New EU and FDA Developments in IVD Evidence and Reporting

Regulatory evidence insights | 6 November 2026

New guidance on orphan IVDs and revised EUDAMED arrangements affect how manufacturers justify evidence and report performance. Each update calls for a different response, from reviewing a performance evaluation report’s (PER’s) remaining uncertainties to managing a validated public summary. For regulatory professionals, the immediate task is to connect study plans and performance claims with the evidence supporting them, while distinguishing binding requirements from nonbinding guidance.

Clinical Evidence for Orphan IVDs

Published on 3 November 2026, MDCG 2026-6, Clinical evidence for orphan in vitro diagnostic medical devices, is nonbinding guidance. The guidance explains when limited pre-market clinical-performance data may be acceptable. Relevant General Safety and Performance Requirements (GSPRs) must still be met, the benefit–risk balance must be acceptable, and the inability to generate more data within a suitable pre-market timeframe needs justification. Scientific-validity and analytical-performance evidence can contribute, with remaining limitations addressed through appropriate post-market surveillance and post-market performance follow-up (PMPF). Its annexes address PER content, extrapolation, and study design.

Orphan IVDs may have limited pre-market clinical performance data, but only where the limitation is justified, benefit-risk remains acceptable, and a rigorous post-market evidence plan closes the gaps.

Core eligibility and evidence standard

An IVD is considered orphan when it targets a disease, condition, or circumstance affecting no more than 12,000 people in the EU per year, and there are insufficient alternatives or it is expected to provide a clinical benefit over available alternatives/state of the art.

The manufacturer must substantiate the epidemiology, clearly define the target population, describe the intended purpose, and justify unmet need or expected clinical benefit. Artificially restricting use to a small subgroup is not acceptable.

The normal IVDR performance-evaluation framework still applies: evidence must support scientific validity, analytical performance, clinical performance, relevant GSPR compliance, safety, and expected clinical benefit.

When less pre-market clinical performance data is acceptable

A reduced amount or quality of pre-market clinical performance data can be accepted only if all of the following are justified:

• All available clinical and non-clinical evidence has been evaluated and remaining gaps are identified.
• Existing evidence supports relevant GSPR compliance, an acceptable benefit-risk profile, and expected clinical benefit.
• Obtaining further pre-market clinical data is not feasible or proportionate within an acceptable timeframe.
• A detailed PMPF and PMS plan will collect data promptly to address the gaps as far as possible.
• Users are informed of orphan status, evidence limitations, and how to report incidents, complaints, and clinical experience.

Evidence gaps should be limited to what is genuinely difficult to obtain. For example, if negative specimens are available, specificity should still be adequately characterized.

Evidence sources and study approaches

Strong non-clinical evidence can reduce the pre-market clinical-data burden when relevant and well justified. It may include analytical data, scientific literature, similar assays, registries, simulations, tissue/ex vivo models, contrived samples, and data from comparable CE-marked or research-use products.

Clinical evidence may draw on literature, expert consensus, registries, biobanks, prior studies, legacy-device experience, derogation use, and data from other jurisdictions.

Clinical performance studies remain the default unless reliance on other sources is justified. Studies should use an appropriately defined target population, clinically relevant endpoints, bias controls, and, where proportionate, multicenter recruitment.

Retrospective biobank studies, adaptive or sequential designs, descriptive statistics, Bayesian or hybrid Bayesian methods, and carefully selected external controls may be suitable for small populations. These approaches must be pre-specified and rigorously justified.

Extrapolation to an orphan indication

Data from a non-orphan population can support an orphan indication when it is relevant, high quality, and scientifically transferable.

Full extrapolation may be possible only if population, disease, sample/pre-analytical factors, device characteristics, and clinically meaningful outcomes do not materially differ, and the data are sufficient on their own.

Partial extrapolation combines transferable data with additional clinical and non-clinical evidence, often using an appropriate statistical model.

Extrapolation is not appropriate where relevance cannot be shown or differences undermine a valid inference.

Post-market evidence is central

PMPF and PMS are critical because they must resolve accepted pre-market uncertainty throughout the device lifecycle. The PMPF plan should specify the evidence gaps, data sources, projected specimen/patient volumes, timing, analysis milestones, risk-management measures, and actions if new safety or performance concerns emerge. Notified bodies can make certification conditional on defined PMPF/PMS activities and may suspend or withdraw certification if binding conditions are not met.

The Performance Evaluation Report should document orphan-status justification, all evidence and residual limitations, benefit-risk reasoning, user information, PMPF/PMS commitments, and planned periodic updates.

