Your existing technical file may contain useful test reports, manufacturing validations and clinical data. The question is whether each record supports the current device configuration, intended purpose and claims, and whether the evidence answers the applicable MDR requirement. A document can exist, remain technically sound and still leave part of that question unanswered.
This guide focuses on turning an existing MDD dossier into a device-specific work plan. It does not determine your eligibility to continue placing legacy devices on the market during transition.
Define the device before comparing the documents
A gap assessment needs a stable scope. Otherwise, one person assesses the marketed device while another prepares documentation for a proposed variant, and their conclusions appear compatible until you compare the details.
Write a short assessment baseline identifying the device family, catalogue references, configurations, accessories, materials, software versions where relevant, manufacturing sites and intended markets. Include the intended purpose, patient population, intended users, use environment and claims. Record the classification rationale you propose to use under MDR rather than carrying forward the MDD classification without review.
Separate the currently marketed configuration from any proposed change. A broader patient population, different material or new use environment can affect the evidence you need, even when the commercial product name stays the same. If you intend to keep marketing a legacy configuration during transition, assess changes through the applicable transition and change-control rules separately.
For example, a laboratory report for one material and manufacturing process does not automatically cover a visually identical variant made with another material or process. Your baseline gives the reviewer enough information to ask the right question before deciding whether to reuse that report.
Organise the MDD to MDR gap analysis around evidence
Use the MDR technical documentation structure as the destination, then map the evidence you already hold into it. Annex II of the adopted MDR sets out six areas covering device description, information supplied by the manufacturer, design and manufacturing, general safety and performance requirements, benefit-risk and risk management, and verification and validation.
Annex III addresses post-market surveillance documentation. Those links show the adopted text, not a current consolidated EU version. Use the applicable current EU text when finalising your device-specific requirement register.
Do not describe every MDR heading as a new obligation absent from MDD. Your existing file may already address risk management, clinical evaluation, labelling and validation. The useful comparison is between the evidence you actually have and the requirement you need to demonstrate now.
Give each evidence decision one of three dispositions.
| Disposition | What it means | What supports the decision |
|---|---|---|
| Reuse | Retain an existing record as supporting evidence | Confirmed device coverage, suitable method, relevant acceptance criteria and a documented rationale |
| Reassess | Keep the record while reviewing whether its conclusions or coverage remain sufficient | A defined technical question, competent review and any supplementary analysis |
| Add | Create or obtain missing evidence, or produce a required document that the file lacks | A specified deliverable, responsible owner and objective completion criterion |
These are working categories, not regulatory classifications. An old test report may be reusable while its applicability assessment needs updating. One clinical report may contain usable data but require a new appraisal. Record those decisions separately rather than forcing an entire folder into one category.
Build a GSPR crosswalk that reaches the evidence
The general safety and performance requirements, or GSPRs, need a reasoned evidence map. The adopted Annex II, section 4, calls for applicability decisions, methods of demonstrating conformity, the standards or other solutions applied and precise references to controlled supporting documents.
Start from your MDD Essential Requirements matrix, but check each applicable MDR requirement on its own terms. Do not assume a one-to-one correspondence between rows or replace the document title while leaving the old reasoning untouched.
A useful row identifies the requirement, explains its applicability to your device and names the evidence location precisely. “Risk management file” is a starting point, not a sufficiently useful destination for someone checking a particular control. Refer to the relevant revision, hazard or control identifier and the verification record that demonstrates the control works.
When you rely on a standard, record the edition used and assess its relevance to the applicable requirement. A newer edition does not, by itself, answer whether you must repeat a historical test. Equally, a successful old test does not resolve a changed requirement or uncovered device configuration. Make the technical comparison visible, including any differences that affect your conclusion.
A worked reuse, reassess and add register
The following example is entirely illustrative. It describes an invented reusable, non-powered surgical instrument family supplied non-sterile for professional use. Document identifiers, revisions, gaps and owners are invented to demonstrate the method. They are not a complete requirement list or an approved assessment of a real device.
