Part 3 of 5-Part Series: Design Verification & Design Validation - Meeting Requirements & Ensuring Intended Use with Objective Evidence

rmforge.io | Part 3 of a 5-Part Series on Design Controls

Design Verification and Design Validation (DV&V) is where engineering, pilot production, preclinical data, and regulatory requirements converge. It establishes whether the product is ready for the next design review, regulatory submission, ethics committee review, or clinical use. It is not just testing. It is structured, documented evidence that Design Outputs meet the Design Inputs and that the device conforms to the User Needs and Intended Uses.

Once Design Inputs (Part 1) are documented and Design Outputs (Part 2) are traceable and controlled, the next move is demonstrating through objective evidence that the product:

• Meets Product Requirements (Design Verification), and
• Conforms to User Needs and Intended Uses (Design Validation).

Design Verification and Design Validation are not Process Validation. Process Validation assesses process consistency. The three concepts can get blended together on occasion. Avoid that mistake.

DV&V vs Process Validation

Three reference points separate them:

• 21 CFR 820.30(f) and (g) (legacy text, now under the QMSR) governed Design Verification and Design Validation. Current FDA reference: 21 CFR 820.10(c).
• ISO 13485:2016 Section 7.5.6 governs Process Validation.
• ISO 13485:2016 separates them: Sections 7.3.6 and 7.3.7 (DV&V) versus Section 7.5.6 (Process Validation).

Note on regulatory citations: as of February 2, 2026, the FDA Quality Management System Regulation (QMSR) replaced the legacy Quality System Regulation. 21 CFR 820.30 subsections are [Reserved]. Cite 21 CFR 820.10(c) and the corresponding ISO 13485:2016 Section 7.3 subclauses.

Sample Size in DV&V: Risk-Based and Justified

DV&V sample sizes must be justified, not guessed. Sample sizes should be:

• Tied to the risk level of the failure mode or hazard, and consequently tied to the Product Requirement being verified
• Selected based on response type (attribute versus variable data)
• Calculated to manage Type I and Type II errors at levels appropriate to the risk

Below is one example of aligning risk with acceptance criteria for confidence and power. This is an illustrative example, not a reproduction from regulations, standards, or guidance.

In DV&V, the team confirms with risk-adjusted confidence and appropriate power that the tested devices meet and conform to the Design Inputs.

Statistical Techniques at DV&V

Statistical tools should match the data type and the conformance goal. Examples:

• Binomial testing for attribute-based tests (pass/fail)
• Equivalence testing or t-tests for variable data
• Tolerance intervals for one-sided or two-sided variable-data acceptance criteria

Always document the sample size rationale in the protocol. Do not state an unjustified quantity to be tested.

Design Verification: Test the Requirement

Design Verification confirms that each Product Requirement (quantitative input) was met by the Design Output through objective testing. Examples: tensile strength, flow rate, dimensions, fatigue durability, electrical isolation, software unit testing.

Best Practices

• Controlled and traceable builds. Each test unit shall be built under controlled conditions with full traceability: components, assemblies, Bill of Materials, Lot History Records, process instructions, and inspection records. Pilot Lots built under the released Quality Management System (QMS), packaged and sterilized as final product.
• One Verification Protocol per Product Requirement (or per coherent test set). Acceptance criteria come directly from the Product Requirement. Do not introduce new criteria in the protocol.
• Test Method Validation (TMV) is required. Per ISO 13485:2016 Section 7.6, any measurement process used in Design Verification shall be controlled and validated. The measurement system must be proven capable through Gage Repeatability and Reproducibility (Gage R&R) or equivalent measurement system analysis.
• No protocol-level Design Input modifications. Do not add new requirements or modify acceptance criteria in the protocol. If a requirement needs to change, revise the Design Inputs through Design Controls per ISO 13485:2016 Section 7.3.9.

Design Validation: Prove It Works

Design Validation evaluates whether the final product, built using production-equivalent materials and documented processes, satisfies and conforms to the User Needs and Intended Uses (qualitative inputs).

When to Design Validate

Design Validation occurs toward the end of development. The signals that you are ready:

• Successful preliminary simulated use
• Successful preliminary preclinical data
• Successful Design Verification across all Product Requirements
• Pilot or production Lots built under released QMS process documentation, packaged and sterilized as final product

How to Validate

• Preclinical Studies: cadaver, simulated use, or appropriate pre-clinical model. Apply Replacement, Reduction, and Refinement (3Rs) principles as appropriate for the pre-clinical model. Biological evaluation per ISO 10993-1.
• Human Factors Validation: per IEC 62366-1:2015+A1:2020 and FDA guidance on Applying Human Factors and Usability Engineering. Summative usability testing with representative users in representative use environments. Make sure you did some Formative Study work to develop requirements during your design input phase.
• Clinical Evaluation: per ISO 14155:2020 and applicable FDA regulations (21 CFR Parts 50, 54, 56, 812). For EU market, clinical evaluation per EU MDR Article 61 and Annex XIV.
• Note: TMV may be necessary for the simulated-use model itself when the model output is treated as objective evidence.

