Team-NB MDR Documentation: Submission Completeness Checks (2026)

TL;DR

Team-NB's 2025 survey shows a persistent completeness problem at the start of technical-documentation review. In its 2025 MDR chart, 23% of respondents fall in the lowest completeness band and 42% in the next band. Together, the displayed rounded percentages put about 65% of responding notified bodies in bands where half or fewer of submissions are complete. These are shares of notified-body responses, rather than the percentage of all device applications that pass or fail. ¹

The survey describes the problem; Team-NB's Version 4 MDR technical-documentation guidance, adopted 21 April 2026, provides a way to examine the submission itself. It identifies recurring issues in completeness, structure, consistent information and retrieval of evidence. Its scope is the MDR. An IVDR dossier needs its own applicable requirements and guidance. ^{2 3}

Pure Global's proposed application is a pre-submission check that follows one claim from the intended purpose through labeling, risk management and supporting evidence, then confirms the relevant post-market plan. Resolve mismatches and missing evidence at that chain's weakest point. This approach targets a reviewable submission; it promises no particular certification time or outcome.

Read the survey at the level it measures

The survey cover describes data from 41 designated Team-NB members. Its MDR and IVDR completeness charts present distributions across reported completeness bands. They describe the share of submissions considered complete within each respondent's experience, then show how responses distribute across those bands. The publication does not supply an application-weighted microdataset for these charts. ¹

Source: Team-NB Medical Device Survey 2025, pages 15–16. The IVDR chart labels its highest band above 75%; the MDR graphic displays the final segment without a repeated label. Figures are reproduced as displayed. Each column totals 101% because the published values are rounded. Question-specific respondent counts are not stated on these slides. ¹

Adding the two lowest displayed bands yields 65% for MDR and 67% for IVDR. Those sums are useful descriptions of the charts. They leave the overall first-pass completeness rate of all dossiers unmeasured. A large notified body and a small one may contribute different application volumes, while these charts provide no weights to reconcile that difference. ¹

The same boundary applies to comparisons across regulations. The IVDR distribution puts more responses in the lowest band, but the charts alone establish neither the cause nor the size of an application-level difference. Device mix, respondent mix and application volume would matter to that analysis.

For a manufacturer, the justified conclusion is operational: completeness merits a deliberate submission check. The evidence supports treating it as a recurring review issue. It supplies no probability that an individual manufacturer's dossier will fail, no ranking of notified bodies by difficulty and no estimate of time saved by a particular consultant.

Completeness and substantive adequacy need separate checks

The survey describes completeness checking before review of the documentation's content. A file can contain every named document and still have weaknesses in the evidence or reasoning. Conversely, a technically strong study can be difficult to assess if its device identity, method, report version or relationship to the submitted configuration is unclear. ^{1 2}

The MDR's Annexes II and III define the technical-documentation framework. Team-NB Version 4 organizes practical expectations around that framework, including device description, information supplied by the manufacturer, design and manufacturing information, general safety and performance requirements, risk management, verification and validation, and post-market surveillance. The guidance reflects Team-NB's interpretation and expressly recognizes that additional information may be requested. ^{2 3}

A useful intake check therefore asks whether the submission is present and navigable. A separate technical review asks whether the evidence supports the intended purpose, claims and conformity argument. Passing the intake check leaves substantive questions to be assessed on their merits.

Pure Global recommends making those two judgments visible in the submission record. For each major evidence item, identify the document's owner and revision, the device or configuration it covers, the requirement it supports and any unresolved substantive question. This is a proposed implementation method, rather than an additional regulatory form.

It helps distinguish two common corrective actions. A missing link to an existing signed report may be a retrieval problem. A report on a different configuration with no adequate bridging rationale raises a different evidence question. Renaming the folder resolves the former more readily than the latter. The reviewer should state which issue the correction actually closes.

Four gaps lead to four different readiness decisions

Team-NB and the MDR already expect coherent, retrievable evidence. Pure Global's proposed comparison distinguishes the work needed when a submission fails that check. The cases below are hypothetical applications of those requirements, not measured survey outcomes or notified-body acceptance decisions. ^{2 3}

Suppose a device family has configurations A and B. The submitted intended purpose covers both. The reviewer finds a performance report and must decide whether the problem is finding the file, applying it to B, supplying missing evidence or revisiting the claim itself.

Pure Global's guidance-derived review aid. The applicable requirements, device and evidence determine the necessary work, including whether further clinical investigation is appropriate. ^{2 3}

The second and third cases can look identical in an upload checklist: both have a test report and a variant description. Their unresolved questions differ. In the second case, evidence for the bridge exists and needs assessment. In the third, the package has not established that bridge. Calling both “missing documentation” conceals a decision about evidence adequacy.

A further boundary separates either case from a changed intended purpose. The broader clinical claim needs evidence covering that proposition as well as a well-organized report. The manufacturer needs the relevant clinical and risk assessment rather than a formatting correction. MDR Annexes II–III and Team-NB V4 supply the evidence framework; the table organizes the corresponding remediation choice. ^{2 3}

Use the comparison to scope a work order. Identify the affected configuration and claim, the existing evidence and the unanswered question. A retrieval repair can be closed by a corrected package check. An unassessed applicability rationale needs a technical decision. An evidence gap or changed claim needs a separately scoped assessment and possibly additional work. These are internal readiness decisions, with conformity assessment remaining under the applicable regulatory process.

