Comprehensive Guide to IVDR Performance Evaluation & Performance Evidence Requirements

What a Defensible Performance Evaluation Report Looks Like?

A defensible Performance Evaluation Report (PER) under EU IVDR is not merely a compilation of studies. It is a structured, scientifically reasoned narrative that demonstrates performance, justifies evidence sufficiency, and explains how performance is monitored and maintained over time.

A high-quality IVDR Performance Evaluation Report allows a reviewer to answer, with minimal interpretation:

• What the device is intended to do and in which clinical context it is used?
• What is the scientific rationale supporting the analyte- clinical `condition relationship?
• How reliably does the device performs analytically under its intended conditions of use?
• How effectively does the device demonstrates clinical performance in the intended population?
• How does evidence support each claim, and what limitations exist?
• Why is the evidence sufficient for device’s intended purpose and risk classification?
• How post-market surveillance and post-market performance follow-up will sustain confidence in performance over time?

Intended purpose as the “lens” for all evidence

One of the most common reasons why performance evidence becomes difficult to defend is that intended purpose is described at a high level, while studies reflect a narrower or different context. For example, a test used in a symptomatic population behaves differently than the same test used in screening settings. Variations in disease prevalence, patient spectrum, predictive values, and reference standard may not behave the same way. A defensible PER anticipates these issues by explicitly demonstrating intended use and ensuring that supporting evidence aligns with that use.

Scientific validity as a structured argument

Scientific validity is often presented as a collection of references, but under the IVDR it is expected to function as a structured and reasoned argument. It should demonstrate that the analyte is meaningfully associated with the relevant clinical condition and that this association supports the device’s stated intended purpose.The strongest scientific validity sections rely on high-quality sources such as clinical guidelines, consensus statements, and peer-reviewed syntheses, rather than relying solely on individual studies. They also clarify relevance: why the evidence applies to the intended population and use.

When scientific validity is weak or generic, Notified Bodies often ask whether the device’s intended purpose is overclaimed relative to the available science. That is why it is useful to anchor scientific validity reasoning to recognized principles such as those described in MDCG 2022-2.

Analytical performance as reliability in real-world conditions

Analytical performance is where performance is translated into measurable and repeatable outcomes. It is also the point at which study design must realistically reflect routine use conditions, including variability in specimen collection and handling, environmental factors, operator-dependent effects (where applicable), lot-to-lot variation, potential interfering substances, and instrument drift over time. A defensible PER describes not only results but the logic connecting analytical designs to intended use conditions.

Analytical sections are also stronger when they connect to risk management and labelling. If certain limitations exist, for example, reduced performance in the presence of an interferent, defensibility improves when the PER explains how that limitation is managed (warnings, IFU instructions, acceptance criteria, QC measures).

Clinical performance as the clinical credibility test

Clinical performance is the most scrutinized pillar for many IVDs because it determines whether a test result meaningfully correlates with a clinical condition or outcome in its intended context. The core challenge is that clinical performance can be inflated if study design introduces bias or if reference standards are imperfectly applied. This is why bias control, endpoint selection, and reference standard justification are central to defensible clinical evidence.

When clinical performance evidence involves diagnostic accuracy, a useful lens is QUADAS-2, a commonly used tool for assessing risk of bias and applicability in diagnostic accuracy studies. QUADAS-2 does not focus on whether results are favorable, but on whether they are credible. A reliable overview is available via QUADAS-2 (University of Bristol). Manufacturers do not need to treat QUADAS-2 as a regulatory requirement but understanding it helps align study reporting with how reviewers think about credibility.

These considerations become especially important when translating clinical performance results into defensible regulatory conclusions, because design choices directly shape how reviewers interpret credibility and applicability. A structured approach to bias control, endpoint selection, reference standard justification, and applicability assessment, such as the methodology described in designing a clinical performance study, helps ensure that clinical performance evidence is not only statistically sound but also clinically meaningful and review-ready under IVDR.

