Medicare Administrative Contractor Palmetto GBA as its new chief medical officer in 2018, things began to change.
Seeing that the MAC — the leader of Medicare's MolDx program, which establishes coverage and reimbursement for diagnostic tests across about 30 states — was writing a lot of very specific local coverage determinations (LCDs) covering individual tests that used technologies that were becoming more widespread, he decided to make a key improvement: broadening the scope of the LCDs to encompass more than one test.
"It seemed like a very inefficient process," he said of the way things were being done when he arrived at Palmetto.
That decision has largely been viewed as a positive by diagnostic test manufacturers, but others noted downsides to the new approach, including an opaqueness over what exactly will be covered.
The shift to foundational LCDs — LCDs written with a broader scope that encompass a type of test, rather than just a single product — started in 2018 and is ongoing, and it has come with its own challenges, Bien-Willner said. Because they are more comprehensive and intended to apply to more than just one test, they take much longer to write and can be more complicated. Bien-Willner said that the average time to create a foundational policy is between 12 and 18 months from start to going into effect.
The major cause of delays in implementation, however, are from new requirements due to the 21st Century Cures Act, which was enacted in 2016 to accelerate medical product development. Palmetto and the MolDx program received the new rules set by the Centers for Medicare and Medicaid Services as a result of the law in 2019, he added.
But once that foundational coverage decision has been unveiled, the process for getting a product reimbursed under it is significantly faster: about a two-month wait time, rather than the nine to 12 months it took to write a separate LCD under the old system, Bien-Willner said.
Writing the LCD is the shortest part of the decision-making process — there's a lot more that goes into the crafting of a decision. Previously, there was no defined procedure to begin the LCD process, but under the 21st Century Cures Act, there are two ways an LCD can be initiated.
First, a test manufacturer with a new or innovative technology can request a new policy in writing and provide its evidence. The MAC has 60 days from that request to notify the company about whether its request is complete or not, and if the submission is complete, Palmetto must write a policy for it, Bien-Willner said. However, it is up to the MAC to prioritize its LCDs, so there's no guarantee that it will be written immediately — Palmetto has a "pretty big backlog of policies" to write.
The MAC can also write up a policy on its own initiative depending on what it sees in the industry, Bien-Willner noted. If there's a provider offering a new service that comes to Palmetto and asks for a coverage policy, Palmetto can review the evidence and write that new policy "in such a way that … if somebody performs a similar service, [they] could also potentially be covered," as long as they meet the criteria for coverage laid out in the LCD.
The move toward foundational LCDs doesn't mean that product-specific coverage decisions have been done away completely — there are some services that may only ever be offered by one manufacturer or provider, such as gene expression profiles that require training large data sets, Bien-Willner said. However, the two most recent product-specific LCDs came out in September 2020, covering Natera's Signatera minimal residual disease test and CareDx's AlloSure test for kidney and heart transplants. Even the Natera and CareDx specific LCDs, however, were left open to potentially be applied to similar tests from other vendors.
The "philosophical change" means that MolDx now writes policies that cover analytes, rather than specific tests or vendors, Bien-Willner said. "Provided that another lab measures the same analytes, they could theoretically get coverage under an existing policy where another lab has already demonstrated clinical utility and clinical validity," as long as that second lab demonstrates analytical validity in measuring the analytes.
Once a policy has been initiated, there are multiple steps that need to be taken to get an LCD to its final form. A technology assessment must be completed, a process where the Palmetto team reviews "peer-reviewed, published evidence to determine if those tests, if those services, are reasonable and necessary," he said. Since the advent of foundational LCDs, what has changed with the technology assessment is mainly the level of expertise within Palmetto. When Bien-Willner joined, he was the primary expert creating every policy and reviewing all of the data. Now, there are four experts reviewing that information, which "hopefully that means we render better decisions," he said.
There is also a technical assessment, which follows the same approach but comes into play once a policy has been finalized. That assessment is dependent on existing policies and what is defined as reasonable and necessary in those policies and is meant to determine whether a specific test meets the requirements for coverage. Two different LCDs may have different procedures or requirements for coverage, and information from the technical assessment is shared directly with the provider seeking coverage. That assessment determines if the test meets the criteria or where it needs to be changed, Bien-Willner said.