Assume the current instructions specify a maximum of 50 reprocessing cycles. Someone has proposed increasing that claim to 100 cycles in the MDR documentation, but the existing validation covers only 50. Keep the proposed claim separate from the marketed baseline while you assess it.
| Requirement area | Existing evidence and applicability | Decision and specific gap | Dependent records | Accountable owner and completion evidence |
|---|---|---|---|---|
| Annex II section 1, device description | SPEC-01 revision 6 lists the marketed instrument sizes and materials | Reuse the specification for the current scope. Add a mapping from each submitted variant to the applicable evidence | Submission index, classification rationale, test coverage matrix | RA lead. Controlled family description and a reviewed variant-to-evidence map |
| Annex II section 3, manufacturing | PROC-04 revision 3 describes finishing, but names a supplier no longer used | Reassess. Establish whether the current supplier uses the same controlled process and whether existing validation covers it | Material specifications, biological evaluation, risk file | Manufacturing lead. Approved process comparison and validation or justified coverage evidence |
| Annex II section 4, GSPR demonstration | ER-01 revision 5 maps MDD requirements to broad folder names | Add the MDR GSPR crosswalk. Existing references do not establish MDR applicability or locate specific evidence | All evidence sections | RA lead. Reviewed applicability rationales and exact controlled-document references |
| Annex II sections 5 and 6, risk and validation | VAL-07 revision 2 supports 50 cycles. RM-02 revision 4 addresses degradation within that limit | Reassess the proposed 100-cycle claim. The current evidence does not demonstrate that claim | Instructions, performance specification, risk controls, clinical evaluation | Engineering lead. A documented scope decision followed by evidence supporting the selected limit |
| Annex II section 2, information supplied | IFU-03 revision 5 states 50 cycles, while an uncontrolled proposed brochure states 100 | Reassess the information set against the selected scope and applicable labelling requirements | Claims register, validation, risk file | RA lead. Reconciled controlled instructions, labels and claims with resolved discrepancies |
| Annex III, post-market surveillance | PMS-01 revision 2 lists complaint collection but does not explain how to assess reports of instrument wear | Reassess the plan and add the missing method for collecting and evaluating relevant field information | Risk review, clinical evaluation, applicable surveillance report | PMS owner. Approved collection and analysis method with documented outputs when available |
This register distinguishes a missing crosswalk from a missing demonstration of performance. You can resolve the former by mapping and reviewing existing evidence. You cannot resolve the latter by writing a more confident paragraph.
In your own register, also record the status, decision date and reviewer. Keep the accountable owner distinct from contributors so that a gap involving manufacturing, clinical and RA work still has someone responsible for moving it forward.
Decide whether historical testing still answers the question
MDR migration is not an instruction to repeat every historical test. The adopted Annex II, section 6.1(b), expressly provides for a rationale where no new testing has taken place. That provision does not make every historical report sufficient.
Assess coverage before making a retest decision. Check the tested device against your current materials, manufacturing process, configuration and intended use. Then review the method, acceptance criteria, results and limitations against the requirement you need to demonstrate. A test summary may also leave you without the protocol or detail needed to understand the result.
In the illustrative instrument example, the existing report may remain useful for the 50-cycle claim. It cannot simply be relabelled to support 100 cycles. Engineering needs to identify what additional evidence the proposed claim requires, including the effects of repeated processing on relevant safety and performance characteristics.
A supplier change requires its own assessment. “Same specification” can be relevant evidence, but it is not the whole conclusion if process differences affect the finished device. Link the assessment to the actual manufacturing records and relevant evaluations rather than relying on a supplier's general assurance.
Where you decide against new testing, write down what you compared, what remains unchanged, what differences you found and why the evidence remains sufficient. Where the comparison identifies an unresolved question, define the investigation or testing needed to answer it.
Resolve claims before finalising linked documents
A single unresolved claim can hold up several otherwise complete documents. Identify that dependency early so that clinical, engineering and labelling owners do not each solve a different version of the problem.
For the illustrative 100-cycle proposal, you have two planning options. You can retain the existing limit in the proposed submission scope, subject to confirming that the evidence remains adequate, or develop the evidence needed to support the increased limit. Neither option excuses inconsistencies in the information supplied with devices already on the market.