Designing Clinical Performance Evidence That Holds Up Under Review

Clinical performance evidence becomes defensible when it is clearly tied to the clinical pathway and the decision that the test informs. A recurring problem in performance dossiers is that evidence is presented without clarifying how the result is used. A screening test has different evidentiary needs than a triage test; a monitoring test has different needs than a confirmatory diagnostic test.

A defensible approach starts by clarifying the role of the test:

• What clinical decision does it inform?
• What happens if the result is positive or negative?
• What are the consequences of false positives and false negatives?
• What patient groups are likely to be tested in practice?

When these questions are answered, performance endpoints become easier to justify. Diagnostic accuracy metrics are not just statistics, they represent trade-offs. Sensitivity may be prioritized when missing disease has significant consequences. Specificity may be prioritized when unnecessary intervention has significant consequences. Predictive values become central in low-prevalence settings, and stratified performance analysis may be warranted for key subgroups if clinical practice indicates meaningful differences.

Bias control is also central. If patient selection disproportionately includes advanced cases, performance may appear artificially high. If reference standards differ across sites, results can be inconsistent. If the index test is interpreted with knowledge of the reference standard, bias can be introduced. A defensible evidence narrative acknowledges these risks and explains how study designs mitigate them.

Evidence Sufficiency Under IVDR: Turning “We Have Data” into “This Is Enough”

One of the most difficult challenges in IVDR Performance Evaluation is explaining evidence sufficiency. Manufacturers often feel that extensive data should be inherently persuasive, but Notified Bodies evaluate sufficiency through a different lens: alignment, relevance, quality, and reasoning.

Evidence sufficiency is judged in context. Factors that often influence sufficiency expectations include:

• Risk class and potential harm from incorrect results
• Intended purpose and the clinical impact of decisions based on results
• Complexity of the test and its interpretive pathway
• Availability of alternative methods and standards of care
• Novelty of the analyte or intended use claims
• The breadth of populations and use settings claimed

This is where MDCG 2022-2 is especially helpful, because it reinforces the idea that evidence sufficiency must be justified and maintained as a continuous process. It also implicitly encourages manufacturers to be transparent about limitations and to manage uncertainty through structured post-market mechanisms.

A defensible PER typically includes an explicit “evidence sufficiency justification” narrative. That narrative explains not only what evidence exists, but why it is relevant and robust for the intended purpose. It acknowledges what the evidence does not cover, and it explains how the remaining uncertainty is managed through targeted PMPF and post-market surveillance.

In many reviews, evidence sufficiency becomes far less contentious when it is argued transparently and grounded in clear regulatory logic. When claims are scoped appropriately, endpoints match the intended clinical use, limitations are acknowledged, and post-market learning is integrated into the evidence system, the sufficiency rationale reads as coherent, defensible, and proportionate to the device’s intended purpose and risk.

Post-Market Surveillance and PMPF: Keeping Performance Defensible Over Time

A defining feature of IVDR is its explicit focus on lifecycle evidence. Devices are not evaluated only at market entry; they are expected to remain safe and performant as clinical practice, populations, and real-world contexts evolve. This is why post-market surveillance is not a standalone compliance task. It is part of the evidence system that keeps performance defensible.

Post-market learning can reveal issues that pre-market studies cannot fully predict such as changes in prevalence, differences across patient populations, unexpected interferents, user behavior variability, lot-to-lot performance changes, and operational factors in routine labs. A mature evidence system anticipates this and uses post-market mechanisms to detect and interpret these signals.

IVDR’s post-market framework includes structured PMS planning and, where appropriate, PMPF activities. PMPF is not simply “more studies.” It is a targeted effort to answer specific performance questions that remain after pre-market evaluation, questions that may relate to clinical subgroups, real-world workflows, long-term trends, or use expansion.

The strongest lifecycle evidence systems show how post-market data feeds back into performance evaluation. That includes how signals are detected, how trends are assessed, and how performance conclusions are updated in the IVDR performance evaluation report. This lifecycle integration is at the heart of IVDR PMS, PMPF, and PSUR, where evidence traceability becomes a practical governance model rather than a theoretical ideal.