A practical starting point is to separate the rationale for orphan eligibility from the assessment of whether the evidence supports the intended purpose. For example, a hypothetical assay for a rare molecular subtype might have a strong epidemiological rationale but limited evidence for one specimen type. Those are different questions, and the PER should make the unresolved specimen question visible. A narrowly worded claim should reflect a scientifically defensible population and use.

Manufacturers can organize the evaluation around the uncertainty attached to each claim. For every gap, record the available evidence, its relevance, and the consequence of an incorrect result. This makes it easier to explain why a proposed study is proportionate and which conclusions remain uncertain. Regulatory professionals can use the same record when discussing the evidence plan with a notified body.

PMPF planning should then be operational. A useful proposed plan identifies how specimens will be obtained, the expected accrual rate, and who will review interim findings. It can also define the action to consider if recruitment falls behind or performance differs from expectations. These are implementation suggestions: the appropriate activities depend on the assay, its risks, and the remaining uncertainty. For SOTA reviewers, the immediate task is to appraise the relevance of alternative tests and the evidence supporting the proposed clinical benefit.

Step-by-step compliance checklist

A compliant orphan IVD program must substantiate orphan status, demonstrate an acceptable pre-market benefit-risk profile, and use PMPF/PMS to resolve justified evidence gaps.

1. Confirm the device qualifies as orphan

• Define the disease, condition, circumstance, or clinically valid subpopulation covered by the intended purpose.
• Document that the relevant condition/circumstance occurs in no more than 12,000 individuals in the EU per year.
• Use EU-relevant authoritative epidemiology where available, such as peer-reviewed literature, public-health statistics, registries, EMA records, or Orphanet.
• If EU incidence is extrapolated, document the method, regional variation, assumptions, and supporting expert or health-service data.
• If relying on a subpopulation, demonstrate why it is clinically valid and why the device is not appropriately usable in the wider population.
• Confirm that the restricted intended use is not an arbitrary means of meeting the incidence threshold.
• Document either: insufficient available alternatives, or an expected clinical benefit versus alternatives/state of the art.

2. Define the intended purpose and target population precisely

• Specify the device, analyte, specimen type, clinical setting, target users, intended function, and claimed clinical benefit.
• Clearly define the target population and any inclusion/exclusion criteria.
• Assess whether a broad screening claim would create unacceptable false-positive risk in a low-prevalence population.
• Align labelling, IFU, study population, performance claims, risk management, and PMPF population with the defined target population.

3. Build the evidence inventory

• Create a complete inventory of available scientific-validity, analytical-performance, clinical-performance, and post-market data.
• Evaluate clinical literature, expert consensus, registries, biobanks, previous studies, legacy-device experience, derogation use, and relevant evidence from other jurisdictions.
• Identify relevant non-clinical evidence, including analytical studies, similar assays, simulations, tissue/ex vivo models, contrived samples, and data from similar CE-marked or research-use products.
• Critically appraise the quality, relevance, scientific validity, and applicability of every evidence source.
• Map the evidence against each applicable IVDR Annex I GSPR and each performance claim.

4. Identify and justify evidence gaps

• Document which GSPRs and performance requirements are fully met, partly met, or unsupported.
• Identify every limitation in pre-market clinical performance data and associated residual risk.
• Explain why additional pre-market clinical performance data are not feasible or proportionate within an acceptable timeframe.
• Limit the evidence shortfall to genuinely hard-to-obtain information. For example, do not leave specificity under-characterized where negative specimens are available.
• Demonstrate that the total evidence supports intended performance, patient safety, expected clinical benefit, and an acceptable benefit-risk ratio.

5. Decide whether to generate additional clinical evidence

• Treat a clinical performance study as the default approach unless reliance on other clinical-performance sources is duly justified.
• Obtain input from appropriate clinical, epidemiological, and statistical experts for the evidence strategy and study design.
• Use multicenter specimen collection where appropriate and proportionate to improve sample size, representativeness, and robustness.
• Predefine clinically relevant objectives and endpoints that support both performance and clinical benefit.
• Select a justified comparator: reference standard where available; otherwise, a suitable CE-marked IVD, or a documented approach for determining true clinical status when no comparator is feasible.
• Control and document risks of selection bias, specimen bias, confounding, and lack of representativeness.

