Clinical Testing for Medical Devices

1. Basic Concepts

Medical device clinical investigation/clinical trial: the systematic evaluation, at qualified clinical trial sites, of the safety and effectiveness of a medical device submitted for registration under normal conditions of use.

Clinical evaluation: a broader concept referring to the assessment of a medical device using clinical data; clinical trials are one way of obtaining such data.

Core purposes:

  • Verify the safety of the product for its intended use

  • Verify the effectiveness/performance of the product for its intended use

  • Provide clinical evidence for product registration

2. Three Routes for Clinical Evaluation

Not every medical device needs a clinical trial before market launch. Depending on the regulations, the following routes are available:

Route When it applies
Exemption from clinical evaluation Low-risk products listed in the “catalogue of devices exempted from clinical evaluation”
Comparison with equivalent devices Demonstrating equivalence with marketed devices of the same type
Clinical trial Innovative products, high-risk products, and those that cannot be justified through the first two routes

3. Key Regulations and Standards

International

Standard/regulation Content
ISO 14155:2020 Good Clinical Practice (GCP) for clinical investigations of medical devices
ICH-GCP E6(R2) GCP for drugs; some principles apply to devices
Declaration of Helsinki Ethical principles for medical research involving human subjects

China

Regulation Content
Regulations on the Supervision and Administration of Medical Devices State Council Order No. 739
Provisions for Medical Device Registration and Filing NMPA Order No. 47
Good Clinical Practice for Medical Devices New GCP, 2022
Technical Guidelines for Clinical Evaluation of Medical Devices Issued in 2021
Catalogue of Medical Devices Exempted from Clinical Evaluation Regularly updated by the NMPA

United States

Regulation Content
21 CFR Part 812 Investigational Device Exemption (IDE)
21 CFR Part 50 Protection of human subjects
21 CFR Part 56 IRB (Institutional Review Board)
21 CFR Part 11 Electronic records and electronic signatures

European Union

Regulation Content
MDR (EU) 2017/745 Medical Device Regulation, including clinical evaluation requirements
MEDDEV 2.7/1 Rev.4 Clinical evaluation guidance (still useful for reference)
MDCG guidance series Specific guidance issued by the Medical Device Coordination Group

4. Types of Clinical Trials

By purpose

Type Description
Feasibility/pilot study Small sample; preliminary assessment of safety and performance
Pivotal study Large sample; provides the main evidence for registration
Post-market studies (PMCF/PMS) Ongoing collection of real-world data after launch

By design

Type Description
Randomized controlled trial (RCT) The gold standard, highest level of evidence
Non-randomized controlled trial Used when constrained by ethics or feasibility
Single-arm trial against a performance goal Comparison with historical/literature data; common for high-risk implants
Paired design The same subject uses both devices
Crossover design Subjects use different devices in sequence

By blinding

  • Open-label trial: both parties know the assignment

  • Single-blind: subjects do not know their group

  • Double-blind: neither subjects nor assessors know the assignment (often hard to achieve for devices due to visible differences)

5. Elements of Clinical Trial Design

1. Subject selection

  • Inclusion criteria

  • Exclusion criteria

  • Withdrawal criteria

2. Endpoints

Type Description
Primary endpoint The core measure reflecting the main study objective
Secondary endpoint Supporting measures
Safety endpoint Adverse events, serious adverse events
Composite endpoint A combination of several events (e.g. MACE)

3. Sample size calculation

Based on the primary endpoint, taking into account:

  • Expected effect size (difference from control)

  • Significance level (α, usually 0.05)

  • Power (1-β, usually 0.8 or 0.9)

  • Loss to follow-up (usually 10–20%)

  • Superiority/non-inferiority/equivalence hypothesis

4. Choice of control

  • Active control: a marketed product of the same type

  • Negative control: sham procedure, placebo (ethically constrained)

  • Historical control: literature data or an objective performance criterion (OPC)

6. Clinical Trial Process

Preclinical studies completed
 ↓
Protocol design and writing
 ↓
Ethics committee review (IRB/EC approval)
 ↓
Regulatory approval (NMPA/FDA/CA, etc.)
 ↓
Selection of sites and investigators
 ↓
Site initiation visit (SIV)
 ↓
Subject recruitment and informed consent
 ↓
Subject screening and enrollment
 ↓
Device use/implantation/treatment
 ↓
Follow-up (short-, medium- and long-term)
 ↓
Data collection, monitoring, auditing
 ↓
Database lock and statistical analysis
 ↓
Clinical study report (CSR)
 ↓
Registration submission

