FDA 510(k) Explained: Medical Device Premarket Notification

Historical Development of Medical Device Regulation

Medical devices were largely unregulated by the FDA until the mid-1970s. The Medical Device Amendments of 1976 (Public Law 94-295) fundamentally changed this landscape. Under the 1976 Amendments, Congress granted FDA authority to classify devices by risk and require either Premarket Approval (PMA) or 510(k) Premarket Notification before marketing. The PMA pathway, akin to drug approval, requires substantial clinical evidence of safety and effectiveness. The 510(k) pathway was created as a less burdensome “clearing” process: new devices could enter the market by proving they were “substantially equivalent” to a predicate device already legally marketed (^[1]) (^[2]). The premarket-notification requirement is codified at 21 U.S.C. § 360(k); the substantial-equivalence criteria are in 21 U.S.C. § 360c(i).

Initially, Congress envisioned most high-risk (Class III) devices being approved via PMAs. However, implementation lagged. The Safe Medical Devices Act of 1990 required FDA to either initiate PMA requirements or reclassify each pre-1976 Class III device, but many high-risk devices remained 510(k)-eligible because formal PMA regulations were delayed (^[24]). In practice, in the mid-2000s thousands of Class I/II devices (90% of submissions) and even several hundred Class III devices (via 510(k)) were being cleared (^[25]) (^[24]).

Since that time, multiple laws and regulations have incrementally updated the framework. Key milestones include:

• 1990 Safe Medical Devices Act: tightened postmarket requirements and called for Class III reclassification (^[24]).
• 1997 Food and Drug Modernization Act (FDAMA): streamlined FDA procedures, allowing Abbreviated and Special 510(k) formats and expediting some processes.
• 2002 Medical Device User Fee and Modernization Act (MDUFMA): introduced user fees for 510(k)/PMA submissions to support faster review, and created the third-party review program for some 510(k) devices (^[26]).
• FDASIA/MDUFA III (2012) and MDUFA IV (2017): reauthorized device user fees and set performance goals, including 510(k) review goals under MDUFA. User fees continue to fund FDA review activities (^[27]).
• 21st Century Cures Act (2016): provided measures to accelerate device approvals, including the new Breakthrough Devices program. It did not fundamentally alter 510(k) but encouraged use of Real-World Evidence.
• Medical Device User Fee Amendments V (2022): current law funding FDA through FY2027, under which FDA is implementing technology goals (eSTAR submission, performance-based reviews, and other initiatives) (^[27]). MDUFA VI negotiations for FY2028–2032 formally began in June 2025, with stakeholder meetings and public comment periods underway; the HHS Secretary must submit the new agreement to Congress by January 2027.

In sum, the 510(k) pathway has evolved under successive FDA user-fee legislative cycles, while substantive structural changes have largely been incremental. The IOM’s 2011 reports (both a 35-year review and a preparatory workshop) were requested by FDA to assess whether the 510(k) framework adequately protects public health (^[28]) (^[29]). These reports found mixed views: significant criticism that 510(k) often places devices on market without sufficient new safety data (^[2]) (^[16]), balanced by industry concerns that overly stringent changes could unduly delay innovation. Ultimately, the 2011 IOM consensus recommended devising a new “integrated premarket and postmarket regulatory framework” and replacing or fundamentally overhauling the 510(k) process (^[16]). FDA has not embraced such a sweeping reform, but has pledged ongoing 510(k) modernization (discussed below).

Device Classification and Regulatory Pathways

FDA-regulated medical devices are assigned to Class I, II, or III, with ascending risk. Class III devices (often implantable or life-supporting) generally require PMA or humanitarian device exemption (HDE). Class I devices (low risk) are often exempt from premarket review. Class II (moderate risk) devices require a 510(k) clearance unless exempt by regulation (^[1]). Thus, 510(k) is the workhorse pathway especially for Class II. Notably, until FDA issues a PMA requirement regulation, even some Class III device types can be cleared by 510(k) if predicates exist (^[30]) (^[24]). GAO found that from 2003–2007, FDA cleared 228 out of 342 (67%) Class III 510(k) submissions, even though class III was intended to be PMA-only (^[31]). In practice, certain Class III device types require PMA. Novel devices without a predicate that present low or moderate risk may be eligible for De Novo classification, while some Class I and II device types are exempt from 510(k) requirements (^[32]).

Other pathways complement 510(k):

• Premarket Approval (PMA) – Required for high-risk Class III devices without predicates. Involves substantive clinical evidence for safety and effectiveness and takes longer (goal ~180 days).
• De Novo Classification – A stream introduced in the 1997 law, for novel low/moderate-risk devices with no valid predicate. De Novo establishes a new device type classification (Class I or II) based on being “safe and effective” (^[33]). After a De Novo is granted, it creates a predicate for future 510(k)s.
• Emergency Use Authorization (EUA) – For public health emergencies (e.g. COVID-19 diagnostics). Not a permanent marketing authorization path.
• Humanitarian Device Exemption (HDE) – A marketing application for a humanitarian use device intended to benefit patients with a disease or condition affecting or manifested in no more than 8,000 individuals in the United States per year. It requires evidence of safety and probable benefit rather than effectiveness.

Table 1 (below) summarizes these main FDA pathways:

Pathway	Device Class	Predicate Required	Typical Evidence	Review Goal	Typical Example Devices
510(k) (Premarket Notification)	Class I/II (moderate risk)	Yes, substantially equivalent to predicate	Engineering/bench testing (biocompatibility, software validation, performance), and clinical data if needed for new claims or safety concerns ([13]) ([7]).	FDA target 90 FDA days (excluding pauses) ([34])	Infusion pumps, blood pressure monitors, surgical staplers, most imaging devices.
PMA (Premarket Approval)	Class III (high risk)	No predicate; full review required	Comprehensive clinical trial data for safety/effectiveness, plus full manufacturing controls.	Under MDUFA V, 180 FDA days when no advisory-committee input is involved; 320 FDA days when advisory-committee input is involved.	Pacemakers, implantable defibrillators, artificial heart valves.
De Novo (New Type)	Class I/II (novel)	No; qualifies if no predicate exists	Varies: bench tests, clinical data (depending on risk). Data to show sufficiency of safety/effectiveness.	FDA decision goal: 150 FDA days for 70% of De Novo requests; this is an FDA review-time goal, not a guaranteed elapsed timeline ([35]).	Novel diagnostics (e.g. a first-of-type AI algorithm), unique monitoring devices.
HDE (Humanitarian Device Exemption)	Class III (rare conditions)	No predicate; for a condition affecting or manifested in no more than 8,000 individuals in the United States per year	Safety and probable-benefit evidence; effectiveness need not be demonstrated.	No standard HDE review goal is stated here.	Humanitarian use devices for eligible rare conditions.