6. Use proportionate study and statistical methods

• Justify prospective or retrospective collection; if using biobank specimens, assess and control for storage, specimen-characteristic, and data-quality bias.
• Consider adaptive or sequential designs only when interim analyses and safeguards against false-positive conclusions are pre-specified.
• Where conventional inferential statistics are unsuitable, justify descriptive, Bayesian, hybrid Bayesian, or other appropriate methods.
• If using historical or registry controls, ensure control patients have comparable disease and clinical/demographic characteristics.
• Pre-specify all models, assumptions, analyses, decision criteria, and use of simulated data.
• Use simulations only to contextualize uncertainty or stress-test assumptions; do not treat them as a substitute for empirical performance testing.

7. Assess whether extrapolation supports an orphan indication

• Determine whether existing data from a non-orphan population are relevant because the indication, endpoint/outcome, or validated surrogate endpoint is the same or similar.
• Do not extrapolate if relevance cannot be established.
• Compare orphan and source populations for meaningful differences in sample collection/matrix/pre-analytics, device characteristics, patient characteristics, disease characteristics, and other relevant factors.
• Use full extrapolation only when there are no clinically meaningful differences and the external data alone are sufficient in amount and quality.
• Use partial extrapolation when external data are relevant but must be combined with additional clinical/non-clinical evidence using a justified method.
• Document the extrapolation rationale, data-quality appraisal, limitations, model assumptions, and conclusion in the Performance Evaluation Report (PER).

8. Establish PMPF and PMS before market access

• Link each accepted pre-market clinical evidence limitation to a specific PMPF objective and activity.
• Specify the post-market data to be collected, including PMPF studies, registry data, routine-use data, and other real-world evidence.
• Set realistic annual projections for tested persons/specimens and define periodic analysis points and milestones.
• Explain how each activity will address the identified performance or safety uncertainty in a timely manner.
• Establish an orphan IVD-specific risk-management plan describing how risks will be monitored, assessed, mitigated, and acted upon if new concerns arise.
• Include long-term follow-up where the true clinical status or clinical benefit emerges over time, such as for predictive or prognostic devices.
• Maintain a PMS system that evaluates post-market clinical-performance data throughout the device lifecycle.

9. Inform users adequately

• Disclose the device’s orphan status and relevant limitations in pre-market clinical performance data.
• Provide clear instructions on reporting incidents, complaints, and other clinical experience to the manufacturer.
• Include this information, as appropriate, in the label, IFU, summary of safety and performance for class C/D devices, and/or accompanying documentation.
• Ensure communications are consistent with the intended purpose, performance claims, known limitations, and risk controls.

10. Complete the Performance Evaluation Report

• Summarise the orphan-status rationale, including epidemiology, target population, alternatives, state of the art, and expected clinical benefit.
• Summarise all evaluated non-clinical and clinical evidence.
• Describe identified clinical-performance limitations, residual risks, and the process used to identify them.
• Justify why limitations are acceptable: all data assessed; relevant GSPRs met; benefit-risk acceptable; expected clinical benefit supported; further pre-market data infeasible/disproportionate.
• Include or reference a detailed PMPF plan, risk-management summary, data-generation methods, timing, annual testing projections, and milestones.
• Record expert-panel interactions and how advice was considered, if applicable.
• Define PER update intervals and require updates when new information could affect benefit-risk.

11. Prepare for notified-body review and lifecycle oversight

• Submit clear orphan-status and clinical-evidence justification early in the conformity-assessment process.
• Be prepared for notified-body conditions requiring defined PMPF/PMS activities, milestones, or PER updates.
• Track fulfilment of all certificate conditions; failure to meet binding conditions can affect certificate validity and may lead to suspension or withdrawal.
• Keep post-market PER updates current with orphan-status changes, sales volumes, changes in alternatives/state of the art, new clinical evidence, benefit-risk conclusions, and PMPF progress.
• For class D orphan IVDs, ensure the PER reflects relevant orphan-status or clinical-evidence advice that may be considered in the Performance Evaluation Consultation Procedure.

12. Consider expert-panel advice where uncertainty is material

• Consider early voluntary expert-panel advice on orphan status and the clinical-development strategy before clinical evaluation or studies.
• Document how the advice was addressed in the PER.
• Where a notified body is involved, coordinate timing so advice can inform the assessment without duplicating or disrupting it.

Applicability note: For in-house orphan IVDs used by EU health institutions, the clinical-evidence principles, GSPR demonstration, extrapolation approach, and post-market evidence concepts remain relevant; notified-body certification and related procedural requirements do not apply under the Article 5(5) exemption.