Make evidence retrievable in the submitted package

Team-NB Version 4 gives practical file-format and navigation recommendations. These include manageable file paths, document sizes, bookmarks and internal navigation. Such instructions support the reviewer's ability to locate information. They should be attributed to the guidance and confirmed against the chosen notified body's submission instructions. ²

The practical test is to use the package as the receiving reviewer will receive it. Open the exported files, follow the evidence references and check that the target document and revision are present. A link that worked only inside the manufacturer's document-management system may fail in the submitted copy.

The same test should cover appendices and referenced prior work. An earlier conformity assessment may supply relevant evidence, but its identity, scope and relevance to the present submission need to remain clear. A historical approval or certificate, by itself, leaves many of those details unstated. The guidance discusses use of evidence from previous assessments within a traceable submission. ²

For translated or multilingual material, use the same controlled terminology for the device, intended purpose and configuration. If an English summary and another-language source differ on a key condition, identify and resolve the discrepancy before submission. Translation quality becomes an evidence-consistency issue when the difference changes what the manufacturer claims.

This work can be tested independently from scientific adequacy. Give a reviewer the exported package and a short list of evidence references without access to the internal repository. Ask the reviewer to retrieve the cited document and confirm its identity and revision. Record retrieval failures separately from findings about the evidence itself.

Plan the work around unresolved evidence

A submission plan should distinguish work the manufacturer can close internally from questions that need discussion with the notified body. Team-NB Version 4 includes communication before an application is lodged. Its guidance can help prepare a specific question, with device scope and existing evidence attached, while the notified body's own processes govern the discussion. ²

For a configuration bridge, an internal team can first assemble the affected variants, differences and existing evidence. For an unclear clinical rationale, it can identify the actual claim, proposed evaluation and known limitation. A focused question gives the discussion a concrete object. Generic requests to "confirm that the file is complete" leave important technical assumptions unstated.

Completeness work also benefits from a clear stopping rule. An administrative check can close when the required documents are present, current and navigable. A material evidence gap stays open until its acceptance basis is resolved. This prevents a checklist completion percentage from obscuring a single consequential weakness.

Pure Global's recommendation is to prioritize the claims and configurations that carry the greatest unresolved evidence burden. There is little value in polishing every folder name while a central performance claim still points to a different device. Conversely, once the substantive relationship is justified, retrieval and consistency defects deserve prompt correction so the reviewer can assess it.

The survey provides context for this discipline, but it supplies no measured effect size for this proposed check. Completing it cannot guarantee intake acceptance, a shorter review, a certificate or access to another market. Certification and market-entry decisions depend on the applicable route, evidence and authority.

What the service budget covers

A dossier finding should be classified before a compilation service is priced. Representation, document compilation, evidence generation and notified-body assessment are different cost categories. Paying for one establishes no completion of another.

The following are Pure Global Master Price List, 2026 company service figures. Annual base rates for one covered registration/device group match the website's published schedule; the quantity unit and market-specific contract terms matter. They are private service fees, separate from government charges and private notified-body charges. Pure Global pricing

For a hypothetical eligible device with one covered unit in each of the EU, UK and Australia, the base representation total is $2,000 + $2,000 + $2,000 = $6,000 per year. This sums representation services only. It assumes each market's quoted base-contract conditions, not automatic eligibility based on a CE mark or a single global registration. The list specifies three-year terms for these base arrangements, a 50% first-year increase for annual contracts, and 50% of remaining contract value on early termination. Multi-registration tiers and the final contract change the quote. Pure Global pricing

Included work varies by market. Reference-approval submission support, modifications, renewals and authority correspondence are included where the relevant service schedule says so. The EU representative row specifically lists document review, free-sale-certificate requests and EUDAMED support; it does not include obtaining an MDR certificate. These boundaries should appear in the written scope. Pure Global pricing

Company-supplied 2026 service-list examples, not government tariffs or fixed quotations for an unspecified device. The list contains expert-scope qualifications and lower-price options for certain China-based expert work; confirm the applicable row and current quote. The public representation guide also identifies the US compilation range. Pure Global pricing Pure Global representation guide

Government charges remain a separate jurisdiction-, product- and route-specific budget line. List the notified body's conformity-assessment quote as a private charge, separately from government fees. Neither the $6,000 annual example nor the compilation ranges represent an all-in authorization cost, an approval promise or a review-time commitment. For the four gap types above, commission the work that resolves the actual question before treating a compilation package as complete.

Method and limits

This report compares the 2025 completeness charts in Team-NB's survey with Version 4 of its MDR technical-documentation guidance and the MDR Annex II–III framework. The survey's original pages 15 and 16 were visually checked against the extracted values. The 2025 rows are presented as published, with their rounding retained. ^{1 2 3}

The survey reports information supplied by members. The completeness slides leave the question-specific response count, application weights and underlying individual observations unavailable. This analysis makes no application-level rate estimate and attributes no causal effect to completeness on certification duration. It also makes no backlog estimate by subtracting cumulative certificates from cumulative applications: those are different units that require a linked lifecycle dataset to reconcile.

Version 4 is MDR guidance. The IVDR survey row is included only as a clearly labeled descriptive comparison; an IVD submission needs the IVDR and appropriate IVD guidance. Before the next submission freeze, choose one consequential claim and follow it through the five checkpoints above in the exported package. Record the precise mismatch or missing evidence, its owner and the acceptance condition for closing it. That produces a focused correction list that a regulatory team can use before handing the submission to its notified body.