Table 1: Comparison of FDA device regulatory pathways for moderate- and high-risk medical devices, with typical evidence and examples (adapted from FDA regulations and publications).

Each pathway requires a detailed submission. Below we focus on 510(k) specifically.

The 510(k) process is governed by Section 510(k) of the Federal Food, Drug, and Cosmetic Act (21 U.S.C. § 360(k)), with implementing requirements in 21 CFR Part 807, Subpart E. Section 513(i) (21 U.S.C. § 360c(i)) governs FDA’s substantial-equivalence determination. A 510(k) premarket notification is required before commercial distribution of a new device or a device changed in a way that could significantly affect safety or effectiveness, unless the device is exempt. The goal of a 510(k) submission is to obtain FDA’s determination that the new device is “substantially equivalent” (SE) to a legally marketed predicate device (^[7]). “Substantially equivalent” means that the new device has the same intended use as the predicate and either (a) the same technological characteristics, or (b) differences in technological characteristics that do not raise new questions of safety or efficacy (^[7]). The predicate may be an older device marketed before May 1976, a reclassified device, or a device cleared through 510(k) previously. If FDA agrees the device is substantially equivalent, it issues a “clearance” (via a “SE Letter”) rather than an “approval”. If FDA finds the device not substantially equivalent (NSE), it cannot be marketed on that 510(k). The submitter may resubmit a 510(k) with new data, request De Novo classification, file a reclassification petition, or submit a PMA (^[32]).

Key Elements of a 510(k) Submission

Per 21 CFR 807.87, every 510(k) submission must include specific content items (^[36]) (^[7]). Among the most critical are:

• Device Identification – Trade (proprietary) name and common name, and FDA classification (Class I/II/III) if known (^[37]).
• Predicate Device(s) – Identification of one or more legally marketed devices alleged to be substantially equivalent. The submitter must explicitly state how the new device is similar to, and/or different from, the predicate(s) and provide supporting data (^[7]). For example, how design, materials, or performance differ, and why those differences do not affect safety/effectiveness (^[7]).
• Labeling and Proposed Directions for Use – Drafts of labels, instructions, advertisements, and promotional materials. These must fully describe the device’s intended use (indications) and user instructions (^[38]). Figures or drawings are included to show the device’s physical design if needed.
• Discussion of Changes – If the device has changes from a prior version (for a “Special 510(k)”), the submission must describe the design controls and verification/validation demonstrating the change’s impact is safe (^[39]) (^[40]).
• Performance Data – Data from nonclinical bench tests (e.g. engineering tests on inputs/outputs, durability, biocompatibility, sterility) are generally required. Clinical data may be needed if the predicate-equivalence can only be supported by human data (e.g. evaluation of a stent in blood vessels) (^[13]) (^[14]). In practice, infrequent Class II devices or novel claims often do use some clinical studies.
• Summary or Statement – A 510(k) Summary or Statement per 21 CFR 807.92/807.93. A 510(k) Summary is a concise summary of safety and effectiveness data, which FDA then posts on its website if cleared; a 510(k) Statement commits the company to providing data to anyone requesting it (in lieu of publication). The submitter chooses one. At clearance, if summary provided, FDA will release that summary along with the clearance letter (^[41]).
• Financial Disclosure – Any financial ties between investigators and company (per Part 54, e.g. for clinical studies).
• Certification and Statements – Certification that the submission’s information is truthful and no material facts omitted (^[42]). For Class III-type predicates (rare legacy cases), a Class III Summary and Certification is additionally required (^[43]).

In short, a complete 510(k) dossier reads like a mini-design history, including device description, how it was tested, labeling, and the path to demonstrating equivalence.

FDA provides detailed guidance on content/format (e.g. the “Refuse to Accept” checklists), but in essence the company must accrue enough evidence to convince FDA that “a reasonable assurance of safety and effectiveness exists for the new device, based on the predicate” (^[3]). If FDA needs more information, it will issue an “Additional Information” (AI) request.

Types of 510(k) Submissions

Although most 510(k)s follow the Traditional 510(k) format (full information per 807.87), FDA offers special pathways for streamlined review in limited cases:

• Traditional 510(k): The standard pathway submitting all required data. FDA review goal is 90 days (MDUFA target) from receipt of a complete application (^[34]).

• Special 510(k): Available for device modifications to a manufacturer’s own previously cleared device (here the predicate is the same firm’s legally marketed device). The premise is that if design controls were properly followed, much of the review can rely on the manufacturer’s verification/validation activities. A Special 510(k) can often be cleared in 30 days if no issues are found. The FDA states this is meant when “design control procedures can produce reliable results that form the basis for an SE determination” (^[39]). It streamlines review by focusing only on the differences/change rather than entire design.

• Abbreviated 510(k): When a device’s review depends on compliance with established standards or guidance. The submission references FDA-recognized consensus standards or special controls, providing declarations of conformity and summary reports rather than raw test data (^[44]). Abbreviated 510(k)s rely on having a well-developed set of applicable standards/guidance. If FDA finds an abbreviated 510(k) insufficient, it will convert the submission into a Traditional 510(k) but keep its original filing date (^[45]).

• Third-Party (Accredited) 510(k): The Third Party Review Program allows certain low- to moderate-risk devices (Class I/II) to be evaluated by FDA-accredited independent organizations (^[46]). If eligible, a company can voluntarily submit its 510(k) to an accredited reviewer instead of FDA. The third-party reviewer conducts a full review, interacts with FDA during review, and then recommends SE or not. FDA then has 30 days to make a final decision on that recommendation (^[47]). Currently about half of all 510(k) device types are eligible for this program (^[48]), which can speed review by offloading simpler devices to accredited entities. The FDA clarifies that third-party-reviewed 510(k)s carry no FDA user fee, since the company pays the private reviewer fee (^[26]). Recent FDA guidances (late 2024) further clarify how 3rd-party reviews work.

(FDA also allows Special Controls documents and guidance documents to shape evidence, and has pilot programs like the Safer Technologies Program, but these do not fundamentally change the 510(k) status.)

The 510(k) Review Process and Timeline

When a manufacturer decides to proceed with a 510(k), a series of steps is triggered. Here we outline the lifecycle of a 510(k) submission to FDA:

1. Pre-Submission Interactions (Optional): Before filing, a company can request a Pre-Submission (roughly akin to an informal Q-Submission) meeting or teleconference with FDA. This is not required but widely used for guidance on study protocols, specific issues, or feedback on a draft submission. For novel devices or borderline cases, a pre-sub can clarify data expectations. Scheduling and interaction timing depend on FDA’s current Q-Submission processes. Pre-Submissions are voluntary and are not listed among submission types subject to medical-device user fees; they can significantly reduce uncertainty. (Such interactions do not count toward official review timelines.)

