IVD Performance Evaluation Under the IVDR

Performance evaluation is the evidence backbone of every IVDR technical file. Under Article 56 and Annex XIII of Regulation (EU) 2017/746, IVD manufacturers must show, with data, that their device does what they claim it does. MedDeviceCorp helps IVD manufacturers plan, generate and document that evidence so it stands up to notified body review.

Get a Performance Evaluation Quote

What Is IVD Performance Evaluation?

IVD performance evaluation is a continuous process of assessing and analysing data to establish or verify a device's scientific validity, analytical performance and clinical performance. It is not a one-time report. It starts at design, supports CE marking, and is updated throughout the device's life with post-market data.

The Three Pillars of Performance Evidence

Scientific Validity

Scientific validity shows that the analyte or marker is associated with a clinical condition or physiological state. Evidence typically comes from peer-reviewed literature, consensus guidelines, or proof-of-concept studies.

Analytical Performance

Analytical performance shows that the device correctly detects or measures the analyte. This covers parameters such as analytical sensitivity and specificity, trueness, precision, accuracy, limits of detection and quantitation, measuring range, linearity and interference.

Clinical Performance

Clinical performance shows that the device produces results that correlate with the clinical condition in the intended population and user. Typical parameters include diagnostic sensitivity and specificity, positive and negative predictive values, and expected values in normal and affected populations.

Master Structure of the Analytical Performance Report (APR)

The Analytical Performance Report brings together all analytical performance data for the device and feeds directly into the Performance Evaluation Report. A well-structured APR makes it easy for the notified body to trace every claim back to its evidence. We build APRs around the following master structure:

Scope & Administrative Identifiers (Device name, UDI-DI, formulation, software version) – defines exactly which device, formulation and software version the report covers.
Traceability Matrix to the Performance Evaluation Plan (PEP) & GSPR Annex I (9.1a) – links each analytical claim to the PEP and to the general safety and performance requirements.
Reference Materials, Calibrators & Metrological Traceability – the calibration hierarchy and the reference materials or methods used.
Analytical Sensitivity: LoB, LoD, and LoQ – limit of blank, limit of detection and limit of quantitation.
Precision (Repeatability, Reproducibility, Within-Laboratory) – variability across runs, days, sites, lots and operators.
Trueness, Accuracy, and Method Comparison – agreement with a reference method or comparator device.
Analytical Specificity (Cross-Reactivity & Exogenous/Endogenous Interference) – the effect of related analytes and interfering substances on results.
Linearity, Measuring Range, and High-Dose Hook Effect – the reportable range and behaviour at very high analyte levels.
Specimen Handling, Collection Matrices, and Pre-Analytical Stability – specimen types, collection tubes, transport and storage conditions.
System Robustness, Carryover, and Reagent Open/Closed-Vial Stability – performance under varied conditions, sample-to-sample carryover and reagent shelf life.
Statistical Protocols, Deviations, and Notified Body Synthesis – statistical methods and acceptance criteria, study deviations, and a summary conclusion for the notified body.

Studies are typically designed using recognised standards such as the CLSI EP guidelines and ISO 17511 for metrological traceability.

CLSI-Aligned Audit Checklist by Parameter

Notified bodies review analytical performance studies against recognised CLSI and ISO methods. The checklist below shows, for each parameter, the applicable standard, what reviewers check and the non-conformities we see most often.

