Coverage with Evidence Development (CED) is a Medicare coverage mechanism the Centers for Medicare & Medicaid Services (CMS) uses inside a National Coverage Determination (NCD) to pay for an item or service that shows promise for Medicare beneficiaries but does not yet have enough clinical evidence to support unconditional “reasonable and necessary” coverage under Section 1862(a)(1)(A) of the Social Security Act. Instead of an outright yes-or-no coverage decision, CMS conditions payment on the patient’s care generating additional evidence — through a registry or a clinical study — that helps close the evidence gap while beneficiaries still get access to the item or service.
For research administrators and clinical trial offices, CED matters because it can turn Medicare coverage itself into a data-collection obligation: a site providing a CED item may need to enroll the patient in a CMS-approved registry, confirm participation in a qualifying study, or bill in a way that reflects the CED requirement, on top of the usual Medicare billing rules.
What CED Actually Requires
CMS describes CED as an evidence-based coverage paradigm for items whose evidence base is not yet sufficiently developed to support standard coverage, but where CMS believes the evidence gap can be closed by conditioning coverage on additional data collection tied to the patient’s own care. A CED determination is issued through the same National Coverage Determination process that produces any other NCD — CED is a coverage condition attached to an NCD, not a separate approval track that bypasses it. See CASRAI’s guide to Medicare Coverage Determinations: NCD vs LCD for how that underlying process works.
The Two CED Categories
CMS’s own CED guidance describes two categories of data-collection requirement, and a given NCD’s CED language specifies which one applies:
- Coverage with Appropriateness Determination. Providers must submit clinical data to a CMS-approved data registry documenting how the item or service was used. This is the lighter-weight of the two mechanisms: beneficiaries receive the item or service in the course of routine care, and the registry captures real-world outcomes data rather than requiring a formal trial protocol.
- Coverage with Study Participation. The beneficiary must be enrolled in a CMS-approved prospective clinical study that is designed to generate the specific evidence CMS identified as missing. This is the more demanding mechanism: it requires an actual study protocol, IRB oversight, and formal enrollment, not just registry data submission.
Some NCDs specify a registry, some specify study participation, and some (historically, transcatheter aortic valve replacement is one example) have used a combination as the evidence base matured over successive NCD reconsiderations.
Real Examples of CED in Practice
CED determinations are relatively rare compared to the full volume of NCDs CMS issues, but they cover some high-visibility technologies. Two frequently cited examples:
- Transcatheter Aortic Valve Replacement (TAVR). CMS’s first TAVR NCD, issued in 2012, established a CED requirement tied to a national registry (the Society of Thoracic Surgeons/American College of Cardiology TVT Registry) as the technology moved from a narrow, high-risk-only indication toward broader use. Subsequent TAVR NCD reconsiderations have progressively relaxed CED and site-requirement conditions as the evidence base matured.
- Amyloid PET imaging for Alzheimer’s and other dementias. CMS’s 2013 NCD conditioned coverage of amyloid PET scans on the beneficiary’s participation in an approved clinical study, reflecting the evidentiary uncertainty at the time about whether the scan actually changed patient management or outcomes. CMS retired that NCD and ended the CED requirement for amyloid PET in October 2023 as national non-coverage determination policy was reconsidered and evidence, and clinical practice, evolved.
The amyloid PET example illustrates the intended lifecycle of a CED determination: it is meant to be a bridge, not a permanent state. As the registry or study data accumulate, CMS can reconsider the NCD and either lift the CED condition, extend it, or, in principle, restrict or end coverage if the evidence turns out unfavorable.
How Often CED Is Used, and How Long It Lasts
CED determinations are uncommon relative to the total NCD workload — independent published tallies place the number of CED-based NCDs issued between 2005 and 2022 at roughly two dozen across a handful of therapeutic areas, with individual CED programs running anywhere from a year or two to well over a decade before CMS revisits them. CMS updated its formal CED guidance document in August 2024, the current reference point for how CMS staff apply the paradigm when drafting a new NCD.
