Investigational Device Exemption

Investigational Device Exemption in Device Trials and Medicare Coverage

An investigational device exemption (IDE) is the permission, granted under 21 CFR Part 812 in the United States, that lets a medical device without marketing approval be shipped and used in a clinical study. The study it permits is usually the source of the safety and effectiveness data that the FDA later reviews in a premarket approval (PMA) application. For health economists it matters because it shapes how much evidence exists when payers and health technology assessment bodies first meet a device, and because the FDA category given to the device shapes whether Medicare pays for the device and routine care in the study.

What the exemption permits and why it exists

The regulation states its purpose as encouraging the discovery and development of useful devices, to the extent consistent with public health, safety and ethical standards. An approved IDE allows a device that would otherwise need premarket approval or compliance with a performance standard to be shipped lawfully for an investigation. It exempts the device from a list of duties that apply to marketed devices, including registration, listing, premarket notification, premarket approval and most quality system requirements.

The FDA explains that clinical studies of devices are most often run to support a PMA, while only a small share of premarket notification [510(k)] submissions need clinical data. Unless a study is exempt, it must have an approved IDE before it starts.

Significant risk, non-significant risk and exempt studies

FDA guidance describes three types of device study under the regulation, which are significant risk, non-significant risk and exempt studies. The type sets how much regulatory work comes before the first patient is enrolled, and so how quickly evidence can start to accumulate.

A significant risk device is one that presents a potential for serious risk to the health, safety or welfare of a subject, for example an implant, a life-supporting device or one of substantial importance in diagnosing or treating disease. For these studies the sponsor must submit an IDE application to the FDA, and both the FDA and an institutional review board (IRB) must approve before the investigation begins. An investigation may begin 30 days after the FDA receives the application unless the FDA tells the sponsor it may not, or once the FDA approves the IDE by order.

A non-significant risk study is any study that does not meet that definition. It follows abbreviated requirements, including IRB approval, informed consent and monitoring, and is treated as having an approved IDE. The sponsor makes the first risk determination, the IRB reviews it, and the FDA is the final arbiter. FDA guidance notes that such a study can start as soon as the IRB approves it. Exempt investigations include some non-invasive diagnostic devices and custom devices.

One ground on which the FDA may refuse or withdraw approval is that the risks to subjects are not outweighed by the expected benefits to them and the importance of the knowledge to be gained. The test weighs harm to participants against the social value of information.

Charging, early access and who pays for device trials

Part 812 limits how far a sponsor can recover the cost of a device trial before the device is marketed. A sponsor or investigator may not promote or test market an investigational device before FDA approval for commercial distribution. Nor may it charge subjects or investigators more than the price needed to recover the costs of manufacture, research, development and handling.

A treatment IDE allows a device under study to be used outside the trial for patients with a serious or immediately life-threatening disease and no comparable or satisfactory alternative, while the sponsor actively pursues marketing approval. It gives early access and further safety and effectiveness data before general marketing.

CMS records that the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 allowed Medicare to pay routine costs of care for beneficiaries in certain categories of IDE study, removing a financial barrier to taking part.

Category A and Category B devices in Medicare coverage

When the FDA approves an IDE it places the device in one of two categories and tells CMS at the same time as the sponsor. A Category A (experimental) device is one for which the absolute risk of the device type has not been established and the FDA is unsure whether the device type can be safe and effective. A Category B (nonexperimental or investigational) device is one for which the incremental risk is the main question, or whose device type is known to be capable of being safe and effective, for example because other makers have approval or clearance.

The category sets what Medicare can pay for. For a Category A study that CMS finds meets its coverage criteria, Medicare covers routine care items and services, but not the device itself, which CMS describes as statutorily excluded. For a Category B study, Medicare may pay for the device and for routine care, provided that CMS decides before the first related claim that the study meets its coverage criteria. Payment for a Category B device may not exceed the amount Medicare would have paid for a currently used device that serves the same medical purpose and has been approved or cleared. Routine care means covered items and services that would have been given even outside the study.

