By Turna Ray
Through a new pilot program with health insurers and collaborations with diagnostic developers, CVS Caremark's genetic testing benefits manager Generation Health is trying to align the goals and expectations of two industry sectors that have thus far been unable to find common ground when it comes to the adoption of genetic tests.
Generation Health is piloting a program with five health plans in which the GBM deciphers stacked CPT codes that insurers use to reimburse for molecular diagnostics and recommends coverage policies based on evidence showing these tests are clinically useful and economically beneficial. Simultaneously, Generation Health is in discussions with several diagnostics firms to assist them in generating the types of clinical validity and utility evidence that payors need to see in order to justify reimbursement for genetic tests.
The CPT coding system for molecular diagnostics is in flux as the American Medical Association develops a new coding structure to address the complexity of emerging genomic technologies. The new coding system proposed by the AMA is slated to go into effect in 2012, at which time the longstanding practice of CPT code stacking will be phased out (PGx Reporter 03/16/2011).
Generation Health's CPT coding project will remain in the pilot phase for the time being so the GBM can accommodate insurers' changing needs as the AMA's coding structure takes hold.
"We're anticipating that this will ultimately become a service," Rick Schatzberg, CEO of Generation Health, told PGx Reporter. While stacked CPT coding could be phased out in the coming years, "that won't make our plans for this service go away," he said. "It will just make the heavy lifting part of it — the deciphering of the codes — easier. We'll just spend less time on deciphering code [stacks] and spend more time on the recommendations aspect of it."
Doing the 'Heavy Lifting'
Unraveling stacked CPT codes is a time-consuming task requiring manual curation of hard-to-find information and good relationships with diagnostic companies. While most insurers would agree that stacked CPT codes don't provide the level of clarity they need to track their payment trends, test providers are understandably nervous when it comes to divulging the way they are getting paid.
Deciphering the codes also requires extensive knowledge of how test developers use CPT codes, not just in terms of lab claims but also by molecular diagnostics firms. Generation Health employees had to manually slog through molecular diagnostics firms' websites to take stock of the types of tests the companies are selling and then try to match them up with the CPT codes that are used to bill health plans.
Making Generation Health's job more difficult is the fact that diagnostics firms often stack different sets of codes to bill different plans for the same tests. Another unknown in the process is the number of units of the test per stack that is being billed for. "We use the term 'fingerprint' internally when you want to identify code stacks instead of type them. That may be futile in a sense — that once you identify them, to expect that the same test will be represented by the same set of codes," Schatzberg explained.
Also, "you need to understand what variation is in [the stacked codes] in terms of the number of units that differentiate which test is which," he said.
On the other side of that coin, some labs use the same handful of codes to describe hundreds of tests in their portfolio. In these cases, "there is no way you can decipher [the codes] without reaching out to the lab and having them tell you," Schatzberg said. "If it's a lab with one test, it's not so hard to figure out [the CPT codes]. But if it's a lab that was just bought by Quest or LabCorp, it's damn hard to figure that out … Then those tests are just not visible any longer."
In order for this project to work, Generation Health has been contacting molecular diagnostics firms to inform them that it is researching the way they bill insurance companies and asking for their help in this task. The GBM is informing these firms that it has the permission of the five plans in the pilot study. "If we did it just as Generation Health, I don't know why they would" provide this information, Schatzberg acknowledged. "But if they were notified by [a] health plan, that we're their agent in this and that we may be contacting them to ask them [for this data], I expect they'd give it to us."
[ pagebreak ]
Before launching the CPT coding program, Generation Health surveyed the top 20 health plans in the US, representing 178 million lives, to figure out the types of policies they had around genetic tests. Among this subset, Generation Health found a wide range — some plans had a few as four genetic testing policies while other plans had more than 120 policies. Then Generation Health assessed whether these policies were being appropriately enforced and identified gaps.
When Generation Health is ready to roll out its reimbursement services program more broadly, the GBM is hoping that this survey of the genetic testing landscape will help it net new customers.
"When we talk to a health plan and see that they have reimbursed a test that's really questionable in terms of its clinical utility … and if we can show them, for example, that among the five health plans in the pilot, for 38 percent of the 178 million [covered individuals], they just flat out exclude that test, that's helpful for the plan we're speaking to to know that," Schatzberg said.
"It doesn't mean that the best recommendation is to just exclude the test," he added. "We might suggest a more nuanced [policy], and recommend that they should cover it under certain circumstances. Or, if we think they should exclude it until there is more clinical data, and they decide to do that, then they're in good company."
For the five insurers in the pilot, Generation Health is conducting critical R&D that shines a light on the types of policies they need to develop as more drug/diagnostic combination products enter the market and more doctors incorporate genetic testing in the care of their patients. Specifically, the information coming out of Generation Health's pilot can help health plans prioritize which tests to reimburse, the types of prior authorization schemes to promulgate, and the kinds of test notification services they need to build.
"It seemed like if we're really getting our hands dirty, truly understanding what's being billed for and reimbursed at a level that no one seems to know at this point, then in collaboration with these clients we can really develop a product they want to buy, rather than [Generation Health] just building something on its own," Schatzberg said.
When considering the type of service offering to build around reimbursement of genetic tests, Schatzberg said that Generation Health considered building an entirely new infrastructure for coding and identifying tests, along the lines of what healthcare information technology provider McKesson has done. The company's InterQual decision-support system for molecular diagnostics helps payors and physicians access the available scientific literature and insurers' coverage criteria for specific genomic tests. As a workaround for the inadequacies of the existing CPT coding system, McKesson created a catalogue that gives a unique identifier to each test (PGx Reporter 11/18/2009).
Building a system such as McKesson's "felt a little bit like 'build it and they'll come' and it does require some real investment to do it and I'm not sure that all the health plans want to go that way," Schatzberg said. "Maybe McKesson will really strike it rich with that approach, but I felt that given our size and resource level, we were better off doing a consulting service on the medical genomics side, and then build a finely tuned niche product based on what [insurers] want."
Furthermore, even if the AMA's new coding structure reduces demand for deciphering codes, insurers will likely need more help sifting through clinical evidence on tests as whole-genomic sequencing technologies brings forth an avalanche of new data on gene-drug associations and disease predisposition markers.
"When whole-genome sequencing comes [into clinical practice], just imagine what types of recommendations health plans are going to need, because now interpretation becomes the big problem," Schatzberg said. "And it all comes back to the thing that everyone talks about: utility … For that discussion to have medical geneticists and real experts talk to plans makes sense. And that's what we do."
Generation Health sees its job as a lot more than just a developer of payment policies for the insurance industry. Rather, the GBM views itself has an intermediary between insurers and test makers, guiding the adoption of genomically guided medicine where it saves healthcare dollars and improves the quality of care. To help generate evidence of this kind, Generation Health has identified "a rich pipeline of opportunities" with diagnostic developers to conduct clinical validity and utility studies.
To an extent, Medco has a similar program to conduct clinical utility studies with test makers. Medco competes with CVS Caremark in the pharmacy-benefits management space, and both companies have launched personalized medicine programs to use genomically guided strategies to make healthcare more efficient and incur cost savings for their payor customers (PGx Reporter 05/26/2010). CVS Caremark holds a majority stake in Generation Health.
"CVS Caremark has payors that may find many of these tests interesting should they prove to have utility," Schatzberg said. "When you do that you are by definition getting [CVS Caremark's payor members] to buy in on the concept that if the test proves useful that could be helpful to them." Since payors will understand the trial designs and may even be able to specify the outcomes they want to see out of the study, "if the trial ends up meeting the endpoints and showing the utility, then in a sense you've really gotten far down in discussions with them about reimbursing for that and that's appealing for the developers of diagnostics," he noted.
In talking to test developers and reviewing their products, Generation Health has encountered "a fair number" of tests that have strong evidence for validity and utility but that aren't being used in healthcare due to lack of awareness. There are also many tests that don't have the necessary real-world clinical utility data, but they look "promising," according to Schatzberg.
Although most of its discussions with test developers are in the beginning stages, Generation Health believes this kind of evidence generation work validates its role in the personalized medicine field as a "neutral arbiter."
"Our role isn't to promote one test or another or one technology or other … What the data says is what the data says," Schatzberg said.
"But, if we've seen the data, then why wouldn't we promote that test?" he posited. "Of course, there are winners and losers in that. We do want to see an uptake of personalized medicine, but not in costs."
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.