Before the 21st Century Cures Act, Bien-Willner said the technical assessment process was "very ill-defined" and confusing for providers and diagnostic test manufacturers. Now, it has been "operationalized and structured in a way to be highly reproducible" to "create a procedure for every kind of service" and make sure "every single service" that comes into the process is reviewed and handled the same way. With each new policy, the procedure for evaluation of tests under that policy is also established.
Labs are now "establishing a chain of evidence from beginning to end to demonstrate something that is reasonable and necessary," he said.
The technical assessment also serves to catch "bad actors," since the assessment will get down "to very granular details regarding the performance of that test" for analytical validity, he said. Clinical validity must also be demonstrated if the test has a different intended use or is measuring those analytes in a different way. If multiple labs are measuring the same analytes, the clinical validity and clinical utility don't change and don't need to be proven by an individual provider. In that case, only the analytical validity of the specific test must be shown.
"It's the analyte that has value, not the person that's processing the sample," he said.
Once that validity is proven and a policy is written, Palmetto has to get buy-in from the other MACs in MolDx, have CMS review the LCD, put the draft up for public comment, review those comments, and edit the draft. The final version is then sent back to CMS, and notice is provided of the final LCD before it goes into effect.
Pre-21st Century Cures, foundational policies weren't needed because every test was required to be reviewed and have a separate LCD published. In MolDx's view, the change has been a good one because these broader policies "have a lot more longevity and allow a lot more providers to obtain coverage for useful and beneficial services," according to Bien-Willner.
"We're seeing an explosion of molecular diagnostic tests … [and] we're going to see more of these tests, not fewer," he said. These services are "the forefront of improved technologies and the improvement of medical services."
Stakeholder reactions
Bien-Willner noted that when the transition first started, not everyone was on board. Stakeholders "resisted changing" at the beginning, but the rumblings of discontent have largely subsided — at least in Palmetto's jurisdiction, which includes North and South Carolina, Virginia, West Virginia, Alabama, Georgia, and Tennessee. Even when labs don't get coverage, the MAC lays out where the provider failed to meet the criteria and has a dialogue explaining its reasoning, Bien-Willner said.
Rob Dumanois, the director of reimbursement strategy at Thermo Fisher Scientific, said that foundational LCDs "[create] a much clearer path around all three legs of the reimbursement stool," meaning coverage, coding, and pricing. The broader coverage determinations allow "an existing lab to understand what's on [the] market, what's required as far as test content is concerned," as well as the validation requirements. It also provides "more clarity around what disease indications … have been covered or are in scope."
Now there are "more accurate building blocks," as well as an idea of the minimum requirements when it comes to panel content. "What it ultimately yields is still a mystery for all of us, but there's more transparency around where value can be assigned than if this process doesn't exist."
On behalf of CMS, the process also "ensures that if MolDx is paying for a test, it's been evaluated and is a robust test," he said.
Another key benefit that both Dumanois and Bien-Willner emphasized is the speed with which a test can be covered, particularly if it falls under an existing policy. The year-plus turnaround time is only relevant when a manufacturer is first to market, Dumanois said — when submitting a dossier for a test under preexisting coverage determination, the turnaround time is more like a few weeks.
When it comes to that actual dossier, the requirements have become "more explicit," with "pretty detailed guidelines" on what is needed for a technical assessment, Dumanois said.
Coding and pricing have also become more transparent with foundational LCDs, he added. The MolDx program uses separate billing and coding articles under an LCD that lays out diagnostic and procedure code scenarios for certain tests. The DEX Diagnostics Exchange from MolDx also publishes individual assay payment levels.
If articles require labs to get a so-called Z code that is unique to one test, this can impact the reimbursement a lab gets. If there is a CPT code that is specific to the type of test — such as 81445 for comprehensive genomic profiling — the lab's Z code will be used to help determine the test's pricing, which generally aligns with Medicare's Clinical Fee Schedule. The price can change based on different variants, some of which are required and some of which are additional.