What CED Means Operationally for a Research Administration Office
If your institution provides a CED-covered item or service, the operational obligations typically include:
- Confirming which CED category applies by reading the specific NCD’s CED language — registry submission and study enrollment carry very different administrative burdens, and the two are not interchangeable.
- Verifying the registry or study is the one CMS actually approved for that NCD. CMS maintains and periodically updates lists of qualifying registries and studies per technology; using an unapproved registry does not satisfy the coverage condition.
- Coordinating between the clinical service line and the research/coverage-analysis function, since CED sits at the boundary between routine clinical billing and clinical-trial-style data collection — exactly the kind of cross-functional coordination a Medicare coverage analysis is built to catch. See CASRAI’s Medicare Coverage Analysis for Clinical Trials guide for that broader process.
- Documenting evidence of registry submission or study enrollment in the billing record, since CMS Administrative Contractors can and do request documentation that the CED condition was actually met, not just asserted.
CED vs. the Other Medicare Coverage Mechanisms Research Administrators Encounter
CED is one of several distinct mechanisms CMS uses around clinical research and emerging technology, and they are frequently confused with one another:
- CED vs. NCD 310.1. NCD 310.1 governs whether Medicare covers the routine patient-care costs of a qualifying clinical trial (drugs, devices, services a beneficiary would need regardless of trial participation). CED is a different, narrower mechanism: it conditions coverage of one specific item or service on evidence-generating enrollment, independent of whether the broader encounter is billed as a qualifying clinical trial under 310.1. A CED-covered item can appear inside or outside a 310.1 qualifying trial.
- CED vs. Category A/B IDE device coverage. The Category A vs. Category B IDE framework (42 CFR 405.201-405.215) governs whether Medicare will pay for an investigational device itself during an FDA-regulated device study. CED is a post-market (or post-NCD) evidence-development condition attached to a specific NCD, not the IDE device-payment framework.
- CED vs. the RAPID Coverage Pathway. CMS and FDA’s RAPID Coverage Pathway, announced in April 2026, is a faster on-ramp to a national coverage determination for a narrow slice of FDA Breakthrough Devices. CED is a separate, unaffected mechanism: RAPID changes how quickly an NCD can be proposed and finalized, not whether that resulting NCD carries a CED condition.
- CED vs. TCET. Transitional Coverage for Emerging Technologies (TCET), finalized in 2024, is a distinct pathway offering time-limited national coverage for certain FDA Breakthrough Devices while further evidence accumulates — conceptually related to CED’s evidence-bridging purpose, but a separate, formally defined program with its own eligibility and timeline rules rather than a CED determination itself.
Frequently Asked Questions
What is Coverage with Evidence Development in Medicare?
It is a mechanism CMS uses within a National Coverage Determination to pay for an item or service with promising but not-yet-sufficient evidence, by conditioning coverage on the patient’s care contributing data — through a CMS-approved registry or clinical study — toward closing that evidence gap.
Is CED still an active CMS policy?
Yes. CMS issued an updated CED guidance document in August 2024, and CED conditions remain attached to a number of current NCDs, even though individual determinations (such as the 2013 amyloid PET NCD) are periodically retired once CMS decides the evidence question has been resolved.
What is the difference between Coverage with Appropriateness Determination and Coverage with Study Participation?
Coverage with Appropriateness Determination requires providers to submit clinical outcomes data to an approved registry as part of routine care. Coverage with Study Participation is more demanding: it requires the beneficiary to be formally enrolled in a CMS-approved prospective clinical study built to answer the specific evidence question CMS identified.
Does every Medicare NCD include a CED requirement?
No. Most NCDs are unconditional coverage or non-coverage decisions. CED is reserved for the comparatively small number of items where CMS judges the evidence promising but insufficient, and where a defined data-collection mechanism can plausibly close that gap within a reasonable timeframe.
How is CED different from a clinical trial’s Medicare routine-costs coverage under NCD 310.1?
NCD 310.1 addresses whether Medicare will pay the routine patient-care costs a beneficiary incurs by participating in a qualifying clinical trial. CED addresses whether Medicare will pay for a specific item or service at all, conditioned on evidence-generating enrollment. The two can apply to the same encounter but answer different coverage questions.
