Under the EU In Vitro Diagnostic Regulation (IVDR) 2017/746, the Scientific Validity Report is not a formality, it is the starting point of the entire clinical evidence framework. Before any IVD can demonstrate analytical or clinical performance, it must first establish that the analyte it measures has a meaningful association with a clinical condition or physiological state. This requirement is formalised in the Scientific Validity Report (SVR), a core component of the Performance Evaluation Report (PER).
While the regulatory definition of scientific validity appears concise, the association of an analyte with a clinical condition or physiological state, as defined in Article 2(38) of the IVDR, the practical implementation remains one of the most frequently misunderstood aspects of IVDR compliance. What has changed compared to the IVDD is not the concept itself, but the level of structure, transparency, and scientific reasoning now required to demonstrate it.
This article examines what the SVR requires under IVDR, the most common deficiencies identified during Notified Body review, and how manufacturers can construct a report that is both compliant and robust, based on real experience reviewing and preparing SVRs across IVD device classes.
What Is a Scientific Validity Report Under IVDR?
The SVR is one of three components of the Performance Evaluation under IVDR, alongside the Analytical Performance Report (APR) and, where applicable, the Clinical Performance Report (CPR). Together, these feed into the Performance Evaluation Report (PER), which manufacturers must submit to their Notified Body as part of conformity assessment.
Under Annex XIII of the IVDR, performance evaluation must include a systematic and documented process for identifying, appraising, and analysing data on the scientific validity, analytical performance, and clinical performance of the device. The SVR addresses the first of these elements.
In practical terms, the SVR must demonstrate:
- That the analyte measured by the device has a biologically plausible and clinically documented association with the condition or physiological state stated in the intended purpose
- That this association is consistent with the current state of the art
- That available evidence has been systematically identified, appraised, and analysed, including both favourable and unfavourable data
- That any limitations or gaps in the evidence are acknowledged and addressed
The SVR does not assess how well the device detects the analyte, that is the role of the APR. Instead, it answers a more fundamental question: does the analyte matter clinically for the purpose the manufacturer is claiming?
Where Scientific Validity Reports Go Wrong: The Most Common Deficiency
“The most common issue we see in SVRs is not a misunderstanding of the definition of scientific validity itself, but a failure to translate that definition into a clear and defensible argument. Many SVRs end up as collections of literature summaries rather than structured scientific justifications.”
Expert insight by Catarina Sepulveda | Regulatory Director | IVD Team
This distinction matters because the IVDR requires performance evaluation to go beyond describing evidence, it requires assessing and interpreting that evidence in a way that supports conformity with the General Safety and Performance Requirements. Simply presenting literature does not demonstrate scientific validity. The SVR must construct a logical, traceable argument showing how the evidence supports the device’s intended use.
A second recurring deficiency is the weak link between the evidence and the intended purpose. Literature may support a biomarker in a general sense, but if the SVR does not demonstrate relevance to the specific clinical context, population, and use scenario defined by the manufacturer, the argument remains incomplete.
Finally, many SVRs avoid addressing contradictory or weaker evidence. The IVDR is explicit: both favourable and unfavourable data must be considered and appraised. An SVR that acknowledges limitations, conflicting findings, and uncertainty, and explains how these are addressed, is far more convincing than one that presents a simplified or one-sided view.
Class C vs Class D: Less Difference Than Expected
The IVDR establishes that the depth of performance evaluation must be proportionate to the risk class and intended purpose of the device.
“In practice, however, the distinction is less pronounced than many manufacturers expect. The core expectations, a systematic literature review, transparent methodology, critical appraisal of evidence, and a clear scientific argument, are effectively the same for both classes. What changes is not the structure of the SVR, but the level of scrutiny applied to uncertainty.”
Expert insight | Catarina Sepulveda
For Class D devices, which typically include tests with high public health impact, Notified Bodies expect a very robust evidence base with minimal residual uncertainty. Any gaps must be thoroughly justified and, in many cases, supplemented with additional data.