The coverage criteria in 42 CFR 405.212 read in part like an appraisal of research value. The study must aim to test whether the device improves health outcomes, must not unjustifiably duplicate existing knowledge, and must have a design and sample size adequate to answer its question. It must also be registered on ClinicalTrials.gov, plan the release of results on all pre-specified outcomes including negative ones, and explain how far the results generalise to Medicare beneficiaries.

The FDA's 2017 guidance on categorisation allows a device to move from Category A to Category B for later studies once new data resolve the initial questions of safety and effectiveness.

Worked example of trial cost shares under each category

The figures below are hypothetical and chosen for illustration. An IDE study enrols 200 Medicare beneficiaries. Each receives a device costing USD 30,000 to supply, routine care costing USD 12,000 and study-specific research costing USD 8,000. A marketed device serving the same purpose is paid at USD 18,000. Medicare is assumed to pay routine care at cost, the sponsor bears everything else, and cost sharing and bundled hospital payment are ignored.

$$C_S = n\,(d + r + s) - n\,(a\,r + b\,p)$$

Here $C_S$ is the sponsor's cost for Medicare participants, $n$ is the number of Medicare participants, $d$ is the device cost, $r$ is routine care cost, $s$ is research cost, $p$ is the Medicare device payment, and $a$ and $b$ equal 1 when Medicare covers routine care or the device and 0 otherwise.

Step 1 (total cost). Per participant $c = 30{,}000 + 12{,}000 + 8{,}000 = 50{,}000$ USD, so the study costs $T = 200 \times 50{,}000 = 10{,}000{,}000$ USD whoever pays. Without CMS approval ($a = b = 0$) the sponsor bears all of it.

Step 2 (Category A). Medicare pays routine care of $M_A = 200 \times 12{,}000 = 2{,}400{,}000$ USD, so $C_S = 10{,}000{,}000 - 2{,}400{,}000 = 7{,}600{,}000$ USD.

Step 3 (Category B). Medicare also pays $D = 200 \times 18{,}000 = 3{,}600{,}000$ USD for devices, so $M_B = 2{,}400{,}000 + 3{,}600{,}000 = 6{,}000{,}000$ USD and $C_S = 10{,}000{,}000 - 6{,}000{,}000 = 4{,}000{,}000$ USD.

The total cost is fixed and only its distribution changes. Category B status moves 60% of the cost for Medicare participants to the public payer, which in effect buys part of the evidence it will later use for its own coverage decision. The payment cap leaves the gap of $g = 30{,}000 - 18{,}000 = 12{,}000$ USD per device with the sponsor.

How IDE evidence feeds economic evaluation and HTA

Because IDE studies usually supply the clinical data for a PMA, their design limits any later cost-effectiveness analysis. Dhruva and colleagues reviewed 78 PMAs for high-risk cardiovascular devices approved between 2000 and 2007. Only 33 of 123 supporting studies (27%) were randomised, 17 (14%) were blinded, 51 of 78 PMAs (65%) rested on a single study, and 187 of 213 primary end points (88%) were surrogate measures.

Of 193 high-risk therapeutic devices approved through the PMA route from 2014 to 2023, Mooghali and colleagues found that the main premarket study had several arms with concurrent controls for 95, a single arm with nonconcurrent controls for 89 and a single arm with no controls for 9. By July 2025, 68 of the 193 devices (35.2%) had had at least one serious recall. Single-arm designs with historical controls give no direct estimate of the incremental effect against the relevant comparator, and surrogate endpoints must be linked to final outcomes before they can yield QALYs.

CMS guidance notes that Medicare beneficiaries have historically been underrepresented in the studies supporting marketing authorisation, because of age, access, disability, comorbidity and concurrent treatment, and the Medicare generalisability criterion asks IDE sponsors to address this at the design stage. Tarricone and colleagues describe device features that challenge current assessment methods, including learning curve effects, incremental innovation that complicates the choice of comparator, and organisational impact with dynamic pricing. A value of information analysis or an early health technology assessment at the planning stage can show whether a larger or randomised design is worth its cost.