2. Filing (Day 0) – Submission and User Fee: On the day of filing, the company submits the 510(k) to FDA (now required electronically via the CDRH Portal as of late 2022 (^[11])). They pay the user fee per MDUFA guidance (e.g. ~$26,067 for standard 2025 fee, or discounted ~$6,517 for small business (^[49])). The device is assigned a unique FDA number (e.g. “K240123” for the 123rd 510(k) in calendar 2024 (^[50])).

3. Acknowledgment of Receipt (≈ Day 7): Within about 7 calendar days of initial submission, FDA performs two checks: confirm the fee is paid and the submission is in acceptable electronic format (eSTAR or eCopy). If all is fine, FDA issues an Acknowledgement Letter confirming receipt date (the official Day 0) and 510(k) number. If fee or format is missing/incorrect, FDA issues a Hold Letter within 7 days, and the applicant gets up to 180 days to correct or else the submission is deemed withdrawn (^[51]).

4. Acceptance Review (15-Day Threshold): Starting from Day 0, FDA has 15 calendar days to conduct an “acceptance review” (also called RTA – Refuse To Accept screening). In this stage, FDA checks that all required elements per 21 CFR 807.87 are present and that it appears complete and in scope (^[52]). Common grounds for RTA include missing sections, unreadable copies, missing summary, or claims of predicate devices without evidence. If the submission fails acceptance, FDA issues an RTA Hold Letter and identifies deficiencies. The firm then has up to 180 days to address them and resubmit (effectively rolling their start date) (^[52]). If acceptance is not resolved, FDA considers the 510(k) withdrawn. FDA estimates that with properly eSTAR-formatted submissions, RTA rates should drop substantially (eSTAR is designed to avoid common RTA triggers) (^[53]) (^[52]).

5. Substantive Review (Goal 60 Days): Once accepted, the 510(k) enters substantive review. The lead reviewer (from the appropriate FDA office) has a goal of issuing an initial “Substantive Interaction” within 60 calendar days of submission receipt (^[54]). This is not the final decision, but rather the team’s first major communication, which can take two forms:

• Interactive Review Continuation: If FDA finds the application largely complete, it may notify the firm they will proceed interactively without delay on hold. In this case, the FDA communicates any needed clarifications or minor data requests through email/phone (“interactive review”), aiming to resolve issues within the overall MDUFA-target timeline (^[55]). The company can respond quickly with missing data or clarifications, keeping the review “live”.
• Additional Information (AI) Request / Hold: If FDA identifies substantive deficiencies requiring more data, it issues a formal Request for Additional Information (AI). This places the review on hold until the sponsor replies with the requested data. The AI letter typically explains specific items needed. The countdown clock (MDUFA days) is paused during the AI period. The submitter must respond by the date specified in the AI (usually 180 days max), or FDA may withdraw the application.

1. FDA Decision (Goal 90 Days): After receiving all information, FDA aims to make a decision by 90 FDA days (i.e., excluding AI hold time) from Day 0 (^[34]). FDA issues a Decision Letter by email indicating one of:

• SE (Substantially Equivalent) Clearance: The device is “cleared”. FDA will check the incorporated labeling/IFU and documentation, then send an official SE letter to the submitter (^[41]). The cleared device is then listed on FDA’s 510(k) database (weekly updates) and the 510(k) Summary (if provided) is posted.
• NSE (Not Substantially Equivalent): The device is not cleared for marketing. FDA’s letter explains why SE could not be established. At this point, the company may try to address issues (submit a new 510(k) or go for PMA) or withdraw.
• Additional Clarification or “Missed MDUFA”: If FDA cannot decide within the 90-day goal, it may issue a “Missed MDUFA Communication” by 100 days, outlining barriers to decision and a new timeline (^[56]). However, FDA’s internal performance goals emphasize a decision within 90 days.

Overall, while 90 days is FDA’s goal for a complete 510(k), in practice median review times often exceed this. For example, an analysis of ~25,000 510(k)s from 2010–2021 reported an average of ~164 FDA days between submission and decision (counting all calendar time including pauses, which exceeds the MDUFA clock) (^[10]). Another analysis found similar averages (~170 days). FDA’s own performance reports strive to meet the 90-day target, but variability exists by product type and complexity. For Class II devices with robust acceptance documentation, many 510(k)s do clear near the 90-day mark, whereas more complex devices with multiple deficiency cycles can stretch longer.

Post-Clearance Requirements

Receiving SE clearance permits marketing but does not end FDA’s oversight. Under Medical Device Reporting (MDR) requirements in 21 CFR Part 803, manufacturers generally must report device-associated deaths, serious injuries, and certain malfunctions within 30 calendar days of becoming aware of them; specified events may require a five-work-day report. Separately, under 21 CFR Part 806, a manufacturer or importer that initiates a correction or removal to reduce a risk to health or remedy a violation that may present a risk to health must report it to FDA within 10 working days. FDA also tracks recalls (voluntary or mandated). Additionally, FDA issued the Unique Device Identification (UDI) final rule on September 24, 2013. Its labeling and database requirements were phased in by device risk and remain subject to certain exceptions and compliance policies.

If after clearance a sponsor implements changes to the device beyond the cleared indications or technology, they may need a new 510(k) or special meeting to confirm continued substantial equivalence. FDA’s guidance on design changes, rebranding (if altering intended use), or modifying indications is a frequent subject of consultation. Failure to submit a new 510(k) when required may result in enforcement.

Volume of Submissions and Clearances

The 510(k) program processes thousands of submissions annually. According to FDA’s CDRH annual reports (summarized by industry analysis groups):

• Annual Volume: Roughly 18,000–21,700 total submissions (all types: 510(k), PMA, De Novo, etc.) have been received by FDA in recent years, trending upward from ~18,800 in FY2022 to ~20,700 in FY2024 and ~21,700 in FY2025 (^[57]) (^[6]). Of these, about 3,100–3,300 were 510(k) authorizations per year. For FY2022, Emergo/UL found FDA cleared 3,229 510(k)s; in FY2023 about 3,326; in FY2024, 3,107; and in FY2025, 3,238 (^[58]) (^[4]) (^[5]) (^[6]). This has been relatively steady, reflecting that 510(k) remains FDA’s primary marketing pathway.