Reference: Medical Device Coordination Group. MDCG 2026-6. Clinical evidence for orphan in vitro diagnostic medical devices. November 2026. Published 3 November 2026. Nonbinding guidance. Sections 4–9 and Annexes 2–4.

EUDAMED Summaries and Document Control

MDCG 2026-4 rev.1, Management of SS(C)P in EUDAMED after mandatory use, was published on 30 October 2026. SSCP refers to the summary of safety and clinical performance for relevant medical devices; SSP refers to the summary of safety and performance for relevant IVDs.

This nonbinding position paper describes the planned phased transfer of responsibility for uploading SSCPs/SSPs in EUDAMED from notified bodies to manufacturers, with key deadlines through March 2027.

The paper plans manufacturer-upload functionality in EUDAMED Production for December 2026. Until the functionality is fully available, notified bodies continue uploading master summaries with relevant certificate registrations and updates. The manufacturer must ensure that the uploaded summary matches the version validated by the notified body. For devices placed on the market before mandatory use of the UDI/Devices module, the paper recommends upload as soon as possible and no later than 27 March 2027. That recommendation concerns the specified transition cohort.

Main changes

• Manufacturers will become responsible for uploading the master SSCP/SSP and its translations in EUDAMED.
• They must ensure that the uploaded master document is the version validated by the notified body during certification.
• Notified bodies will identify the validated SSCP/SSP (or relevant parts) by selecting the applicable Basic UDI-DI when registering or linking the certificate.
• The same approach applies to SSPs for IVDs.

Implications for Devices already on the market

• Because manufacturers will not be able to upload SSCPs/SSPs in Production until December 2026, the practical time available is shorter than the original six-month registration transition.
• Notified bodies have until 27 May 2027 to upload relevant certificates.
• Consequently, some device records may initially lack an associated certificate and SSCP/SSP.
• For devices placed on the market before mandatory use of the UDI/Devices module, manufacturers should upload SSCPs/SSPs as soon as possible and no later than 27 March 2027, coordinating timing with their notified body.

Manufacturers can prepare by agreeing on a document handover process with their notified body. A useful internal register would identify the master summary, its revision, the validation record, and the related device identifiers. Translation tracking can sit alongside that record so teams can determine which language versions correspond to a particular master. Assigning a named owner for the handover reduces uncertainty about who checks the file before publication.

A practical example is a PER update that changes a performance estimate or adds a limitation. The regulatory team can assess the consequences for the SSP, identify any validation steps, and track the version intended for upload. Quality personnel can then reconcile the publication record with the controlled document. This proposed workflow helps prevent a revised evaluation and an older public summary from presenting inconsistent information.

Regulatory professionals should also distinguish operational readiness from evidence readiness. A file can be technically ready to upload while its claims still need review. Before release, teams can reconcile the summary with the underlying evaluation and approved labeling, making sure that qualifications and residual limitations have been carried through accurately. Planned deployment dates should be monitored rather than assumed to confirm that a function is already available.

The highest compliance risks are failing to upload legacy-device SS(C)Ps by 27 March 2027 and uploading a version that was not validated by the notified body.

Risk-ranked SS(C)P requirements:

1. Critical — Meet the legacy-device deadline

• • For devices placed on the market before mandatory use of the UDI/Devices module, upload the SSCP/SSP as soon as possible and no later than 27 March 2027.
  • Why this is critical: manufacturers only gain Production upload capability in December 2026, substantially reducing the effective registration window.

2. Critical — Upload only the notified-body-validated master document

• • Ensure the master SSCP/SSP submitted in EUDAMED is exactly the version validated during certification.
  • Why this is critical: the manufacturer is expressly responsible for this alignment; a mismatch undermines the validity and traceability of the public document.

3. High — Maintain certificate and Basic UDI-DI linkage

• • Ensure the notified body identifies the validated SSCP/SSP, or relevant section, through the appropriate Basic UDI-DI when registering or linking the certificate.
  • Why this is high risk: an incorrect or missing linkage can prevent clear association between the device, its certificate, and its validated SS(C)P.

4. High — Follow the transitional responsibility split from 28 May 2026

• • Until the new Production functions are available, ensure the notified body uploads the master SS(C)P when registering new certificates and updates for affected devices.
  • Why this is high risk: responsibility changes by system phase; assuming the manufacturer can upload in Production before December 2026 risks an absent SS(C)P.