For Class C devices, there may be slightly more flexibility, but this is often limited, particularly when the clinical implications are significant. A well-prepared Class C SVR often closely resembles a Class D SVR in structure and depth.
Attempting to rely on a ‘minimum viable SVR’ for Class C devices is a risky strategy. Building a robust, well-justified scientific validity argument from the outset is far more effective than relying on perceived differences in risk-class expectations.
Building Scientific Validity When Evidence Is Limited
Novel biomarkers and emerging diagnostic approaches present some of the most challenging scenarios in SVR preparation. Limited or fragmented evidence does not necessarily prevent compliance, but the strategy for building the scientific validity argument must be carefully constructed.
The most effective approach is to expand the concept of evidence beyond direct, high-level clinical studies. When direct literature is limited, it is often possible to build a scientifically coherent argument through a combination of:
- Biological plausibility and mechanistic evidence
- Proof-of-concept or exploratory studies
- Evidence from related biomarkers or pathways
- Expert guidelines, position statements, and consensus documents
“Another key strategy is recognising that the boundaries between scientific validity and clinical performance are not always rigid. Clinical performance data may indirectly support scientific validity by demonstrating that the measurement of the analyte leads to clinically meaningful differentiation between patient groups. While this does not replace the need for an SVR, it can significantly strengthen the overall evidence package.”
Expert insight | Catarina Sepulveda
Perhaps most importantly, transparency is essential. Attempting to present limited evidence as stronger than it is almost always leads to deficiencies during conformity assessment. A better approach is to clearly identify gaps, explain their implications, and outline how they will be addressed, for example, through post-market performance follow-up (PMPF) or targeted studies. This aligns with the IVDR principle that performance evaluation is a continuous lifecycle process.
What an Acceptable Literature Search Looks Like in Practice
An acceptable literature search under IVDR is not defined by the number of publications retrieved, but by the transparency, reproducibility, and justification of the methodology used to identify and select evidence.
Start with a Formal Literature Search Protocol
The process begins with a clearly defined scope using structured frameworks such as PICO or, more appropriately for diagnostics, the PIT framework (Population, Index test, Target condition). These ensure that the search strategy is aligned with the intended use of the device, not a generic exploration of the wider literature.
A formal Literature Search Protocol (LSP) must be defined before the search is conducted. This protocol specifies the databases, keywords, Boolean strategies, and inclusion/exclusion criteria upfront. The absence of an LSP is one of the fastest ways to generate a deficiency, because it raises immediate concerns about bias and reproducibility.
Databases, Selection, and PRISMA
In practice, Notified Bodies expect the use of recognised biomedical database, typically PubMed or MEDLINE, Embase, and Cochrane, supplemented by guideline repositories and relevant professional society publications. What matters is not only which databases are used, but how well the search strategy is adapted to each one. Running the same keywords across multiple platforms without adjusting syntax or indexing terms is often considered insufficient.
The study selection process must be fully traceable. Manufacturers often present a final list of included papers without showing how those papers were selected. In practice, Notified Bodies expect a clear and documented selection process, typically aligned with PRISMA principles, including a flow diagram showing records identified, screened, excluded, and included at each stage.
Evidence Appraisal Is Non-Negotiable
“What ultimately convinces a reviewer is not the size of the dataset, but the coherence of the reasoning. The literature search must lead to a clear, traceable conclusion: that the available scientific evidence is sufficient, or not, to support the claimed association between the analyte and the clinical condition. If a reviewer cannot follow how the evidence leads to that conclusion, the SVR will not be considered robust, regardless of how many references it contains.”
Expert insight | Catarina Sepulveda
Once studies are selected, evidence appraisal is where many SVRs fail. Notified Bodies expect a critical evaluation of study quality, potential bias, and relevance to the intended use, not simply a summary of results. Under IVDR, it is not enough to list evidence; manufacturers must demonstrate that they have assessed its suitability for the specific clinical claim.