• Clearance Rate: GAO (2009) found that ~90% of reviewed Class I/II 510(k) submissions were cleared, whereas only ~67% of Class III 510(k)s were cleared (^[31]). More recent CDRH data on overall acceptance (or RTA) rates are not routinely publicized.

• Device Categories: Cleared 510(k) devices span dozens of regulated categories (“Device Panels” and “Product Codes”). Feigal (then FDA CDRH Deputy Director) noted that approximately 6,700 device types exist (^[59]), each with different performance standards. Panel experts and standards facilitate review, but the sheer diversity means review teams see widely varying technologies.

• Timeline Performance: Under MDUFA V, the 510(k) decision goal is for 95% of applicable 510(k)s to receive an SE or NSE decision within 90 FDA days. Elapsed calendar time can be longer because it includes time awaiting the submitter’s response, as noted earlier (^[10]).

• User Fees: As of FY2026, the standard 510(k) user fee is $26,067; small businesses pay a reduced fee of $6,517 (^[49]). The annual establishment registration fee rose 23% to $11,423 in FY2026. These fees (and establishment registration fees) fund FDA’s review staff. CDRH staff grew from 2,011 in 2021 to ~2,230 by 2023 and ~2,260 by 2024 (^[60]) (^[5]). A December 2025 GAO report found that staffing constraints limited FDA’s oversight of medical-device recalls; FDA officials reported forgoing some oversight activities and prioritizing higher-risk recalls and earlier recall stages. The ‘510(k) program’ is financed by a mixture of federal and user-fee dollars, and FDA reports that “the page count of a 510(k) submission has more than doubled… to over 1,000 pages” since 2009, reflecting more extensive content and presumably greater review effort (^[12]).

• Diversity of Submissions: Table 2 shows a breakdown of FDA marketing authorizations in 2022–2023, illustrating how 510(k)s dominate numerically, with PMAs and De Novo much fewer.

Submission Type	2022	2023	2024	2025
510(k) clearances	3,229	3,326	3,107	3,238
Emergency Use Authorizations	222	138	48	9
De Novo reclassification requests	23	47	47	27
Premarket Approvals (PMA)	22	36	33	41
PMA Supplements	2,126	2,180	2,217	2,210
Humanitarian Device Exemptions (HDE)	1	3	4	1
HDE Supplements	108	77	108	114
Total FDA marketing authorizations	5,731	5,807	5,564	5,640

Table 2: FDA marketing authorizations in 2022–2025. Data from FDA CDRH Annual Reports summarized by Emergo/UL (^[58]) (^[4]) (^[5]) (^[6]).

This shows that in 2022–2025, roughly 5,500–5,800 total authorizations were granted annually, of which 510(k)s consistently made up the majority (~55–58% of all marketing authorizations). The decline in EUAs as the COVID-19 emergency receded was offset by steady growth in PMA supplements and De Novo classifications. The fact that over half of authorized devices are 510(k) underscores its centrality in the U.S. device market.

Postmarket Safety and Recalls

Concerns about 510(k) often center on postmarket safety. Some high-profile device failures raised questions about whether the 510(k) clearance process adequately assures long-term device safety. For example, metal-on-metal hip implants (initially cleared via 510(k)) led to thousands of revisions; similarly, the widespread recall of transvaginal mesh implants was tied to their rapid market penetration largely via 510(k) pathways (^[2]).

Empirical analyses provide perspective:

• Recall Frequency by Pathway: A 2021 JAMA Netw. Open study examined 28,556 devices cleared 2008–2017 (28,246 by 510(k), 310 by PMA). It found 3,012 (10.7%) of 510(k) devices were ever recalled, vs 84 (27.1%) of PMA devices. After risk adjustment, PMA devices had a significantly higher hazard of recall (HR ≈2.74) and of high-risk (Class I) recall (HR ≈7.30) compared to 510(k) devices (^[18]). In absolute terms, only 0.8% of 510(k) devices had a Class I recall, versus 5.2% of PMAs. Thus, PMA devices (generally higher risk) had more recalls per device. However, because so many more devices use 510(k), most recall events numerically involve 510(k) products.

• Predicates with Recalled Status: A 2023 JAMA cross-sectional study (Kadakia et al.) examined 156 510(k)-cleared devices that received Class I recalls in 2017–2021. Among the 127 index devices for which predicates were identified, 44.1% had used predicates with prior Class I recalls. In a matched analysis, devices cleared on predicates with past Class I recalls had 6.4 times higher risk of their own Class I recall than devices using Class I recall–free predicates (^[17]) (^[61]). The authors concluded that clearing devices based on problematic predicates has measurable safety consequences and that “stronger safeguards are needed” in the 510(k) pathway (^[17]) (^[23]).

• High-Risk (Class I) Recalls: In the Kadakia study, of 156 devices with Class I recalls, most (76.3%) had entered the market with no new clinical testing disclosed in their 510(k) submissions (^[62]). Only 5.8% of those recalled devices had any clinical data in the original submission. In this recalled-device cohort, the finding shows that many submissions had no disclosed clinical testing; it does not estimate clinical-testing practices across all 510(k) devices. (FDA has since asked for more clinical data in certain 510(k) contexts, as noted below.)

• Time to First Recall: The same study found that for devices eventually having Class I recalls, the median time from clearance to first Class I recall was 7.3 years (IQR 2.9–13.2) (^[63]). This highlights that safety issues can emerge long after market entry, complicating feedback into the premarket process.

A December 2025 GAO report (GAO-26-107619) further underscored postmarket challenges. Across FY2020–2024, FDA oversaw 3,934 medical-device recall events, all voluntarily initiated by manufacturers. GAO found that 74% of recalls exceeded FDA’s policy for terminating a recall within three months after the manufacturer’s completion request, and recommended that HHS and FDA address workforce-planning limitations and assess FDA’s authority over manufacturer-initiated recall strategies (^[64]).

Overall, these data illustrate the systemic trade-offs: while 510(k) enables fast market entry, it also means most marketed devices have not been through rigorous human testing at time of clearance. Regulatory and academic communities have debated whether this acceptance of uncertainty is justified, particularly for devices likely to be life-critical. The 510(k) statute relies on iterative improvements and a strong postmarket system, but critics argue the balance has shifted too far toward speed.

FDA’s Viewpoint and Initiatives

From the FDA’s perspective, the 510(k) program must balance its dual mission: protecting public health while helping bring safe device innovations to market quickly (^[65]) (^[66]). FDA identified Jeffrey Shuren as CDRH director on July 15, 2024. Michelle Tarver, MD, PhD, was serving as acting director by September 2024 and was selected as the permanent director in October 2024 (^[67]) (^[68]) (^[69]). In FDA statements (e.g. Sept 2023), FDA noted it “believes in the merits of the 510(k) Program” and has taken steps “over the past decade to ensure it meets patient needs and provides appropriate safeguards” (^[70]). They emphasize ongoing commitments to science-based review, transparent guidance, and tools to expedite review (like eSTAR and IT portals) (^[70]) (^[19]).

