A Clinical Performance Study Plan, or CPSP, defines how an IVD clinical performance study will answer a specific clinical question while protecting study subjects and generating reliable evidence. Under Annex XIII of Regulation (EU) 2017/746, a clinical performance study must be performed on the basis of a CPSP.
The plan should connect the device’s intended purpose, performance claims, target population, specimen type, study endpoints, comparator, statistical methods and operational controls. It should also establish the applicable regulatory route before the study begins.
A CPSP is not the same as a Performance Evaluation Plan, Clinical Performance Study Report or Performance Evaluation Report. Each document has a different role in the IVDR evidence lifecycle.
What is clinical performance under the IVDR?
Under the EU IVDR, clinical performance is the ability of an IVD to produce results that correlate with a particular clinical condition, physiological or pathological process or state, in accordance with its intended purpose and target population. Clinical performance forms part of the device’s performance evaluation and must be supported by appropriate clinical performance data.
Do all IVDs require a new clinical performance study?
Not necessarily.
Article 56(4) of the IVDR states that clinical performance studies must be performed unless the manufacturer provides due justification for relying on other sources of clinical performance data. Those sources may include peer-reviewed scientific literature, previously conducted studies or published experience gained through routine diagnostic testing.
The manufacturer must first identify the clinical performance evidence needed to support the intended purpose and claims. If the available evidence is insufficient, a new clinical performance study may be required.
When a clinical performance study is undertaken, the study must have a documented CPSP that complies with the applicable requirements of IVDR Annex XIII.
CPSP, CPSR, PEP and PER: what is the difference?
| Document | Purpose | When it is used |
|---|---|---|
| Clinical Performance Study Plan (CPSP) | Defines the rationale, objectives, design, conduct, monitoring and proposed analysis of a specific study | Before and during the clinical performance study |
| Clinical Performance Study Report (CPSR) | Documents the study methods, results, conclusions and negative findings | After completion of the study |
| Performance Evaluation Plan (PEP) | Defines the overall strategy for generating and evaluating scientific validity, analytical performance and clinical performance evidence | