Korea Class III Medical Device Clinical Evidence Requirements

What Counts as Clinical Evidence for MFDS

MFDS accepts three categories of clinical evidence for Class III/IV devices:

1. Literature-based evidence — Published clinical studies on the device or substantially equivalent devices
2. Equivalence-based evidence — Performance comparison with a device already approved in Korea
3. Local clinical study — Korean clinical investigation (sometimes required)

The choice depends on device novelty, intended use, and existing approval status.

Decision Framework: Which Evidence Path?

Device Situation	Likely Evidence Path	Timeline	Cost
Same device approved by FDA/EU MDR, same intended use	Literature + equivalence	4–8 weeks	$15K–$40K
Same device with modified indications for Korea	Literature + bridging analysis	8–16 weeks	$30K–$80K
Novel device with FDA De Novo/PMA	Literature + bridging + possibly local data	6–18 months	$100K–$500K
Novel AI/ML SaMD	Literature + Korean validation study	12–24 months	$200K–$1M+
High-risk implantable, no FDA/EU approval	Local Korean clinical investigation	18–36 months	$500K–$2M+

Literature-Based Evidence

The most common path for Class III devices with established global use.

What MFDS requires:

• Systematic literature search (PubMed, Embase, Cochrane, Korean clinical databases)
• Search strategy documentation (keywords, inclusion/exclusion criteria, date range)
• Quality assessment of included studies (PRISMA-compliant)
• Synthesis demonstrating safety and performance for intended use
• Korean clinical context analysis

What strengthens the submission:

• Studies including Korean or Asian patient populations
• Studies citing Korean clinical practice guidelines
• Post-market data from FDA / EU MDR vigilance databases
• Real-world evidence from Korean hospitals (when available)

Common gaps:

• Search strategy not transparent or reproducible
• No Korean patient population justification
• Outdated literature (>5 years without current updates)

Equivalence-Based Evidence

Comparison with a device already approved by MFDS.

What MFDS requires:

• Identification of the equivalent device (MFDS-approved KMD number)
• Technical characteristics comparison (device design, materials, performance specifications)
• Biological evaluation comparison (ISO 10993)
• Clinical performance comparison
• Discussion of any differences and impact on safety/efficacy

Critical: The equivalent device must be MFDS-approved, not FDA-approved or CE-marked. This is a common error among foreign manufacturers.

How to find equivalent devices:

• MFDS Medical Device Database (의료기기 정보포털)
• Korean Medical Device Industry Association (KMDIA) catalogs
• Korean hospital procurement records (limited access)
• Engage Korean clinical KOL for equivalence guidance

Local Korean Clinical Studies

Required when literature and equivalence evidence are insufficient.

When MFDS requires local studies:

• Novel devices with no MFDS-approved equivalents
• Devices with Korean-specific population concerns (e.g., genetic variants affecting drug-device combinations)
• AI/ML devices trained on non-Korean populations
• High-risk implantables with limited global clinical history

Korean clinical study structure:

• IRB approval at Korean clinical site (typically 6–12 weeks)
• Patient enrollment (varies by indication)
• Korean Good Clinical Practice (KGCP) compliance
• Korean Clinical Trial Registry registration
• Data analysis by Korean clinical research organization

Typical timelines:

• Pilot study (n=30–50): 6–12 months, $200K–$500K
• Pivotal study (n=100–300): 12–24 months, $500K–$2M
• Multi-center pivotal (n=300+): 18–36 months, $1M–$5M

How Existing Clinical Evidence Maps to MFDS

Source	Reusability for MFDS
FDA 510(k) clinical performance testing	High (with bridging analysis)
FDA De Novo clinical study	High (Korean context analysis required)
FDA PMA clinical study	High
EU MDR Clinical Evaluation Report	High (with Korean adaptation)
Real-world evidence from FDA/EU vigilance	Medium (supplementary, not primary)
Asian population studies (Japan, Singapore)	Medium-high (especially Japan)
Korean published literature	Highest priority

Common Pitfalls

Pitfall 1: Direct Translation of CE CER

The EU Clinical Evaluation Report is a starting point, not a finished MFDS submission. Direct translation without Korean adaptation triggers deficiency notices.

Fix: Add 2–5 page Korean context section, including:

• Korean patient population relevance
• Comparison with MFDS-approved equivalents
• Korean clinical practice guidelines reference

Pitfall 2: Insufficient Equivalent Device Analysis

Citing FDA-approved devices as "equivalent" without MFDS-approved equivalents is rejected. MFDS requires Korean-approved comparators.

Fix: Identify MFDS-approved equivalents before submission. If none exist, prepare for literature-based or local study evidence.

Pitfall 3: Assuming Local Clinical Studies Are Optional

For some Class III devices, MFDS will require local Korean clinical data even if FDA/EU MDR clearance exists. This is more common for:

• AI/ML SaMD devices
• Devices with novel Korean indications
• Implantable devices with limited global clinical history

Fix: Engage MFDS pre-submission consultation to clarify clinical evidence expectations before starting Korean STED preparation.

Pitfall 4: Underestimating Korean IRB Timelines

Korean Institutional Review Boards typically take 6–12 weeks for initial approval. Multi-site studies add coordination time.

Fix: Engage Korean clinical sites early; budget 3 months minimum before patient enrollment begins.

Pre-Submission Clinical Evidence Planning

Before submitting MFDS Class III registration:

1. ✅ Identify intended use precisely (drives clinical evidence requirements)
2. ✅ Search MFDS database for approved equivalent devices
3. ✅ Inventory existing clinical evidence (literature, FDA/EU clinical data, real-world data)
4. ✅ Gap analysis: what additional evidence is needed for Korea?
5. ✅ MFDS pre-submission consultation (optional but valuable)
6. ✅ Korean clinical KOL engagement (validates approach)
7. ✅ If local study required: site identification, protocol development, IRB strategy

Frequently Asked Questions

Q: Can we use FDA pivotal trial data for MFDS submission?

A: Yes, FDA pivotal data is high-value evidence for MFDS. Required additions: Korean patient population bridging analysis (2–5 pages) and discussion of any Korean-specific risk factors.

Q: How does MFDS view real-world evidence (RWE)?

A: As supplementary, not primary. MFDS accepts RWE from FDA MAUDE, EU Eudamed, and post-market literature as supporting evidence for safety claims, but typically requires prospective clinical data for novel devices.

Q: Do we need Korean KOLs as study investigators?

A: For local Korean studies: yes. For literature-based evidence: not required but Korean KOL endorsement strengthens the submission.

Q: How does MFDS evaluate AI/ML device clinical evidence?

A: MFDS published 2023 AI/ML medical device guidance requiring (a) algorithm performance validation on Korean-representative datasets, (b) bias analysis for Korean demographics, (c) post-market performance monitoring plan. Korean validation studies are typically required for AI/ML SaMD.

Q: Can Leanabl help plan our clinical evidence strategy?

A: Yes. Leanabl's Regulatory Pathway Strategy service includes clinical evidence gap analysis and Korean clinical strategy.

How Leanabl Helps

• Regulatory Pathway Strategy — clinical evidence planning
• Korea Medical Device Registration — full Class III/IV submissions
• Technical File Gap Analysis — clinical evidence readiness assessment

Contact Leanabl for clinical evidence strategy.

---

Last updated: 2026-05-15.