FDA also acknowledges that device complexity has grown: in 2023 the agency noted that average 510(k) submissions have doubled in volume since 2009, now exceeding 1,000 pages each (^[12]). This indicates more data (often clinical) accompany modern 510(k)s. In line with this, FDA reports that clinical evidence is being requested more often than before in 510(k) reviews (^[12]) (^[14]). FDA’s “Safety and Performance Based Pathway” guidance (2019) even creates an alternative to Traditional 510(k) for certain well-characterized device types with clear performance criteria (^[71]). The January 2023 CDRH Program Update highlighted 10 device-specific final guidances under this new pathway.

FDA’s recent initiatives (often under MDUFA V performance goals) include:

• eSTAR Submission Templates: Mandatory electronic templates to ensure completeness and consistency. eSTAR became mandatory for 510(k)s in October 2023 and for De Novo submissions in October 2025. A May 2025 draft guidance proposes extending mandatory eSTAR to Q-Submissions (Pre-Subs). eSTAR is also now available voluntarily for IDE and PMA supplement submissions (^[11]) (^[72]).
• Third-Party Program Upgrades: Final guidance (Nov 2024) to clarify third-party reviews and eliminate redundant FDA re-reviews (^[47]).
• Enhanced Review Procedures: More thorough interactive review to resolve questions earlier. Emphasis on meeting intermediate milestones (60-day interactions) has been codified in the 2018 Q-Submission guidance and implemented via the CDRH Portal.
• Additional Clarifications: FDA has issued guidances clarifying scientific requirements (e.g. for sterile devices, biocompatibility testing, software validation, cybersecurity, interoperable standards). In 2022–23, CDRH published dozens of new/updated guidances (including 5 explicitly on the 510(k) program) (^[73]). For a premarket submission for a device that meets the statutory definition of a cyber device, Section 524B requires cybersecurity information, including an SBOM; these requirements do not apply to every medical-device submission. Other recent guidances include the Computer Software Assurance (CSA) guidance (September 2025), replacing burdensome CSV with a risk-based approach; final Predetermined Change Control Plans (PCCPs) guidance for AI-enabled devices (August 2025); and the QMSR (Quality Management System Regulation), which became effective on February 2, 2026 and incorporates ISO 13485:2016 by reference.
• Medical Device Safety Action Plan: Launched in 2018 to spur total product life cycle (TPLC) surveillance, including improved adverse event analysis and device tracking information (UDI system) (^[71]).
• Breakthrough and Safer Technologies Programs: Though mostly postmarket or pre-certification, these encourage higher-risk roadmaps with more FDA collaboration; some cleared devices initially had 510(k) predicates.
• NEST (National Evaluation System for Health Technology): An FDA initiative (in partnership with NIH and MDIC) to collect real-world evidence to inform safety/emergency signals post-clearance. This may feed back to future regulatory changes.

The overall trend from FDA is toward greater data expectations and transparency. One example: FDA’s statements acknowledge that older predicates (over 10 years old) may rely on outdated technology, and thus one modernization proposal was to publicly identify when a 510(k) device is based on a very old predicate (^[74]). Even though such specific proposals generate debate, they illustrate FDA’s recognition of a need for evolutionary change.

On the resource side, FDA notes that CDRH’s budget (~$634M, 2021) is funded 35% by device user fees (^[75]), which underscores industry commitment to funding review timelines. Nonetheless, GAO has labeled medical device regulation as a high-risk federal program, and recommended bolstering staff expertise and enforcement capacity. FDA’s CDRH grew from ~2,011 staff in 2021 to ~2,260 by 2024 to handle growing workloads (new regulations, MDUFA mandates, and emergent technologies like AI). However, GAO reported that FDA officials described staffing constraints as limiting medical-device recall oversight and that the agency had not completed a current assessment of the staffing resources and skills needed for that work (^[64]).

Industry and Innovation Perspective

Device manufacturers largely view 510(k) as vital for bringing iterative improvements to market. Because 510(k) requires no new clinical trial as long as substantial equivalence is shown, companies, especially smaller ones, can introduce modified designs or new generations of products relatively cheaply and quickly. The financial and time costs of a PMA (typically tens of millions of dollars and several years) are untenable for most Class II innovations. Thus “the 510(k) has fostered innovation in the medical device industry” by enabling a favorable return on development investment. For example, Stanford med-tech expert Josh Makower noted decades ago that the U.S. ecosystem incentivizes clinicians and engineers to rapidly convert ideas into products — often via 510(k) (^[76]).

From industry trade associations’ standpoint, any 510(k) reforms should not unduly delay patient access. The BioMed Alliance and AdvaMed (industry groups) have pointed to FDA’s high clearance rate (often over 90%) and the need for predictability. They argue that evidence requirements should be commensurate with risk — echoing what FDA asserts (^[13]). Indeed, FDA’s Dr. Janice M. Feigal and others have emphasized that lower-risk devices may be adequately vetted with bench testing and standards use, without costly clinical trials (^[13]). Industry often asserts that mandatory trials for class II devices would reduce the pace of device innovation and availability of beneficial therapies.

Nonetheless, developers recognize that incremental changes may still benefit from robust biocompatibility or small clinical studies. The fact that 510(k) cleared devices have, on average, grown information-rich (thousands of pages of data) suggests industry is adapting to FDA’s higher expectations. Many companies nowadays include toxicology, performance testing under varied conditions, and even limited human study to strengthen SE claims.

Another industry perspective is global competitiveness. Since the U.S. 510(k) pathway is often compared to the EU’s CE-marking process (sometimes criticized as less rigorous), U.S. firms generally welcome a clear regulatory system. However, they must also navigate stricter new EU rules (EU MDR/IVDR) and often design dual submissions. The 510(k)’s similarity to CE marking (predicates vs. “equivalence” in EU) is noted, but U.S. law requires more explicit justification of equivalence (^[7]) and FDA maintains a national device registry. Companies weigh these international differences in global market strategies.

