By Bruce Quinn
Decade by decade, healthcare undergoes big changes. But we can still make enormous improvements in innovations for unmet needs, more effective healthcare, better coordinated services, and cost-effectiveness. In particular, this is true in the field of precision medicine, where genomic and proteomic sciences have made rapid progress. Bringing these technological advances into healthcare devices and services is challenging, exciting, difficult, and important.
Most frequently, when we talk about healthcare investment and innovation, we focus on the FDA. This usually reviews the evergreen concerns that the FDA is either too strict, and keeping innovations from patients, or too lenient, letting unvalidated drugs or devices slip through the cracks and into the marketplace. Both the focus and the tension was captured in the title of a 2011 Congressional hearing – “Impact of Medical Device Regulation on Jobs and Patients,” 1 which was stimulated in part by an insightful study of device innovation by Josh Makower MD 2, who now leads the Byers Center for Biodesign at Stanford University.
Less discussed is the critical role that Medicare policies have on device and diagnostics innovation. Yes, Medicare pays for healthcare technologies, and CMS makes coverage decisions which, like FDA decisions, may be viewed as too lax by some and too burdensome by others. In this review, we argue that the role of CMS policies in accelerating valuable innovation is far more important than is usually appreciated, and spans a range of national and even local Medicare policy rules that install unnecessary roadblocks. Let’s look more closely at some of these policy rules, praise the good ones, and suggest changes for rules that are dysfunctional.
National coverage decisions – The value of open-ended promises
Medicare makes national coverage determinations (NCDs) via a nine-month national comment process, run by a small number of physicians and other professionals in the agency’s Coverage and Analysis Group.
These policymakers have written several recent national coverage decisions that have important and innovative ramifications. In 2017, CMS announced it would cover tumor genome diagnostics with next generation sequencing for on-label indications if they were FDA-approved. It wasn’t just about testing; CMS wanted to encourage developers to enter the FDA review pathway because of the guaranteed Medicare coverage at the end of it. In a 2019 extension, CMS added coverage for hereditary cancer tests as long as they were similarly FDA cleared or approved.
In a third decision, in 2021, CMS created an innovative coverage category, promising to cover any liquid biopsy tests that screen for colorectal cancer (in place of stool tests or colonoscopy) as long as they were 74% sensitive (catching 3 cancers in 4) and 90% specific (less than 10% false positives). CMS ratified this coverage position even though no such tests had yet been approved by the FDA.
These decisions are very important. Together, in three major areas of genomics Medicare provides coverage for broad categories of important tests, each predicated on FDA review and approval, but providing first-day coverage from Medicare when that happens. Each policy is self-implementing on a rolling basis, as the FDA approvals occur. Interestingly, all three of these innovative decisions, which had no precedent in precision medicine, occurred during the Trump administration (2017-2020). The timing of these innovations may have reflected a loosened resistance to policy change and “the way we’ve done things” during that interval.
Local coverage decisions – Foundational LCDs define coverage rules for future tests
Similar remarkable innovations have been rolled out in the last several years at the local level as well, where Medicare contractors work. Only about 60% of Medicare beneficiaries are still in the slowly shrinking fee-for-service program, but its policies also apply to patients in Medicare Advantage and Medicare Direct Contracting programs (now called REACH ACOs).
Some 80% of genomic tests in the Medicare program are covered by contractors in a special carve-out called the MolDx program, which has a dedicated staff of molecular pathologists in charge of making uniform coverage decisions for about 30 states.
Since 2018, the vision of the MolDx program has been to broaden umbrella coverage statements, called “foundational local coverage determinations,” that define areas of coverage as “use cases” for genomic tests. The most dramatic is an LCD finalized in November, 2021, which provides coverage for DNA-based minimal residual disease monitoring for two use cases – tumor relapse aftercurative surgery, and treatment monitoring. To achieve coverage, new tests just need to meet or exceed performance criteria laid out in the LCDs, and each test passes a close inspection (“technological assessment”) by MolDx professionals. The index test was the Natera Signatera test when used to manage colorectal cancer, but the LCD is open-ended enough to allow coverage of future tests in lung cancer, breast cancer, and other conditions.
That’s not all. MolDx has also sped up the review process for new genomic tests, and reduced uncertainty, by writing and posting public templates (spreadsheets up to 20-30 lines long) explaining exactly how clinical utility and validation quality will be evaluated. In this approach, MolDx policymakers were ahead of the FDA, which only recently began using clear pre-defined templates for diagnostic test submissions during the COVID crisis (as an addition to the Emergency Use Authorization rules). FDA has remarked it finds the template approach so successful it may be used for non-EUA submissions in the near future. 3
It’s not all good news – The demise of “MCIT”
In September 2020, CMS announced a major new regulation called “MCIT” – Medicare Coverage for Innovative Technology. After favorable public comments, CMS finalized the regulation in January 2021. The theme was simple: if a device received the FDA Breakthrough Device Designation – a process established in the 21st Century Cures bill – and goes on to successfully complete its clinical trials and FDA review, CMS would guarantee four years of provisional coverage for on-label applications. Unfortunately, under new leadership, CMS put this policy on hold in Spring 2021 and canceled it by the end of the year. Fortunately, as of Spring 2022, CMS is holding town halls on other ways it might help innovation.
