Skip to main content
Premium Trial:

Request an Annual Quote

CMS Stance on Reimbursement of Multi-Analyte Algorithm Assays Raises Industry Concerns


Originally published Sept. 18.

Developers of molecular diagnostics and genetic tests are concerned about the Centers for Medicare & Medicaid Services' intention to deny reimbursement for the algorithmic analysis underlying multi-analyte tests through the payment pathway dedicated to laboratory services.

In a recently released document describing its preliminary stance on the reimbursement of so-called multi-analyte algorithm-based assays, or MAAAs, CMS notes that these types of tests don't fit the agency's traditional definition of diagnostics performed at laboratories, since they use a mathematic formula to analyze multiple markers associated with patients' risk for disease or their need for particular healthcare interventions. MAAAs are considered to be "boutique" tests since they are performed at a single laboratory, which owns the algorithm. After analyzing patients' biological samples using MAAAs, these labs report a composite score or result, and don't break out the presence of individual markers in test reports.

"Medicare does not recognize a calculated or algorithmically derived rate or result as a clinical laboratory test since the calculated or algorithmically derived rate or result alone does not indicate the presence or absence of a substance or organism in the body," CMS states in the document. "Medicare uses other codes for payment of the underlying clinical laboratory tests on which the MAAA is done and we continue to recommend not separately pricing the MAAAs codes."

The American Medical Association has assigned new Tier 1 and Tier 2 current procedural terminology codes – identifiers that describe medical services and that healthcare providers submit to payors for reimbursement – for approximately 100 genetic tests. Tier 1 lists CPT codes for commonly performed tests described by the specific analytes they gauge. Tier 2 lists codes for less commonly performed tests and describes them based on their complexity.

MAAAs, however, are not listed in the Tier 1/Tier 2 CPT code list. The AMA has said it will grant CPT codes for MAAAs that it has reviewed and considers analytically and clinically valid. For MAAAs that the association hasn't yet reviewed, the AMA has proposed to list them in an appendix that identifies the tests by their brand name. The MAAAs with CPT codes will not be identified by their proprietary name.

Meanwhile, CMS has been gathering stakeholder input on whether to price and reimburse these new molecular diagnostic test codes via the clinical lab fee schedule, the pathway through which most lab services are paid for, or the physician-fee schedule, the process by which doctors are compensated for their services. The lab industry is concerned that reimbursement will be negatively impacted if these tests are placed in the PFS, while physicians and pathologists believe that they won't be appropriately compensated for interpreting complex diagnostics if such tests are placed in the CLFS.

In its preliminary guidance, CMS is essentially expressing its view that the regulations that outline the types of lab services that can be reimbursed under the CLFS prohibit reimbursement of MAAAs under that pathway. CMS' view is troubling for industry observers who believe that in the next few years MAAAs, more so than analyte-specific tests, will dominate the diagnostics market as a way to personalized healthcare.

MAAA developers further hold that the proprietary algorithms that these tests use to yield a test result is what distinguishes them and increases their value to the healthcare system over other tests on the market. As such, CMS's position has test developers worried that the agency is planning to undercut the reimbursement for these assays, which often carry list prices of several thousand dollars. CMS did not respond to e-mailed questions for this article.

For example, Genomic Health's Oncotype DX, tagged with a $4,175 list price, gauges the expression of 21 genes and yields a recurrence score that places early-stage breast cancer patients at high, low, or intermediate risk for disease recurrence. Doctors can use this score to help them decide whether to treat a patient with chemotherapy. "Although we do not have a MAAA specific code for 2013, we are aware of the CMS determination and are assessing the potential implications both internally and with our partner trade groups," a Genomic Health spokesperson told PGx Reporter. "The role of bioinformatics has been well documented in peer-reviewed published studies, contributes to cost savings, and is a key component of personalized medicine that exists today in companion diagnostics and lab developed testing."

Oncotype DX is currently reimbursed by CMS for Medicare recipients and by most major private payors as a tool to assess breast cancer recurrence in women with early-stage disease that hasn't spread to the lymph nodes. Genomic Health, like the majority of diagnostics developers that currently market MAAAs, hasn't applied for a new CPT code from the AMA and has continued to garner reimbursement from CMS using a "miscellaneous" CPT code until the agency finalizes its stance on this subset of molecular tests.

"We've kept MammaPrint on the sidelines, while there was so much uncertainty about what CMS was going to do," said Wendy Wifler, senior director of governmental affairs at Agendia. In the short term, Wifler believes the company's multi-gene expression breast cancer recurrence test, which is reimbursed by the CPT code 84999 for "an unlisted chemistry procedure," will not be impacted by CMS's reimbursement plans for MAAAs. "But we're very interested in any precedent that might be set based on the decisions they make this year," she said. Agendia's MammaPrint test, listed at $4,200, analyzes the expression of 70 genes to determine whether patients are at high or low risk of breast cancer recurrence.