Parameter	Applicable Standard	Key Notified Body Review Checkpoints	Common Audit Non-Conformities
Detection Limits (LoB, LoD, LoQ)	CLSI EP17-A2	Distinct calculation of Limit of Blank (zero-analyte), Limit of Detection (parametric vs non-parametric) and Limit of Quantitation Minimum of 2 reagent lots and 3–5 sample days LoQ verified against a clinically acceptable total error (%TE) or coefficient of variation (%CV) budget	Conflating LoB and LoD Failing to state the target %TE at the functional sensitivity/LoQ limit Using synthetic buffer instead of natural target matrix
Precision	CLSI EP05-A3	Standard 20-day evaluation (2 runs per day, 2 replicates per run; 20 × 2 × 2) or multi-site reproducibility (3 × 5 × 5) Breakdown of variance components: within-run (repeatability), between-run, between-day and total within-laboratory Testing covers clinical decision cut-offs (medical decision levels)	Conducting only a 1-day or 3-day repeatability study Omitting 95% confidence intervals around the SD and %CV Testing only at high-positive levels
Trueness & Method Comparison	CLSI EP09c CLSI EP29-A	Comparison against an accepted Reference Measurement Procedure (RMP) or designated predicate with verified performance Minimum 100–140 patient samples evenly distributed across the measuring interval Deming or Passing-Bablok regression with slope, intercept and 95% CI (avoiding ordinary least squares when there is measurement error on both axes)	Relying on Pearson's r correlation coefficient instead of Deming/Passing-Bablok bias estimates Sample spiking without proving matrix equivalence
Analytical Specificity & Cross-Reactivity	CLSI EP07-A3 CLSI EP37	Evaluation of common structurally similar pathogens, homologous analytes and related proteins Tested at high concentrations (e.g. 10⁶ CFU/mL for bacteria, 10⁵ PFU/mL for viruses, or 10–100× therapeutic ranges) Both zero-analyte (false-positive check) and low-positive analyte (interference check) tested	Incomplete pathogen panel omitting regional variants Not evaluating cross-reactivity at clinically relevant high concentrations
Endogenous & Exogenous Interference	CLSI EP07-A3 CLSI EP37	Paired-difference testing vs baseline control samples Endogenous: haemoglobin, conjugated/unconjugated bilirubin, triglycerides/lipaemia, albumin, total protein, rheumatoid factor, HAMA Exogenous: common OTC medications, disease-specific therapeutics, collection tube additives, anticoagulant types	Testing only endogenous interferents while ignoring common prescription/OTC drugs No predefined cut-off for "significant interference" (e.g. bias exceeding ±10%)
Linearity & Measuring Range	CLSI EP06-Ed2	Polynomial evaluation (linear vs 2nd-order vs 3rd-order fit) At least 5–7 dilutions/admixtures covering the full claimed analytical measuring range (AMR) Identification of lower and upper limits of quantitation (LLoQ, ULoQ)	Visual assessment without formal statistical polynomial regression Extending the claimed range beyond validated linear dilution points
High-Dose Hook Effect	CLSI EP06 / assay-specific	Crucial for sandwich/one-step immunometric assays Testing supraphysiological analyte concentrations (e.g. 10×–1000× upper normal limit) to confirm the signal does not drop back below the positive threshold	Omission of prozone testing where assay architecture is theoretically vulnerable to analyte excess
Matrix Comparison & Anticoagulants	CLSI EP35	Direct evaluation across every claimed specimen matrix (serum, K2-EDTA, K3-EDTA, Li-heparin, Na-citrate, capillary whole blood) Minimum 40–50 paired draws per alternative matrix Deming regression assessing slope close to 1.0 (±5%) and negligible intercept	Validating only serum and extrapolating the claim to all plasma types Capillary fingerstick claims based solely on venous whole blood data
Specimen & Reagent Stability	CLSI EP25-A	Specimen: transport simulation, ambient hold time, refrigerated (2–8 °C), frozen (−20 °C, −80 °C) and 3–5 freeze-thaw cycles Reagent: real-time shelf life (accelerated Arrhenius data must be confirmed by real-time tracking), in-use/open-vial stability, on-board analyser stability	Using accelerated shelf-life studies exclusively without ongoing real-time data Missing freeze-thaw validation for biological reference samples
Metrological Traceability	ISO 17511:2020	Clear calibration cascade linking assigned calibrator values back to SI units or an international certified reference material (e.g. WHO International Standard, NIST) Explicit calculation of combined measurement uncertainty (u_c) through each step of the calibration hierarchy	Merely naming a WHO standard without detailing the uncertainty propagation Uncalibrated internal "master lots" lacking traceable links

Key IVDR Performance Evaluation Documents

Performance Evaluation Plan (PEP): defines the intended purpose, claims, evidence needed and methods to obtain it.
Performance Evaluation Report (PER): summarises scientific validity, analytical and clinical performance data and concludes on conformity.
Performance study documentation: study plans and reports for analytical and clinical performance studies, including those under Articles 57 to 77 where applicable.
Post-Market Performance Follow-up (PMPF) plan and report: keep the evidence current after the device is on the market.
Summary of Safety and Performance (SSP): required for Class C and Class D devices.

How Requirements Change by Risk Class

Every IVD class needs a performance evaluation, but the depth of evidence and level of scrutiny rise with risk. Class D devices may also need to meet common specifications and undergo verification by an EU reference laboratory. For legacy CE-IVD devices, data generated under the old IVD Directive often needs to be supplemented to meet IVDR expectations.

Our IVD Performance Evaluation Services

Gap assessment of existing performance data against IVDR Annex XIII
Scientific validity literature reviews
Performance evaluation plans and reports (PEP and PER)
Analytical and clinical performance study design and protocol review
PMPF plans, PMPF reports and SSP preparation
Support with notified body questions on performance evidence

Performance evaluation is one part of the wider IVDR CE certification process, alongside classification, technical documentation and quality management. Planning it early is the best way to avoid delays during notified body review.

Get Help With Your IVD Performance Evaluation

Tell us about your device, its intended purpose and the data you already have. We will identify the gaps and propose a clear plan for your performance evaluation.

Request an IVDR Performance Evaluation Proposal