Why SVRs Take Longer Than Expected
SVR timelines are consistently underestimated. While a straightforward report for a well-established single-analyte test may be completed relatively quickly, more complex scenarios extend timelines considerably. The main drivers include:
- Volume and heterogeneity of literature: In areas such as oncology or infectious diseases, the screening and appraisal process can be extremely resource-intensive.
- Novel analytes: Limited or fragmented evidence requires more effort in interpretation, justification, and structuring of the argument.
- Multiplex assays: Each analyte may require its own scientific validity argument, as well as consideration of how analytes interact within the overall intended purpose.
- Poorly defined intended purpose: Changes to scope often require the literature search and analysis to be repeated, significantly impacting timelines.
Establishing a well-defined intended purpose and a clear LSP at the outset, using frameworks such as PICO or PIT, is the most effective way to control timelines. Preparing a high-quality SVR is a scientific and regulatory reasoning process, and the time required reflects the rigour expected under the IVDR.
Where Notified Bodies Are Raising Deficiencies
Deficiencies in SVRs most commonly cluster around three areas: methodology, evidence appraisal, and scientific reasoning.
1. Literature search methodology
Missing or poorly defined search strategies, unclear database selection, and insufficient justification of inclusion/exclusion criteria are common issues. These deficiencies raise concerns about potential bias and completeness of the evidence base.
2. Lack of critical appraisal
Many SVRs summarise evidence but do not evaluate its quality, limitations, or relevance. This creates a disconnect between the data presented and the conclusions drawn. Under IVDR, summarising studies is not the same as appraising them.
3. Weak link between evidence and intended use
Even when the literature is robust, SVRs often fail to explicitly demonstrate how that evidence supports the specific clinical application claimed for the device. This is particularly problematic for devices with narrow or highly specific intended uses.
“Since the publication of MDCG 2022-2, these expectations have become more clearly defined and consistently enforced. The overall trend is a shift from evaluating whether evidence exists, to assessing whether the reasoning linking evidence to claims is scientifically and regulatorily sound.”
Expert insight | Catarina Sepulveda
What Happens When Scientific Validity Cannot Be Established
A conclusion of insufficient evidence is a critical outcome, but not necessarily the end of the regulatory pathway. What matters is how the manufacturer responds.
The realistic options are:
Generate additional data
This is the most direct option for higher-risk devices or novel analytes where the existing evidence base is insufficient. This may involve conducting clinical performance studies or targeted research to establish the missing association.
Refine the intended purpose
Narrowing the clinical claims, adjusting the target population, or modifying the use scenario can reduce the evidence burden and bring the scientific argument into alignment with available data.
Justify residual uncertainty
For lower-risk devices with limited clinical impact, it may be possible to justify remaining uncertainty with a strong, transparent rationale. This is unlikely to be accepted for higher-risk applications.
Reconsider market placement
Where no scientifically valid association can be established, the most realistic outcome may be device redesign or reconsideration of the market claim. Under IVDR, scientific validity is the first pillar of clinical evidence. If it cannot be demonstrated, the entire performance evaluation framework is invalid.
Legacy IVDD Data: Can It Still Be Used?
“Short answer: almost never. A legacy SVR developed under the IVDD is almost never sufficient on its own. The reason lies in the fundamental shift introduced by the IVDR: from a largely declarative approach to a systematic, evidence-based, and fully documented performance evaluation process.”
Expert insight | Catarina Sepulveda
Under the IVDD, requirements for clinical and scientific evidence were relatively limited and often inconsistently applied. Many manufacturers relied on general knowledge of the analyte combined with a small number of references, without performing a structured or reproducible literature review. As a result, most legacy SVRs lack the elements now considered essential under IVDR: systematic search strategies, explicit inclusion/exclusion criteria, and critical appraisal of evidence.