Traceability: The Feature That Makes a PER Reviewable and AI-Extractable

Traceability is one of the most important “quiet success factors” under IVDR. It determines whether a reviewer can follow your reasoning without ambiguity. It also influences how well AI search tools can extract and summarize your content: clear relationships between claims, endpoints, evidence, and conclusions increase “answer readiness.”

A defensible Performance Evaluation Report (PER) typically makes traceability visible in the narrative, not only in tables. It explains:

• The intended purpose and the clinical pathway
• The claims derived from intended purpose
• The endpoints and acceptance criteria derived from claims
• The evidence sources selected to support each endpoint
• The appraisal logic used to weigh evidence
• The conclusions and limitations derived from that appraisal
• The post-market mechanisms used to monitor and update performance confidence

This approach reduces review friction because it prevents “gaps” a reviewer must fill in. It also improves internal governance: when evidence is traceable, updates become manageable. You can identify exactly which claim and endpoint are affected by new information and adjust evidence sufficiency rationales accordingly.

Organizations that align evidence traceability with broader clinical and performance evaluation practices often find that performance evidence becomes more sustainable over time. The key is treating traceability as a design feature of the evidence system, not as an after-the-fact documentation activity.

Common Notified Body Questions and How Defensible PERs Pre-empt Them

Notified Bodies often ask questions that cluster around a few predictable areas. A defensible performance evidence system anticipates these areas and addresses them proactively.

One common question is whether study populations represent intended use. Evidence that looks strong in a narrow population may not justify broad claims. Another common question is whether the reference standard is appropriate and consistently applied, inconsistent reference standards can undermine clinical performance credibility even when results appear positive. Reviewers also frequently ask whether endpoints reflect clinical utility i.e., does the performance measure actually matter in the intended clinical decision pathway?

Literature use is another area. Notified Bodies often look for critical appraisal, not just citation. They want to see that literature is relevant, that limitations are understood, and that conclusions are not overstated. This is aligned with the evidence appraisal principles emphasized in MDCG 2022-2.

Finally, post-market planning is increasingly scrutinized. If uncertainty exists (and it almost always does), reviewers expect to see a proportionate plan for how post-market surveillance and PMPF will reduce uncertainty and confirm performance over time. When post-market evidence governance is coherent, it strengthens the entire PER because it demonstrates that performance confidence is continuously maintained, not assumed.

Conclusion

The EU IVDR framework reframes performance evaluation as a continuous discipline, not a one-time documentation exercise. Manufacturers are expected to demonstrate that performance claims are scientifically grounded, analytically reliable, and clinically meaningful, and that these claims remain defensible as scientific knowledge evolves and real-world evidence accumulates.

A defensible IVDR Performance Evaluation approach is built on clear planning, coherent evidence logic, and transparent traceability. The Performance Evaluation Plan establishes the methodological foundation that links intended purpose to claims, endpoints, evidence sources, and appraisal methods. The IVDR Performance Evaluation Report then becomes the narrative that demonstrates performance, justifies evidence sufficiency, acknowledges limitations, and explains how residual uncertainty is managed over time.

Post-market learning is the final ingredient that makes performance evidence durable. Robust post-market surveillance and targeted PMPF activities turn performance evaluation into a living system, one that confirms performance in routine use, detects meaningful trends, and supports appropriate updates to evidence and conclusions when needed. When these elements are integrated into a coherent lifecycle evidence strategy, performance evaluation becomes more than compliance: it becomes a durable foundation for trust in diagnostic results and confidence in clinical decision-making.

How Freyr Can Help

Freyr supports IVD manufacturers with end-to-end Performance Evaluation, from defining intended purpose and strengthening Performance Evaluation Plan development to creating defensible Performance Evaluation Reports aligned with Notified Body expectations. Additionally, Freyr helps manufacturers in developing MDR-aligned Clinical Evaluation Reports (CERs) and lifecycle-based clinical evidence systems that stand up to Notified Body scrutiny.

If you need support generating performance evidence, designing or appraising studies, closing evidence gaps, or building a traceable lifecycle evidence system integrating post-market surveillance, PMPF, and PSUR obligations, talk to our experts for guidance across every stage of the IVDR evidence lifecycle.