Boundaries with marketing approval, IND and coverage with evidence development

An IDE permits a device to be studied and makes no finding that it works. The regulation forbids representing an investigational device as safe or effective for the purposes under study. Marketing needs a separate FDA decision, such as premarket approval or 510(k) clearance. The investigational new drug application (IND) plays the same role for medicines under 21 CFR Part 312.

Coverage with evidence development (CED) is a different Medicare tool, in which a national coverage determination covers an item or service only within approved clinical studies or with further data collection. IDE coverage applies only to devices under an approved IDE, whereas CED can apply to any item or service whose evidence is not yet adequate for coverage under the reasonable and necessary standard. Both resemble the managed entry agreements used elsewhere, in which a payer links access to further evidence.

Limitations and common misreadings

The FDA can refuse an IDE for a scientifically unsound investigation, but an approved IDE says nothing about whether the design will answer the questions a payer or HTA body asks. It is also specific to United States law.

1. An IDE means the device has been approved. It allows lawful shipment for study only, and safety and effectiveness remain to be shown.
2. A non-significant risk study escapes oversight. The FDA need not be told about the study, but the IRB acts for the FDA and the abbreviated requirements still apply.
3. Medicare pays for the device in every IDE study. Medicare never pays for a Category A device, and a Category B device is covered only when CMS accepts the study, at no more than the payment for the existing device.
4. IDE evidence is ready for economic evaluation. Single-arm designs, surrogate end points and limited Medicare-age enrolment often leave large uncertainty in the incremental effect.
5. The category is fixed for the life of the device. The FDA can move a device from Category A to Category B for later studies when new data resolve the initial questions.

Sources

• US Code of Federal Regulations. 21 CFR Part 812, Investigational Device Exemptions (sections 812.1, 812.2, 812.3, 812.7, 812.20, 812.30, 812.36, 812.42, 812.66). eCFR as of 1 October 2026. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-H/part-812
• US Code of Federal Regulations. 42 CFR Part 405 Subpart B, Medical Services Coverage Decisions That Relate to Health Care Technology (sections 405.201 to 405.212). eCFR as of 1 October 2026. https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-B/part-405/subpart-B
• US Food and Drug Administration. Investigational Device Exemption (IDE). Content current as of 22 August 2024. https://www.fda.gov/medical-devices/premarket-submissions-selecting-and-preparing-correct-submission/investigational-device-exemption-ide
• US Food and Drug Administration. Significant Risk and Nonsignificant Risk Medical Device Studies. Information sheet guidance. January 2006. https://www.fda.gov/media/75459/download
• US Food and Drug Administration. FDA Categorization of Investigational Device Exemption (IDE) Devices to Assist the Centers for Medicare and Medicaid Services (CMS) with Coverage Decisions. Guidance, 5 December 2017. https://www.fda.gov/media/98578/download
• Centers for Medicare & Medicaid Services. Medicare Coverage Related to Investigational Device Exemption (IDE) Studies. Page last modified 10 March 2026. https://www.cms.gov/medicare/coverage/investigational-device-exemption-ide-studies
• Centers for Medicare & Medicaid Services. Coverage with Evidence Development. Guidance document, 7 August 2024. https://www.cms.gov/medicare-coverage-database/view/medicare-coverage-document.aspx?mcdid=38
• Dhruva SS, Bero LA, Redberg RF. Strength of study evidence examined by the FDA in premarket approval of cardiovascular devices. JAMA. 2009;302(24):2679-2685. https://doi.org/10.1001/jama.2009.1899
• Mooghali M, Arshad F, Rathi VK, et al. Premarket clinical evidence strength and recalls of high-risk therapeutic medical devices. JAMA Health Forum. 2026;7(8):e262626. https://doi.org/10.1001/jamahealthforum.2026.2626
• Tarricone R, Torbica A, Drummond M. Challenges in the assessment of medical devices: the MedtecHTA project. Health Economics. 2017;26(Suppl 1):5-12. https://doi.org/10.1002/hec.3469