For well-established analytes, legacy literature can still serve as a valuable foundation. However, the methodology must be rebuilt. Evidence must be re-evaluated within a compliant framework, with full transparency and traceability.
In practice, most legacy SVRs require significant rework or complete redevelopment. The most effective strategy is to treat the IVDR SVR as a new document and integrate relevant legacy evidence into a robust, compliant structure.
Top 10 Scientific Validity Report Mistakes
Based on real Notified Body feedback, the following are the ten most common mistakes manufacturers make when preparing an SVR under IVDR:
- Treating the SVR as a literature summary instead of a scientific argument, presenting information without explicitly demonstrating the analyte–disease association.
- Weak or missing link to intended purpose, evidence is generic rather than clearly aligned with the device’s specific clinical context, population, or use scenario.
- Absence of a structured Literature Search Protocol, without a predefined and documented methodology, the review cannot be considered systematic.
- Inadequate search strategy, search terms are too narrow, too broad, or not adapted to each database, leading to incomplete or biased evidence retrieval.
- Poor documentation of study selection, included studies are listed, but the selection process is not traceable or justified.
- No critical appraisal of evidence, studies are summarised, but their quality, limitations, and relevance are not assessed.
- Ignoring contradictory or unfavourable data, presenting only supportive evidence undermines credibility and raises concerns during review.
- Overstating conclusions from weak evidence, limited or indirect data is presented as conclusive without sufficient justification.
- Failure to address evidence gaps, missing or incomplete evidence is not acknowledged, or no mitigation strategy is proposed.
- Reusing IVDD-era documentation without methodological update, legacy reports are often insufficient because they lack systematic, reproducible methodology.
From Literature Compilation to Scientific Justification
A strong Scientific Validity Report is not defined by how much literature it contains, but by how convincingly it answers a single question: does the analyte measured by the device have a clinically meaningful association with the intended condition?
Under the IVDR, the SVR is not just a supporting document, it is the starting point of the entire clinical evidence framework. If scientific validity is weak or poorly justified, analytical and clinical performance cannot compensate for it.
For manufacturers, the key shift is moving from a documentation mindset to a scientific reasoning mindset. The goal is not to collect evidence, but to build a transparent, structured, and defensible argument that can withstand regulatory scrutiny.
That is what ultimately defines a strong and compliant Scientific Validity Report under IVDR.
Frequently Asked Questions about Scientific Validity Report Under IVDR
An SVR is a core component of the Performance Evaluation Report (PER) under EU IVDR 2017/746. It documents the association between the analyte measured by an IVD and a clinical condition or physiological state, demonstrating that the measurement is clinically meaningful for the device’s intended purpose.
Yes. All IVD devices subject to IVDR require a performance evaluation, which must include demonstration of scientific validity under Annex XIII. The depth and extent of the SVR must be proportionate to the risk class and intended purpose of the device.
Scientific validity addresses whether the analyte is clinically relevant for the intended purpose. Analytical performance addresses how well the device detects or measures that analyte. Both are required components of performance evaluation, but scientific validity must be established first.
Rarely, if ever, in its original form. Legacy documentation typically lacks the systematic methodology (structured literature search, defined protocols, critical appraisal) now required under IVDR. Most legacy SVRs need significant rework or complete redevelopment.
A finding of insufficient evidence is not the end of the regulatory pathway. Options include generating additional data, refining the intended purpose to reduce the evidence burden, or justifying residual uncertainty for lower-risk applications. For higher-risk devices, additional clinical performance studies are typically required.
MDCG 2022-2 (Guidance on general principles of clinical evidence for In Vitro Diagnostic medical devices) is the primary guidance document. It clarifies the requirements for scientific validity, evidence appraisal, and the structure of the literature review within the performance evaluation framework.
How MDx Supports Scientific Validity Reports
MDx CRO prepares Scientific Validity Reports for IVD manufacturers across all device classes, combining regulatory expertise with structured literature methodology to deliver SVRs that withstand Notified Body scrutiny.