Validation Ties to Intended Use

If your Intended Use is, for example, "minimally invasive delivery of an implant through a tortuous vascular pathway," the Design Validation might include:

• Simulated use through deployment in a tortuous-anatomy model at 37 +/- 1 degree C in saline
• IACUC-approved, Good Laboratory Practice (GLP) preclinical testing in a relevant model
• Summative testing for surgeon usability in cadaver or another preclinical model
• IRB-approved clinical investigation (typically late-stage development, pre-submission for some pathways, or supporting post-market evidence)

Validation Passes When

• The device conforms to the User Needs and Intended Uses in the preclinical or clinical context
• Users confirm the device meets User Needs and Intended Uses (e.g., IEC 62366-1 summative testing)
• No unacceptable use errors are observed

Required DV&V Documentation

• Plan: when, how, and how many you will verify and validate
• Protocol: who, what, where, how, and acceptance criteria
• Test Method Validation (TMV): validated test method, referenced in or attached to the protocol per ISO 13485:2016 Section 7.6
• Raw Data: captured under document-controlled protocol
• Summary Report: conclusions based on the evidence

All records shall be maintained per ISO 13485:2016 Section 4.2.4.

Typical Sections of a DV&V Protocol and Report

A standardized structure supports consistency, traceability, and audit-readiness.

Protocol Sections

• Purpose: intent of the testing activity
• Scope: device, subassembly, or function under test
• References: applicable regulations, guidance documents, and standards
• Definitions: any terms with specific technical meaning
• Responsibilities: execution, review, and approval roles
• Background and Justification: relevant design history and rationale
• Acceptance Criteria: succinct recapitulation of the pertinent Design Input(s); measurable or pass/fail
• Test Materials: consumables and reagents (IPA, swabs, vials, beakers)
• Test Equipment: fixtures, tools, calibrated instruments
• Test Samples and Statistical Rationale: lot and unit IDs, quantity tied to statistical sample size justification
• Test Method: step-by-step instructions
• Deviations Section: documented procedural deviations and rationale

Report Sections (Protocol Sections, Plus)

• Results and Discussion: raw and summary data with interpretation
• Conclusion: explicit statement of whether the acceptance criteria were met
• Appendices: raw data, annotated photos, calibration certificates, traceability tables

Best practice: the Verification Report recapitulates the Verification Protocol in past tense. The report becomes one record for the entire test, which speeds turnaround, supports regulatory submissions, and improves future reference value.

Design Verification vs Design Validation: Summary

Common Errors to Avoid

1. Blending DV&V with Process Validation

Problem: teams run "validation" without separating Design Validation (does the device meet user needs?) from Process Validation (does the manufacturing process consistently produce conforming product?). Auditors find this routinely.

Solution: treat them as three distinct activities under three distinct sections of the QMS. DV per ISO 13485 Section 7.3.6, Design Validation per Section 7.3.7, Process Validation per Section 7.5.6.

2. Modifying Acceptance Criteria in the Protocol

Problem: the protocol introduces a new acceptance criterion (AKA product requirements) or relaxes an existing one because the team realized late that the original requirement was too tight.

Solution: if the criterion needs to change, revise the Design Input through formal Design Change per ISO 13485:2016 Section 7.3.9. The protocol does not author requirements.

3. Unjustified Sample Sizes

Problem: "n = 30" with no statistical rationale. A reviewer asks why and the team has no answer.

Solution: document the sample size rationale in the protocol. Tie it to the risk level of the failure mode and the response type (attribute or variable). Reference the statistical method used.

4. Missing or Inadequate TMV

Problem: the test method is unvalidated. The measurement system contributes more variability than the product itself. Verification data is unreliable.

Solution: complete TMV before executing verification. Gage R&R or equivalent. The measurement system shall be capable for the tolerance band of the requirement.

5. Design Verification or Validation Without Production-Equivalent Units

Problem: Design Validation runs on prototypes built with non-QMS materials or processes. The validation does not represent the marketed product.

Solution: use pilot or production Lots built under released QMS process documentation. Package and sterilize as the final product would be.

Final Thought

Design Verification and Design Validation are the checkpoints before design transfer and submission. Done well, they:

• Confirm Design Outputs meet the Design Inputs
• Demonstrate compliance with applicable standards and regulations
• Reduce the risk of downstream failure and post-market non-conformance
• Support efficient audits and regulatory approvals or clearances

DV&V reports should be explicit: a recap of the protocol, lot numbers and sample IDs, objective and analyzed data, discussion, execution and approval signatures, clearly stated rationales for any deviations, and a succinct conclusion stating what passed (or did not, and why).

The regulatory reviewer should never have to interpret the company's test intent or data. Tell them how it was done, what the outcome was, and why it matters, in the correct context.

Done poorly, DV&V protocols and reports create ambiguity that triggers rebuilding, retesting, and delayed market entry.

Verification proves you built it right. Validation proves you built the right thing.

The 5-Part Series

• Part 1: Design Inputs
• Part 2: Design Outputs
• Part 3: Design Verification and Design Validation (this article)
• Part 4: Risk Management in Design Control
• Part 5: Design Input and Design Output Matrix

References

Quality Management System Regulation (QMSR), 21 CFR Part 820. 89 FR 7523, Feb. 2, 2026.

ISO 13485:2016. Medical Devices, Quality Management Systems, Requirements for Regulatory Purposes.

ISO 14971:2019. Medical Devices, Application of Risk Management to Medical Devices.

Design Control Guidance for Medical Device Manufacturers. FDA, 1997.

FDA Guidance. Applying Human Factors and Usability Engineering to Medical Devices. FDA, Feb. 3, 2016.

IEC 62366-1:2015+A1:2020. Medical Devices, Part 1: Application of Usability Engineering to Medical Devices.

ISO 14155:2020. Clinical Investigation of Medical Devices for Human Subjects, Good Clinical Practice.

ISO 10993-1:2018. Biological Evaluation of Medical Devices, Part 1: Evaluation and Testing Within a Risk Management Process.

Regulation (EU) 2017/745 of the European Parliament and of the Council of 5 April 2017 on Medical Devices. Official Journal of the European Union, 2017.

Questions or feedback: mike@rmforge.io

Try the platform: rmforge.io