MCIT was smarter than its detractors realized, as later noted by policymakers like Scott Gottlieb MD and Joe Grogan, a physician and an attorney, respectively, who worked at high levels in the 2017-2020 administration. The goal of MCIT was to boost investment in medical products that were highly promising, original, and important, as defined by others – by Congress and the FDA. The goal was to reduce the “valley of death” between early trials and full funding for clinical development, which is caused by two kinds of risk – science risk (will it succeed and pass FDA) and payor risk (will it meet uncertain payor criteria or suffer slow payor timelines). MCIT nixed the payor risk.
I think objections to MCIT were overestimated. For example, if a product had too much science risk, it wouldn’t pass FDA muster, and might not even try for a final review. If the product was so-so, it’s unlikely there would have been much clinical adoption, so CMS payments would have been few. In short, some of the problems proposed by opponents were self-correcting. In a simple model, using reasonable timelines, success rates, and discount rates, I found that MCIT (coupled with earned Breakthrough Status), raised the economic value of a Phase I product idea by 2.5X (such as from $15M to $36M). In contrast, another CMS policy trick, Coverage with Evidence Development, destroyed value and slashed the product net present value by half, from $15M to only $8M. Where the rubber meets the road is when you add a new factor, how much the clinical trials will cost. If they cost $10M, and the product value under CED is $8M, you won’t invest, and the product dies. But under MCIT, the exact same product, with the same odds of success, is worth $36M, so the $10M in development funds will quickly be invested. (See how the math works in an online article. 4
Two unnecessary headaches: Glacially slow local decisions and counterproductive bundling policies
There are anti-innovative policies at CMS that could be corrected fairly easily. For one thing, the timeline for local coverage determinations – LCDs – has gotten far too slow, with proposals waiting in limbo for a year or more, and the timeline for executing a new LCD taking a year from first publication to finalization. This comes largely from a CMS decision that even expansions in coverage – such as adding a new covered service to a relatively minor LCD – should go through a lengthy public comment cycle in advance. In a quick Medicare fix, Congress could add a few words to the statute and solve this recently-created problem. The slow timelines for LCD modifications has led some contractors to issue LCDs of maddening vagueness, such an LCD that covers microbiology genomic panels when “the test is timely and will improve care.” That’s too vague to provide a safe harbor for test utilization, and a lightning rod for disputes in recoupment cases where for-profit reviewers, called recovery audit contractors, will take the provider’s money today and essentially dare him to try and get it back in court years from now. In the hospital outpatient setting, where much specialty care like cancer care is delivered, CMS has been overzealous in applying bundling rules to lab tests. Beginning in 2014, CMS has bundled nearly all lab tests to patient visits and procedures, such as hospital biopsies. CMS makes an exception for tests of human DNA and RNA – but where does that leave advances in proteomics, metabolomics, molecular pathogens, and tests that rely on artificial intelligence. 5 Those are all bundled to the hospital visit, prostate biopsy, or other procedure. Soon we’ll have accurate proteomic tests for Alzheimer’s disease, but if they are paired to a hospital neurology clinic visit or spinal tap, the value of the test will be “bundled” inside the $100 office visit or the procedure payment. And soon we’ll have organoids grown on biopsy tissue which predict chemotherapy responses in cancer patients, but under current policy, testing costs would be bundled if they’re ready in 7-10 days rather than 14-30 – a highly counterproductive policy. 6 Bundling of the most innovative categories of precision medicine diagnostics potentially risks impeding innovation in these areas, and investors learn to navigate away from these areas where old rules are toxic, even though the unmet clinical need is high.
Awareness will support change
For one thing, we need to bring some of these findings out of the dark basement of the reimbursement policy world, and into the daylight. When CMS makes smart decisions that encourage innovation and improvement, at reasonable or even lower costs, we should celebrate it and look for more opportunities like that. When CMS does stuff that’s just dysfunctional and counterproductive, we shouldn’t let them get away with it, and we should provide constructive feedback for fixes. Also in the prior administration, CMS established a new department, a “Technology Coding and Pricing Group,” designed to be a forum for hearing about problems and taking steps forward. We should all recognize how important that kind of department and leadership is at CMS, and support them wherever we can.
References
1. Subcommittee on Health, House Committee on Energy and Commerce , February 17, 2011.
2. Makower, J., Meer, A., & Denend, L. (2010) FDA Impact on U.S. Medical Technology Innovation. 44pp.
3. Herr, T., (2021) US FDA publishes EUA templates for COVID-19 IVD submissions.
4. Quinn, B. (2022) Revisiting MCIT with Numbers. MCIT Could Triple the Value of
Medtech Investments. Discoveries in Health Policy (blog).
5. Goldfarb, A., & Teodoris, F. (2022) Why is AI adoption in health care lagging? Brookings Institute.
6. Letai, A. Functional precision cancer medicine – moving beyond pure genomics. Nature Medicine 23:1028-35.
Bruce Quinn, a physician-MBA, is a full time consultant for companies bringing innovative new technologies towards Medicare coverage. He holds an MD-PhD from Stanford University and an MBA from the Kellogg program at Northwestern University. After working in academic medicine, he’s been a strategy consultant for Accenture, two health policy firms in Washington DC, and for his own consulting practice based in Los Angeles and San Francisco.