7. Key Parties and Responsibilities

Role Responsibilities
Sponsor Initiates, manages and funds the trial; overall responsibility for data quality
CRO (Contract Research Organization) Manages the trial on behalf of the sponsor
Principal investigator (PI) Responsible for conducting the trial at the site
Ethics committee (IRB/EC) Reviews the protocol and protects subjects’ rights
Monitor (CRA) On-site monitoring to ensure protocol compliance and data accuracy
Auditor Independently assesses trial quality
Data management (DM) Data entry, cleaning and database lock
Biostatistician Sample size, analysis plan, statistical report
Regulatory authorities Approval and oversight by the NMPA/FDA/EMA, etc.

8. Ethics and Subject Protection

Core ethical principles

  1. Respect: informed consent, autonomy

  2. Beneficence: minimize risks, maximize benefits

  3. Justice: fair selection of subjects

Informed consent

Key information that must be included:

  • Purpose, procedures and duration of the trial

  • Expected benefits and potential risks

  • Alternative treatments

  • Subject rights (withdrawal at any time)

  • Privacy protection

  • Compensation for injury

  • Contact information

Protection of vulnerable groups

Children, pregnant women, people with mental disorders, prisoners, etc. require additional protective measures.

9. Examples of Common Endpoints (Different Devices)

Device type Example primary endpoint
Coronary stents Target lesion failure (TLF) at 12 months
Artificial joints 5/10-year survival rate, Harris score
Pacemakers Pacing threshold, sensing, impedance stability
IVD reagents Sensitivity, specificity, agreement with the reference method
Surgical robots Surgical success rate, complication rate
AI-assisted diagnosis AUC, sensitivity, specificity

10. Adverse Event Management

Term Description
AE (adverse event) Any untoward medical occurrence
ADE (adverse device effect) An AE related to the device
SAE (serious adverse event) Death, life-threatening, hospitalization, disability, etc.
SADE (serious adverse device effect) Serious and related to the device
UADE (unanticipated adverse device effect) A serious event not anticipated in the protocol

SAE reporting timelines:

  • China: report to the sponsor within 24 hours; the sponsor reports to the authorities within 7 days (death/life-threatening) or 15 days

  • US FDA: 10 working days (UADE)

  • EU: in accordance with MDR requirements

11. Special Types of Studies

1. Real-world studies (RWS/RWE)

  • Using data from real clinical practice

  • In China, the Hainan Boao Lecheng pilot zone has cases of RWE supporting registration

2. Adaptive design

  • Adjusting sample size, dropping arms, etc. based on interim data

  • Improves efficiency; must be pre-specified in the protocol

3. Registry studies

  • Long-term, large-sample observational studies

  • Commonly used for post-market surveillance of implants

4. Multi-center, multi-regional clinical trials (MRCT)

  • The same protocol conducted in different countries

  • Data can be used for registration in multiple countries

12. Post-Market Clinical Follow-up (PMCF/PMS)

Under the EU MDR, high-risk devices require ongoing PMCF:

  • Collect real-world use data

  • Regularly update the clinical evaluation report (CER)

  • Submit periodic safety update reports (PSUR)

Similar requirements in China are reflected in the medical device adverse event monitoring and re-evaluation systems.

13. Common Challenges

  1. Slow recruitment: enrollment is difficult for rare and serious diseases

  2. Protocol deviations: execution departing from the protocol

  3. High loss to follow-up: a common problem in long-term follow-up

  4. Data quality: inconsistent source data, incomplete records

  5. Time-consuming ethics review: multi-center trials need approval from several ethics committees

  6. Regulatory differences: multinational trials must reconcile different regulatory requirements

  7. Difficulty blinding: differences in device appearance and operation make double-blinding hard

  8. Choice of control: sham controls are hard to use for high-risk implants

14. Key Trial Documents

  1. Clinical investigation plan (protocol)

  2. Investigator’s brochure (IB)

  3. Case report form (CRF)

  4. Informed consent form (ICF)

  5. Statistical analysis plan (SAP)

  6. Clinical study report (CSR)

  7. Data management report (DMR)

  8. Monitoring and audit reports

  9. Adverse event summary

Scroll to Top