The work should follow the decision. First, record the selected claim and configuration. Next, assess the evidence and update the relevant risk analysis. Then align the performance specification, instructions and clinical evaluation with the supported scope. Finally, check that the GSPR references point to those same controlled revisions.
Do not treat narrowing a claim as an administrative shortcut. Assess its effects on intended purpose, benefit-risk conclusions, existing information and any applicable change controls. If the discrepancy affects devices already supplied, consider that issue through your quality and post-market processes rather than leaving it in the migration spreadsheet.
Connect clinical evidence and post-market surveillance
Your technical-file gap analysis should identify clinical evidence questions without attempting to settle every clinical strategy decision in the same worksheet. Record which claims the clinical evaluation supports, where the evidence needs appraisal and whether proposed changes create an additional evidence need.
For the instrument example, the technical-file owner should flag the proposed increased reuse limit to the clinical and risk owners. The register should identify what they need to assess, not assume that either a revised clinical report or a post-market clinical follow-up plan will compensate for missing validation.
Post-market surveillance, or PMS, also feeds back into the evidence you intend to reuse. Complaints about wear, unexpected failures or difficulties following reprocessing instructions may affect your view of risk controls and existing performance conclusions. An empty complaints table alone does not explain whether your collection methods could detect the issue you are assessing.
Keep legacy surveillance and MDR certification preparation connected but distinct. MDCG 2022-4 rev.2 discusses appropriate surveillance of legacy devices and the MDR requirements applicable during transition. It is non-binding guidance. Do not assume that preparing a future MDR submission postpones your current surveillance duties, or that legacy surveillance constitutes a full MDR conformity assessment.
Turn the gaps into a dependency-led work plan
Prioritise decisions that change the scope or evidence strategy before polishing documents that depend on them. A classification question, unsupported intended use or uncertain manufacturing configuration can alter substantial parts of the file. Resolve those questions before spending time formatting references that may change.
Next, identify evidence that takes external coordination or a defined observation period to generate. Obtain technical proposals and assess prerequisites rather than assigning a generic duration to every gap. Document reconciliation and submission assembly come after the underlying evidence decisions, although you can improve indexing while technical work proceeds.
Use completion criteria that demonstrate an answer. “Update validation” is an activity. “Approved report covers the submitted variants and selected reuse limit, with deviations assessed and risk conclusions updated” describes evidence of completion. If the final report fails to support the proposed claim, the gap remains open even though the testing activity has finished.
Before treating the file as ready for assessment, choose a few important claims and trace them in both directions. Can you follow the claim from the instructions to its supporting evidence? Can you follow a significant test limitation back to the scope, risk conclusions and information supplied? That check exposes contradictions that a document-count dashboard will miss.
Questions that change the assessment
Can you reuse the MDD technical file as the MDR submission?
You can reuse suitable evidence, but assess its relevance and organise the submission around the applicable MDR requirements. An MDD certificate or a complete historical index does not establish that every MDR requirement has sufficient supporting evidence.
Does every gap require a new document?
No. Some gaps need clearer references or a documented applicability assessment. Others require additional analysis, testing or changes to the device or claims. Classify the evidence problem before assigning a writing task.
What should you do with an unavailable supplier report?
Identify exactly what you need the report to demonstrate, request the relevant evidence and assess available alternatives. Keep the gap open if you cannot substantiate the conclusion. A purchase order or certificate may answer a different question from the validation evidence you need.
Does finishing the gap assessment establish transition eligibility?
No. Technical-documentation readiness and eligibility to rely on transitional arrangements are different decisions. Keep a separate device-specific assessment of the applicable conditions. A completed migration register does not remedy an unmet legal condition.
Leave the assessment with decisions you can act on
A useful gap analysis gives you a defensible evidence-reuse decision, a defined set of unresolved questions and a work plan that follows their dependencies. You should be able to explain why an old report remains relevant, why another needs reassessment and what evidence will close each remaining gap.
If you need help deciding which technical-file gaps to tackle first, talk to Dovetail about your MDD to MDR gap analysis.