Perspectives on Safety and Criticisms

Patient safety advocates and some lawmakers have called for stricter oversight of 510(k). Key criticisms include:

• Lack of clinical evidence: Because 510(k) traditionally relies on bench and similarity data, some fear that devices may enter market without adequate proof of real-world safety/effectiveness. The IOM and GAO reports emphasized this gap. For instance, the 2011 IOM consensus committee stated that in many cases, “clinical data may be needed to address the human factors” of a device’s use (like actual patient outcomes) (^[14]), yet such data are often absent. The GAO found that essentially none of the cleared premarket notifications between 1980–1987 included data to verify performance claims (^[1]) (although performance standards have since increased).

• Use of older predicates: Critics note the phenomenon of “predicate creep,” where devices cleared today may trace lineage back to decades-old products, potentially drifting from modern safety standards. Legislative proposals like the 2012 SOUND Act (not enacted) sought to ban predicates with serious recall histories. While not law, these ideas highlight concerns that bad-actor devices can spawn chains of new products. The Kadakia JAMA study quantifies the risk of bad predicates (^[17]).

• High-risk devices via 510(k): GAO 2009/2011 showed that some implantable/life-sustaining devices (e.g. hip implants, AEDs) were still cleared via 510(k) because FDA had not reassigned them to PMA since the 1990 reclassification mandate (^[24]) (^[77]). Since then, FDA has reclassified some types (e.g. hip resurfacing went to PMA), but as of 2011 it allowed ~26 device types to remain 510(k)-eligible (^[77]). More recently, FDA has continued classifying devices to PMA where warranted, but advocates remain vigilant about any high-risk technology slipping through 510(k).

• Adverse outcomes of 510(k) clearances: As noted, cases like metal-on-metal hips, power morcellators, certain stents, and others have served as rallying points. For example, after FDA warnings in 2014 about “unsuspected” uterine cancers, several J&J power morcellator devices were voluntarily recalled (^[78]), and senators demanded removal of others. Critics point out these devices were originally cleared by 510(k) without premarket evidence of non-worsening cancer outcomes. Similarly, transvaginal mesh implants saw large-scale injuries; over 40,000 lawsuits later, FDA reclassified them in 2016 requiring PMAs. These events fuel arguments that 510(k) can let defective devices reach patients.

Despite these concerns, it is important to note that the absolute safety risk of most class II devices remains low. The Dubin et al. study shows low percentages of overall recalls, and Class I recalls are rare relative to the millions of devices used. Moreover, many 510(k) devices are non-invasive monitors or simple tools that pose little danger by design. Nevertheless, the trend toward more data, better postmarket surveillance (including NEST and device tracking when FDA has designated a device type for tracking), and FDA communication (MedWatch alerts and safety communications) indicates regulators and industry are acknowledging the need for lifecycle oversight.

Characteristics of Cleared 510(k) Submissions

FDA and academic analyses provide insight into what typical cleared 510(k)s look like:

• Review Outcomes: GAO’s analysis (2009) of 510(k) submissions from FY2003–2007 found a very high clearance rate: FDA cleared 90% of 13,199 Class I/II 510(k)s reviewed during that period (^[31]). Even for Class III device 510(k) submissions, 67% were cleared (^[31]). Denials and withdrawals were rare. This indicates that once a submission is accepted, only a minority (~10%) are rejected as not equivalent.

• Advancement vs. New: GAO Appendix (2009) observed that the vast majority of 510(k) submissions are incremental changes. For example, >90% of cleared 510(k)s from 2005–2007 were for devices with the same intended use and similar tech to predicates (^[31]). Very few were first-of-type or “breakthrough” technologies. In practice, novel devices often go via De Novo or PMA.

• Clinical Evidence Usage: In the Kadakia et al. study’s cohort of 156 510(k)-cleared devices subject to Class I recalls from 2017–2021, 119 devices (76.3%) had no clinical testing reported in their publicly available 510(k) materials, while 9 (5.8%) included clinical data; evidence could not be assessed for the remaining 28 devices because of missing or insufficient public information (^[62]). These figures do not estimate clinical-evidence practices across all Class II 510(k) devices. However, FDA scrutiny has increased. Feigal (IOM workshop 2010) noted that “clinical evidence is needed more often for class II products than one might expect” (^[14]), suggesting a gradual shift toward requiring human data even for some moderate-risk devices.

• Standards and Guidance: FDA encourages using consensus standards (ISO, ASTM, etc.) in Abbreviated 510(k)s (^[44]), and many device types have detailed performance standards (e.g. certain pacemaker standards). About 1,000 classification guidance documents exist (covering many devices) (^[79]) to streamline determinations. Verified compliance with standards often forms a large part of the review, especially for software and imaging devices.

• Page Count and Complexity: As noted, the “average page count” of 510(k) submissions has more than doubled since 2009 (^[12]), now often exceeding 1,000 pages per submission. This metric suggests increased incorporation of testing reports and detailed information (for instance, full bench test protocols, expanded labeling, extensive ISO declarations). The FDA view is that this reflects growing device complexity and patient safety considerations.

Case Studies (Examples of 510(k) Applications and Outcomes)

While detailed proprietary data on specific devices is confidential, several public examples illustrate 510(k) granting or denial decisions:

• High-Profile Recalls: Numerous devices that entered via 510(k) have triggered major FDA actions postmarket. For instance, transvaginal mesh implants (for pelvic organ prolapse) were mostly Class II products cleared on predicate meshes. Years of adverse event reports led FDA to reclassify these devices in 2016 to Class III, effectively ending most 510(k) clearance for new devices (^[2]). Similarly, the dePuy ASR hip system (metal-on-metal hip implant) was cleared by 510(k) in 2003 but later recalled widely; it eventually underwent a PMA for its successor. These cases highlight how 510(k)-cleared devices are subject to later scrutiny.

• Innovative Clearances: Many technologies are reviewed through the 510(k) program, but the marketing status and characteristics of any individual device should be verified in FDA’s public database. FDA’s record for iRhythm’s Zio XT ECG Monitoring System and Zio AT ECG Monitoring System (K190593) lists a decision date of August 23, 2019; it does not establish a claim that the product was the first AI-driven ECG analysis software (^[80]).

• Special and Abbreviated Usage: Consider a surgical patch manufacturer seeking minor design change (new adhesive) to its own legally marketed patch. They can file a Special 510(k) emphasizing design controls verification of the change (^[39]). If approved, the change can be cleared in ~30 days (instead of running a full new study). This contrasts with a Traditional 510(k) which might take 90+ days.

• Third-Party Example: An eligible Class II device could be submitted to an accredited reviewer. If the reviewer makes a positive recommendation, FDA makes the final decision. The review fee is set by contract between the submitter and the review organization, and FDA does not collect a user fee for third-party submissions (^[81]).