5. High — Coordinate timing with the notified body for pre-mandatory-use devices

• • Align registration schedules because notified bodies have until 27 May 2027 to upload corresponding certificates.
  • Why this is high risk: a device may be registered before its related certificate and SS(C)P appear in EUDAMED, creating temporary traceability and publication gaps.

6. Medium — Manage translations under the applicable MDCG guidance

• • During certificate registration in the transition, translations are not uploaded with the master SS(C)P; the manufacturer remains responsible for their management under MDCG 2021-1 rev.1 and MDCG 2022-12.
  • Why this is medium risk: treating translations as part of the notified body’s certificate-registration upload can leave translation obligations unmanaged.

7. Medium — Test the revised workflow in EUDAMED Playground

• • Use the planned July 2026 Playground deployment to test roles, uploads, and linkage controls before the planned December 2026 Production deployment.
  • Why this is medium risk: the document frames this interval as the period for adapting internal processes and ensuring compliance; insufficient testing increases operational-error risk.

Priority control focus

The immediate control priority is a reconciled inventory showing, for every applicable device: the Basic UDI-DI, certificate status, notified-body-validated master SS(C)P, translation status, EUDAMED upload owner, and applicable deadline. This directly addresses the document’s main risks of late upload, wrong-version upload, and missing certificate linkage.

Use this checklist to prepare for the transfer of SS(C)P upload responsibilities in EUDAMED. It applies to SSCPs under the MDR and SSPs for IVDs under the IVDR.

Governance and document control

• Identify every device and relevant Basic UDI-DI that requires an SSCP/SSP.
• Maintain a controlled master SSCP/SSP for each applicable device or device group.
• Confirm that each master document uploaded to EUDAMED is the exact version validated by the notified body during certification.
• Establish version-control records linking the master SSCP/SSP, validation outcome, certificate, and relevant Basic UDI-DI.
• Manage SSCP/SSP translations under the applicable MDCG translation guidance; translations are the manufacturer’s responsibility.

Responsibilities with the notified body

• Agree a documented handoff process with the notified body for SSCP/SSP validation and certificate registration.
• Confirm that the notified body identifies the validated SSCP/SSP, or relevant part, by selecting the applicable Basic UDI-DI when registering or linking the certificate.
• Align timelines with the notified body, especially for legacy devices whose certificate may not yet be registered in EUDAMED.

Transitional period: 28 May 2026 to Production release

• For new certificates and certificate updates for affected devices, verify that the notified body uploads the master SSCP/SSP with certificate registration.[ ] Do not treat translation upload as part of certificate registration during this period; the transitional upload covers the master document only.
• Update internal procedures to reflect the temporary split: notified body uploads the master SS(C)P, while the manufacturer manages translations.
• Use the EUDAMED Playground release, planned for July 2026, to test access, roles, document controls, and upload workflows.

Manufacturer upload readiness: Production release

• Prepare manufacturer user roles, internal approvals, and evidence retention for direct SSCP/SSP upload in EUDAMED Production, planned for December 2026.
• Test that the upload process preserves the link between the uploaded document, the notified-body-validated version, certificate, and Basic UDI-DI.
• Train regulatory and quality personnel on the revised workflow before Production deployment.

Devices placed on the market before mandatory UDI/Devices-module use

• Create a separate inventory of legacy devices requiring SSCP/SSP registration.
• Prioritize devices whose certificates may be uploaded late during the notified body transition period, which runs until 27 May 2027.
• Upload SSCPs/SSPs as soon as EUDAMED Production permits manufacturer uploads.
• Complete uploads for these devices no later than 27 March 2027.
• Reconcile legacy-device records against certificate registration status, because some devices may initially appear without a corresponding certificate or SS(C)P.

Evidence to retain

• Notified body validation evidence for each master SSCP/SSP.
• Upload confirmation and date from EUDAMED.
• Certificate and Basic UDI-DI linkage records.
• Translation-management records.
• Records of coordination and agreed timelines with the notified body.

The key compliance control is ensuring that the manufacturer-uploaded master SSCP/SSP always matches the version validated by the notified body, while meeting the transition dates and legacy-device deadline.

Reference: Medical Device Coordination Group. MDCG 2026-4 rev.1. Management of SS(C)P in EUDAMED after mandatory use. October 2026. Published 30 October 2026. Nonbinding position paper.

Across these two topics, a useful immediate exercise is to review one active product from its intended purpose through the supporting studies to the public summary. Record where evidence is limited, where a study does not fully represent clinical use, and where a document revision needs to be carried into another record. Those findings can become specific actions in the evidence plan and document-control process, with an owner and a review date.