CMS hasn't yet finalized under which fee schedule it will place the 100 Tier 1/Tier 2 test codes or the few MAAA codes the AMA has issued. For example, Vermillion's OVA1, a test that gauges whether a woman's ovarian mass is malignant and requires surgery, falls into the MAAA category and received a new CPT code from the AMA earlier this year. Therefore, the most immediate impact from any CMS decision in this matter will likely be felt by the handful of MAAAs with CPT codes that go into effect in Jan. 1, 2013.

The agency will release its final payment determinations for both schedules in November. Stakeholders have 30 days to comment on CMS's initial decision.

Why Not PFS?

According to some experts closely tracking the evolving reimbursement landscape for molecular diagnostics, CMS's initial determination that the algorithmic portion of MAAAs doesn't fit into the CLFS signals an effort to push these types of pricey tests into the PFS, which could allow the agency to reduce the price of these tests over time.

"We believe that CMS is interested in placing the new codes on the PFS because of their ability to adjust valuation and pricing in the future – and ultimately control costs. CMS views MAAA as the most cutting-edge tests and believes that prices will come down in the future," Danielle Pambianco, director of health policy and reimbursement at the consultancy HillCo Health, told PGx Reporter over e-mail. "The CLFS is more limiting for CMS because once tests are priced, the pricing remains fairly stagnant, according to the way that fee schedule was set up." As such, a $4,000 MAAA reimbursed under CMS would likely maintain that price, as would the price of lower cost analyte-specific tests.

The PFS includes a professional component, which includes payments for mental and physical work directly provided by doctors, and a technical component, through which CMS pays for doctors' equipment, supplies, and clinical staff. If CMS decides to reimburse MAAAs via the PFS, labs would likely have to bill for these tests under the technical component.

Pambianco estimates that approximately 90 percent of the reimbursement for MAAAs under PFS would be paid for through the technical component. "Labs would also be able to bill the professional component if they had a pathologist on staff review the test," she noted, adding that the algorithm component of MAAAs may be accounted for as a practice expense under the PFS.

Unlike services billed under CLFS, however, lab tests billed under PFS would be subject to certain payment setting procedures, such as sustainable growth rate adjustments, the method by which CMS controls spending on Medicare beneficiaries.

Through SGR calculations each year, CMS assesses whether what it spends per Medicare beneficiary exceeds the growth in the gross domestic product. If Medicare beneficiary spending is more than the target amount for one year, then physicians' reimbursement levels face cuts the following year. In January 2013, based on the SGR formula, CMS is expected to slash doctors' pay rates by 27 percent.

However, over the last 10 years, Congress has helped physicians avoid deep SGR-driven reimbursement reductions through short-term fixes. Although physicians' groups and other stakeholders have floated various suggestions to Congress for fixing the payment system for doctors, including repealing the SGR process, no solution appears to have broad support among lawmakers. If CMS decides to reimburse molecular diagnostics through the PFS, it would place labs and test developers in the midst of this kerfuffle.

Additionally, most labs are concerned that placing molecular diagnostics in the PFS would underprice their tests because they'd be calculated by the schema used to determine reimbursement levels for doctors' services and costs. CMS calculates how much doctors should be reimbursed for their services and costs via resource-based relative value units, or RVUs, which are adjusted based on geographic location.

Finally, a 20 percent copay would be associated with lab tests reimbursed through PFS. "The … copay also worries labs since they do not have the collection procedures in place to chase the money," Pambianco noted.

A Unique Code

The latest changes in the medical claims coding and payment structure is due to the limitations of stacking or bundling existing CPT codes, which is the way healthcare providers currently bill payors for most molecular diagnostics. The AMA decided to develop new codes for molecular diagnostics and genetic tests at the urging of payors that felt the longstanding practice of stacking CPT codes obscured which procedures healthcare providers were performing and that this lack of clarity in payment policy contributed to unnecessary healthcare spending.

Instead of using stacked CPT codes, developers of MAAAs, such as Genomic Health, have had success getting reimbursement for tests through so-called "miscellaneous" or "unlisted" codes. However, for tests billed via such codes, companies must negotiate rates and claims with each payor, which is a time-consuming and expensive strategy.

The current reimbursement and coding system also causes wide variance in payment rates. In a report released in June, the Office of the Inspector General tracked the reimbursement of 16 genetic tests across state Medicaid and federal employee health benefits programs. OIG found that analysis of BRCA 2 gene mutations with Myriad Genetics' BRACAnalysis test garners $1,000 in reimbursement from Medicaid programs in Pennsylvania, but nearly $5,000 in Iowa. Meanwhile, Medicaid in Illinois reimburses around $2,000 for Oncotype DX, while in Minnesota the payment for the same test is around $3,400 (PGx Reporter 6/20/2012).