A more detailed fictional scenario: A new blood pressure cuff (Class II) manufacturer prepares a Traditional 510(k) including device description, bench tests (accuracy against reference), packaging and labeling, and a 510(k) summary (^[37]) (^[82]). They identify a predicate (an existing cuff brand), explain their device is similar except for material improvements, and show validation data (e.g., passing ANSI/AAMI standards). They submit via eSTAR in October. FDA issues an Acknowledgment Letter in about 7 days and conducts acceptance review within 15 calendar days (^[52]). By Day 45, FDA asks a simple AI question: “provide data on temperature stability” (substantive interaction). The company promptly sends test results. By Day 90, FDA issues an SE letter. After clearance, the company must register its establishment and list the device as applicable. A year later, if the company wants to add Bluetooth connectivity, it may need to submit a new 510(k), depending on the change and supporting data.

Table: Volume and Clearance Rates by Class (2003–2007)

GAO (2009) analyzed FDA data on submissions (Table excerpt):

Submission Type	Device Class	Reviewed	Cleared or Approved
510(k)	Class I and II	13,199	11,935 cleared (90%)
510(k)	Class III	342	228 cleared (67%)
Original PMA	Class III	217	170 approved (78%)
Supplemental PMA	Class III	784	664 approved (85%)

Table 3: GAO data on FDA 510(k) and PMA review decisions (FY2003–2007) (^[31]). Note: “reviewed” means submissions formally examined. ‘Cleared’ means SE; ‘Approved’ means PMA.

This table illustrates that in that era about 90% of 510(k) submissions were cleared, versus roughly 78% of original PMAs. It also shows that a nontrivial number of Class III devices still went through 510(k), with 67% of those cleared (though GAO recommended closing that loophole (^[15])).

Evidence from Research Studies

Academic researchers have further probed 510(k) outcomes:

• Recalls by Device Specialty: Dubin et al. found that, among specialties, orthopedics (spine, hip, etc.) had the highest recall rates. Devices in interventional cardiology (stents, catheters) and anesthesia (ventilators, pumps) also had notable recalls. The study’s hazard regression indicated specialties differ even within the 510(k) vs PMA category. The takeaway: the risk of recall is more dependent on device type and intended patients (and premarket pathway) than 510(k) status alone (^[83]).

• Efficacy vs. Safety: Some recall analyses suggest 510(k) might bias toward performance (they are frequently performance-oriented devices) while PMA tends to emphasize human trials. A viewpoint is that PMA devices often treat difficult conditions and thus by nature have more potential safety signals, while many 510(k) devices are ancillary (e.g. pumps, consumables).

• Export vs. Import: In a global context, studies compare FDA with other regulators (e.g. EU, Japan) in device clearance stringency. Some find FDA’s ratio of recalls to market share is similar or lower than other countries, but methodologies vary. The 510(k)’s reliance on predicates is somewhat unique; the EU historically did not formally use predicates (CE marking is a self-declaration with Notified Body review), though both systems have evolved.

Below are examples (with citations where possible) illustrating how the 510(k) process plays out in practice.

1. Transvaginal Mesh (TVM) for Pelvic Prolapse: Hundreds of thousands of women in the U.S. received mesh implants cleared by 510(k). Over time, thousands of adverse reports (pain, infection, erosion into vagina/bladder) emerged. Consumer lawsuits and FDA review led to a 2016 order requiring all TVM for prolapse to undergo PMA (reconverted to Class III) (^[2]). This case is often cited in discussions of 510(k) shortcomings, though technically the original clearances were lawful under the statutes at the time.

2. Laparoscopic Power Morcellators: These devices, used to fragment tissue during minimally invasive surgery, were cleared by 510(k) in the mid-1990s. In 2014 FDA issued a safety communication citing risk of spreading unsuspected cancerous tissue (^[78]). The largest manufacturer (J&J) voluntarily withdrew its devices; others remained on market. This episode prompted debate about whether the 510(k) review should have required more testing for unrecognized risks. A law firm commentary on the morcellator case invoked the IOM’s critique of 510(k) as “flawed” (^[84]) (^[85]), illustrating the political fallout that can follow 510(k)-cleared device failures.

3. Infusion Pump Recalls (2017): In 2017, FDA highlighted a wave of Class I recalls of large-volume infusion pumps (from different companies) due to programming errors and hardware failures. Many pumps in question had been grandfathered or 510(k)-cleared. The FDA convened a public workshop to improve pump safety. While FDA did not revoke 510(k) approvals directly, it has since updated guidances on pump software validation and cybersecurity, underscoring that some devices may need enhanced scrutiny despite being moderate risk on paper.

4. COVID-19 Pandemic Devices: During 2020–2021, FDA issued hundreds of EUAs for diagnostic tests and ventilators. Some diagnostic tests later sought full 510(k) clearance; FDA provided pathways to transition from EUA to fully cleared product, demonstrating flexibility in extraordinary situations. These actions show how 510(k) processes adapt to public health needs.

5. Novel Technology (AI/ML Algorithms): FDA maintains a public list of AI-enabled medical devices that have received marketing authorization. Computer-aided detection algorithms for radiology may be cleared by comparing outputs against legally marketed predicate devices and providing appropriate performance data. FDA’s final Predetermined Change Control Plans (PCCPs) guidance for AI-enabled device software functions explains how manufacturers may include planned modifications in a marketing submission. Section 524B imposes specified cybersecurity submission requirements only for devices that meet the statutory definition of a cyber device (^[86]; FDA cybersecurity FAQs).

The debate over 510(k) often frames two goals: ensuring patient safety vs. promoting innovation and timely access. The data and cases above illustrate that the 510(k) route does reduce premarket burden—and indeed nearly all low/moderate-risk devices enter this way—but at a cost of less upfront evidence.

Supporters of change argue that certain reforms could strengthen safety without sacrificing innovation. Proposals include: requiring summary of prior device problems, excluding recalled devices as predicates, routine use of postmarket data for clearance, and more mandatory clinical data for any implantable or high-risk tech entering by 510(k). FDA has already moved in some of these directions. For instance, the “Safety and Performance Based Pathway” mandates strict performance thresholds. For premarket submissions for devices that meet the statutory definition of a cyber device, Section 524B requires specified cybersecurity information; these requirements do not apply to medical-device submissions generally (^[87]). In Sept 2023, CDRH claimed the 510(k) process is being continually “strengthened and modernized” with new guidances to clarify complex topic areas (^[70]).

Opponents of drastic changes caution that adding premarket hurdles (like routine clinical trials) will slow important device introductions, potentially hurting patient health by delaying improved diagnostics, therapeutics, and quality-of-life tools. They also note that devices can often be iteratively improved after initial clearance through supplemental 510(k) filings, rather than withdrawing beneficial technology. The industry perspective emphasizes that real innovation in medical devices is often incremental, and that the 510(k) statute was designed to allow “long cycle” improvements. For example, an insulin pump manufacturer might tweak a sensor or algorithm yearly; forcing each tweak through PMA would be impractical.

