Included in the original law that established Medicare in 1965 was a short but consequential mandate that the program only pay for items and services that are “reasonable and necessary” for the treatment of illness and injury. Over the decades that followed, this requirement was generally viewed as excluding from coverage those medical services and products that were experimental or investigational, or for which there was still uncertainty about the benefits and harms for Medicare patients.
As a result, many new products, though approved by the FDA, would not meet Medicare’s “reasonable and necessary” standard. In some cases, this led to long delays in Medicare patient access to promising, emerging technologies.
In an effort to comply with its statutory mandate while avoiding delayed access to promising technologies, Medicare in the mid-2000s unveiled a “coverage with evidence development” (CED) program, which provided beneficiaries early access to promising medical technologies while it investigated gaps in evidence of effectiveness and safety. Since then, Medicare has applied the program in roughly two dozen cases, mostly for medical devices and procedures.
For example, it made transcatheter aortic value replacement available to patients for whom studies suggested a likely benefit, providing more rapid access than traditional evidence requirements for national Medicare coverage would have permitted. The CED program mandated additional data collection about the procedure’s real-world safety and effectiveness in a national clinical registry, which provided useful evidence for Medicare officials as they reconsidered whether to pay for this technology.
However, data collection through CED-approved clinical trials and registries has often proven costly and prolonged, and yielded evidence of variable quality and relevance. As a result, only a handful of technologies under CED have been converted to full coverage or had their coverage restricted or withdrawn. Furthermore, the CED policy, which debuted in an era of BlackBerrys and DVDs, has not been substantially updated, despite advances in data collection and analysis. CED’s implementation has proven challenging. The program is long overdue for a substantial overhaul.
In August the Centers for Medicare and Medicaid Services unveiled a modification of the CED program, called transitional coverage of emerging technologies (TCET). The new framework includes some procedural changes that may shorten timelines for Medicare decision-making, allow for more interaction between medical device developers and Medicare officials, and potentially enable more efficient evidence generation approaches. The new policy helpfully encourages “fit-for-purpose” study designs, including those that leverage real-world data. But the impact of these changes will likely be minimal for Medicare beneficiary access to emerging technologies, and the life sciences ecosystem that develops them.
Meaningful improvement in the CED program’s goals will require more fundamental reforms, both within Medicare’s authority and beyond CMS’s jurisdiction. A truly extreme makeover would embed CED into a broader and more ambitious vision for expanding the nation’s capacity for post-approval evidence generation.
TCET takes important steps in the right direction. It will allow medical device companies to coordinate with CMS prior to clearance or approval by the FDA for selected FDA-designated “breakthrough” devices (a label extended to promising devices targeting life-threatening or serious conditions). Manufacturers will provide an evidence development plan to address critical evidence gaps. Once CMS finalizes and approves the plan, the CED mechanism provides a transitional coverage period of at least five years, after which CMS would conduct an updated evidence review and reconsider its coverage decision.
Still, because TCET is built on the existing CED chassis and can be deployed only through Medicare’s national coverage authority, its impact will be limited. Importantly, TCET will apply to just a few breakthrough devices each year, while excluding drugs and almost all diagnostics.
It will also omit the large number of medical devices that Medicare pays for through existing codes, bundled payments, and local coverage decisions, which do not involve national coverage review. For these devices, manufacturers have little incentive to conduct studies to address uncertainties about the real-world benefits and risks of their products.
Furthermore, TCET studies are designed as “one-offs.” That is, they are not intended to coordinate with FDA post-approval study requirements or other industry-funded post-market studies, or consider the potential to leverage shared data collection infrastructure. For example, the recent CED for monoclonal antibodies for Alzheimer’s disease relies on multiple independent registries for data collection, with different study protocols and endpoints, and no plans for data aggregation across studies.
The need for improvement in post-approval data collection is not unique to Medicare. Evidence generation challenges for new technologies exist across payers and treatment settings, reflecting the fragmented nature of U.S. post-approval data collection efforts, which lack the coordination and shared infrastructure to create scale and efficiency. Dozens of drugs, devices, and diagnostics are approved or cleared by the FDA each year even if their benefits and harms in real-world use, in patient subgroups, and over time are unclear. Medicare’s ability to address critical evidence gaps is limited to the handful of medical technologies that are reviewed through the national coverage process.
Medicare should broaden the TCET program’s scope and ambitions. In practice, this could allow for new studies examining how emerging treatments for diabetes, heart failure, or Alzheimer’s disease affect the health and quality of life of Medicare beneficiaries in real-world settings.
Such efforts would require additional resources for CMS. The Bipartisan Policy Commission has recommended up to $10 million per year to modernize the Medicare coverage process — a modest figure in light of Medicare’s trillion-dollar annual outlays, and an order of magnitude too small to make any meaningful difference. For comparison, the budget for the FDA center that reviews medical devices was $748 million in 2023. The combined budget for the centers that review drugs and biologics is substantially higher.
More resources could support additional CMS staff with methodological and clinical expertise, as well as more frequent meetings of the Medicare Evidence Development and Coverage Advisory Committee to inform the development and implementation of CED programs. Dedicated funds could also strengthen CED data collection infrastructure and help CMS coordinate with other government agencies, such as the National Institutes of Health.
Legislation to clarify CMS’s authority to implement CED would also help. While courts have generally supported CMS discretion over coverage policy, CED’s legal underpinnings need strengthening, especially given the Supreme Court’s recent decision to overturn the “Chevron doctrine,” ending the longstanding practice of courts deferring to federal agencies’ interpretations of their statutory authorities.
CMS would further benefit from new statutory language clarifying the “reasonable and necessary” coverage standard, including explicit language related to evidence requirements for emerging technologies, and allowing Medicare to consider a technology’s cost or cost-effectiveness.
Congress should also explore policies that link evidence quality to reimbursement levels. Capping payment for emerging technologies until confirmatory evidence is provided, for example, would create stronger incentives for product developers to conduct studies rapidly to address critical uncertainties.
An enhanced CED program could also serve as one element of a more ambitious vision to expand post-approval evidence generation in the U.S. A truly extreme makeover for CED would leverage the program in support of a national evidence generation ecosystem. CED would have more impact if it were formally linked to efforts by other federal agencies to advance evidence generation. Medicare is the only payer in the U.S. with the size and power to drive this change.
A broader effort could enable the conduct of pragmatic randomized trials that accrue patients quickly, as well as support registries in selected contexts to capture data on large numbers of patient in real-world settings. The data could help address national priorities for which private industry lacks sufficient incentives to gather evidence. Private payers and health systems could also participate in clinical studies on FDA-approved products to augment the evidence base on key safety and effectiveness questions for drugs and devices. Implementing this broader vision will require champions, substantial investment, new legislation and answers to many technical questions. STAT’s own Matt Herper has observed that our country’s focus on scientific advances in health care has outpaced our ability to assess them. “We try to solve problems by building faster and more expensive Ferraris,” he wrote in 2022, “when what we really need are better roads.” The TCET program plugs up some of the potholes in the roadway, but the country’s evidence generation efforts need a state-of-the-art interstate highway system.
Peter J. Neumann is director of the Center for the Evaluation of Value and Risk (CEVR) at Tufts Medical Center. Elliott Crummer is a research assistant at CEVR. Anna Kaltenboeck is a principal and head of prescription drug reimbursement practice at ATI Advisory. Sean R. Tunis is a senior fellow at CEVR. This opinion piece is informed by work performed under a grant from Arnold Ventures. The views expressed are those of the authors, and do not necessarily reflect the views of Arnold Ventures.