For some multiplex tests, MolDx requires labs to bill a Z code with the 81479 CPT code for an unlisted molecular pathology test. That Z code provides more information on what exactly is being performed and how it's being used, which MolDx can then incorporate when determining pricing, Dumanois said.
Overall, for a manufacturer seeking a new policy, "the road map is relatively clear." There's the chance to start discussions with MolDx "pretty early in the process," with Palmetto having a "willingness … to provide some guidance on how they'd like to see data presented to them," he added. It also may be able to provide insight into some gaps or potential blind spots at the beginning of the journey. A test developer can "begin with the end in mind."
Not everyone has been so effusive in their praise of the program, however. An industry insider who requested anonymity used the recent LCD for molecular syndromic panels for infectious disease as an example of the drawbacks of the system, particularly its impact on the broader payor landscape. Some private payors use Medicare's coverage determinations as guidance for their own plans, despite the coverage policy being developed for a different population, and if a MolDx coverage policy has a restricted scope, there can be little recourse for patients covered under a commercial payor to get a policy changed. If somebody questions a policy, the payor can say "Medicare said it was OK, and it's a federal government program," the insider said. "It really does make it easier to limit access to testing."
Meantime, reimbursement expert and consultant Bruce Quinn has taken the shift to foundational LCDs as a mixed bag, noting that while the broader LCDs are "more open-ended as to what they will cover in the future," that also lends itself to vagueness and a lack of clarity. Coverage can evolve as new clinical applications appear, and future products can get approved more quickly because the LCD "says a certain kind of care is inherently reasonable and necessary." As a result, test developers "don't need to invent the wheel again" and "only need to show that they're accurately capturing the biomarker" that has been deemed useful by MolDx.
Larger questions of what exactly is covered may not be as clear, he said. MolDx has templates available on its website for the technical assessments that tests have to go through, but they are very specific — focused more on things like specimen stability, Quinn said. "You could still get into an argument that some extension of the LCD is medically reasonable or not. That's where you could still have a vague area."
Quinn also noted that it could stifle innovation in the industry. There's a "risk that this directs a lot of resources toward replicating things that have already been done" in hopes of getting reimbursement more quickly. "It's still very time-consuming and uncertain to do something that's truly novel." "With the advent of these foundational LCDs, and the devotion of a lot of resources to a lot of fast-follower test approvals, MolDx will have to be sure it still devotes enough time and energy to novel products," he added.
John Warren, founder of Gettysburg Healthcare Consultants and a former CMS official, said that MolDx has been "trying to balance their requirement to make comprehensive … coverage decisions with their desire to do that in a timely manner." The full LCD process for individual tests can "pose a significant workload burden on Palmetto's staff" that could result in delays in coverage. "As technology continues to advance and more and more companies are getting into the molecular diagnostic space," MolDx is making sure it doesn't get overwhelmed, he said.
But Warren echoed Quinn's thoughts that the "trade-off for … faster turnaround time is transparency." For a test-by-test procedure, everyone can see exactly how Palmetto came to its coverage decision for that test. A more general LCD has a "little bit less of that transparency," with CMS using their billing and coding articles to confirm whether a specific test is covered and what the pricing will be, he said. Those articles have no comment period or opportunity for public decision-making, and it's possible for two tests to meet the requirements of an LCD without receiving matching coverage, he said.
Some manufacturers may also view the lack of transparency as taking away the opportunity to learn from competitors and requiring them to create their own pathway to coverage, rather than being able to follow another manufacturer's process, he said.
Because these LCDs take so long to write, getting them changed if there's an issue isn't easy, although that's why the public comment period exists. Bien-Willner noted that there are mechanisms in place to rescind or change an existing LCD. Providers can request reconsideration of an existing policy, and if it turns out that assumptions were made about a test's or technology's validity or there was insurmountable bias, data could come in that would overturn a policy — it has happened before, although not recently, and it is rare. It is "up to the community" to tell Palmetto that it was wrong about an LCD, he said.
You "do your best with your expertise … and determine whether you think that the bar for reasonable and necessary is met," but it is subjective, he said. "The only thing that will change our mind is new evidence," and "it will happen again."
"That's evolution," he said.