In the end, 510(k) is a compromise codified by law: devices can be cleared on similarity, but postmarket surveillance must catch failures. FDA’s task is to make that system robust enough. Recent agency efforts aim to make postmarket obligations (UDI, registries, mandated clinical studies) fill in any evidence gap not obtained premarket.

Future Directions

Looking ahead, several factors will shape 510(k):

• Regulatory Science and R&D: FDA is investing in “Regulatory Science Tools” (statistical models, test methods, quantitative safety margins) to streamline device evaluation (^[88]). As tools are qualified, they may be incorporated in future 510(k) guidance (e.g. computational models in lieu of animal testing).

• Real-World Evidence (RWE): Integration of clinical use data (registries, electronic health records) could support both preclearance (for supplements or new claims) and postmarket surveillance. FDA’s NEST and FDA evaluations allow some existing clinical registries to be used as “clinical studies” for device safety. In future, 510(k) submissions might attach retrospective RWE analyses to bolster safety claims.

• Global Harmonization: The International Medical Device Regulators Forum (IMDRF) has issued guidance on software medical devices and lab-developed tests, which FDA now follows. As countries converge, companies may prepare one dossier to satisfy multiple regulators, which could pressure FDA to align with global standards.

• Legislative Pressure and Political Context: MDUFA VI reauthorization negotiations began in June 2025, with stakeholder consultations ongoing. The new FDA Commissioner Marty Makary (confirmed March 2025) has emphasized both scientific rigor and a deregulatory approach. Broader HHS reorganization has led to significant FDA workforce reductions, raising questions about sustained review capacity. Past proposals like the SOUND Devices Act (to bar recalled devices as predicates) continue to be discussed. Incremental statutory changes may accompany the MDUFA VI reauthorization, potentially including expanded requirements for clinical data in certain device categories.

• Technology Evolution: Emerging technologies such as artificial intelligence, machine-learning medical software, digital therapeutics, and 3D-printed devices are testing the limits of 510(k). FDA maintains a public list of AI-enabled medical devices that have received marketing authorization. FDA’s final Predetermined Change Control Plans (PCCPs) guidance for AI-enabled devices explains how manufacturers may pre-specify certain modifications in a marketing submission. The eSTAR requirement—mandatory for 510(k)s since October 2023 unless exempted—standardizes electronic submission for that pathway (^[86]; FDA’s 510(k) overview).

• Patient Involvement: FDA has increased patient representation via advisory committees and committee summaries. Future trends may include soliciting patient preference data for device tolerances (already done in some premarket studies), or requiring patient registries post-clearance for higher-risk devices (some proposals in Congress).

In summary, the 510(k) program continues to evolve through guidance, policy updates, and incremental regulatory changes rather than wholesale statutory overhaul. It remains a living process that FDA, industry, and other stakeholders regularly tweak. With the rapid growth of digital health and personalized medicine, future modifications will likely emphasize data-driven review and ongoing safety monitoring.

The FDA 510(k) premarket notification program is a central, workhorse mechanism regulating the majority of medical devices in the United States. It enables relatively fast market access by leveraging the concept of “substantial equivalence” to previously cleared devices, thereby avoiding full clinical testing in most cases. As this report has detailed, the process has many well-defined steps and requirements—device description, predicate comparison, labeling, performance data, and so on—and FDA’s official procedures establish clear timelines and interactive review phases (^[52]) (^[34]). Operationally, the program approves over 3,000 devices per year (^[58]) (^[4]) with high clearance rates, supporting the medical device industry’s rapid innovation cycle.

At the same time, the 510(k) pathway has drawn criticism, especially whenever device failures occur. Independent reviews (GAO, IOM) and recent studies have highlighted that problematic predicate selection can propagate risk. In the Kadakia et al. study, the finding that 76.3% of the 156 510(k)-cleared devices subject to Class I recalls had no reported clinical testing applies only to that recalled-device cohort, not to all 510(k)-cleared devices (^[17]) (^[18]). Such findings give policymakers and regulators pause, and indeed FDA has been gradually modifying the program to enhance rigor: requiring more premarket data and expanding postmarket surveillance, while also streamlining and clarifying submission requirements through guidance (^[12]) (^[46]). The tension between encouraging innovation and ensuring safety will continue to shape 510(k) in coming years.

Veteran FDA reviewers note the agency’s challenge: it must clear devices both swiftly and carefully. According to Dr. Feigal, FDA has to clear about 15 510(k) submissions per business day to handle the volume (^[9]). This reality explains why the 510(k) process is not as resource-intensive (in reviewer time) as drug or PMA review; many decisions must be made in weeks. The challenge is shaped by the complexity and volume of submissions. In areas of high risk, additional evidence may be needed to support a substantial-equivalence determination. Indeed, FDA acknowledges that substantial clinical data are increasingly expected, and that older predicate devices require careful scrutiny (^[12]).

Looking forward, the 510(k) process is likely to become more data-driven. The MDUFA VI reauthorization (FY2028–2032), currently under negotiation, may introduce new performance goals and digital infrastructure requirements. QMSR became effective on February 2, 2026, and the expansion of AI/ML-specific frameworks (PCCPs and lifecycle management) signals a maturing regulatory approach. We may see more use of registries and postmarket data, as well as more sophisticated bench models (including computational modeling), to inform SE determinations. We may also see stricter criteria for predicate acceptance (for example, FDA could choose not to allow a new device to claim SE to a recalled predicate, even if not required by law). The December 2025 GAO report's findings on recall oversight gaps may spur legislative action. Lessons from COVID-19 EUAs and digital device approvals continue to spill over to the general 510(k) process, further integrating clinical real-world outcomes.

In conclusion, the 510(k) pathway remains the pivotal link between device innovation and patient care in the U.S., and while it has been remarkably successful in delivering large quantities of medical hardware and software to the market, it is also being continuously reexamined. This report has underscored that thorough understanding of the 510(k) process—its strengths, weaknesses, and evolving contours—is essential for stakeholders on all sides: industry innovators aiming for market clearance, regulators upholding public health, and ultimately the patients relying on safe and effective medical technologies (^[2]) (^[23]).

Sources: Selected primary and secondary sources are cited throughout this article (^[89]) (^[31]) (^[13]) (^[12]) (^[58]) (^[4]) (^[34]) (^[49]) (^[17]) (^[18]), among others as cited throughout the text.