According to payors that want more consistency in payment policies and the establishment of a "value-based reimbursement system" for molecular diagnostics, this can most efficiently be achieved by assigning a unique claims code for every test developed by a specific test maker or lab. Such a coding system would allow payors to identify a particular test being administered, track its use in a population, and set reimbursement rates based on the impact it has on patient care. However, the AMA and some in the lab industry have balked at advancing such a system, fearing it will lead to price inflation for boutique tests.

"[S]ince codes are based on the market expense of inputs, a sole-source manufacturer who raised prices sharply would see 'his' proprietary CPT code rise as well, in a lockstep fashion, as his branded device became more expensive," Bruce Quinn, senior health policy specialist at the law firm Foley Hoag, wrote in a white paper last year. "Conversely, at least in principle, if the code is unbranded and workable equipment is multi-source, the provider would aim to buy the least expensive equipment compatible with the service provided, creating market competition rather than a monopoly among the corresponding suppliers."

On the flip side, a general code describing branded MAAAs could allow some labs to bill a higher reimbursement price for tests that aren't robustly validated. "The most common high-value tests (with the exception of complex cystic fibrosis testing) are branded tests, such as Oncotype DX, [Monogram Bioscience's] HIV diagnostic Trofile, and BRCAnalysis test," Quinn wrote. "Labs and payors might worry that if the descriptor for such tests was sufficiently general (e.g., 'prognostic test for breast cancer, 5 or more genes') then some labs might substitute very thinly validated tests for highly validated ones while receiving payment and coverage under the same CPT code."

At a personalized medicine conference held in Boston last week, Premal Shah, Genomic Health's director of business development, argued that the price of novel diagnostics should be high for tests that labs have built "from scratch" and the price should incrementally increase over time – not decrease – if the test developers have invested in conducting studies proving that the assay improves patients' health and saves the healthcare system money.

In a 2005 study sponsored by Genomic Health and published in the American Journal of Managed Care, researchers led by John Hornberger of Stanford University looked at how the use of Oncotype DX impacted survival and cost-effectiveness in a hypothetical cohort of 100 women with early-stage, estrogen receptor-positive breast cancer. Based on the assumption that patients who were deemed to be at intermediate or high risk of recurrence by Oncotype DX receive chemotherapy and patients at low risk don't, the researchers concluded that Genomic Health's test increased quality-adjusted survival by 8.6 years and reduced overall costs by $202,828.

While CMS is trying to figure out how best to integrate AMA's new molecular diagnostic CPT codes within its existing reimbursement framework, the agency is also running a program to evaluate whether the "one code, one test" model might be a better alternative. CMS has charged Medicare contractor Palmetto GBA to launch the MolDx program under which labs must submit data on the clinical validity and utility of their tests in order to receive Medicare coverage. Under the effort, each lab must also submit a diagnostic claim with a unique identifier to enable Palmetto to track its utilization. Based on the clinical evidentiary data submitted by labs and the utilization information gathered by Palmetto, the contractor will determine CMS's reimbursement policy and pricing for molecular diagnostics performed in six US states and territories (PGx Reporter 11/16/2011).

Although this program is still evolving, Palmetto Medical Director Elaine Jeter has indicated that payors would likely consider well-validated MAAAs of greater value than analyte-specific tests in terms of the impact they have on healthcare decisions (PGx Reporter 2/29/2012). "KRAS to me is an analyte. Is the mutation there? It provides a 'yes' or 'no' answer," said Jeter, speaking at a reimbursement conference earlier this year. "I know there was R&D [investment] put into that, but for the 25 to 50 laboratories in California that are going to be doing KRAS, that's [testing for] an analyte, which to me is no different than [a] glucose or … a [lactate dehydrogenase test]."

However, with multivariate algorithm-based diagnostics, "we're into a decision tree, where if Jane has this particular score or classification, then she will not have chemotherapy," Jeter said at the time. "That has an entirely different value, at least to us, in how we're trying to develop the reimbursement for these assays."

Palmetto is using healthcare technology firm McKesson's Diagnostics Exchange module to collect evidentiary data on tests and track the utilization of every molecular diagnostic with a unique identifier called a Z-Code. According Matthew Zubiller, VP of McKesson's decision management business, one way to identify MAAAs under the current coding and payment structure would be to add a Z-Code in addition to the appropriate CPT codes in each claim. This is how Palmetto is tracking utilization of molecular diagnostics under the MolDx program, and McKesson is working with private payors to implement a similar system.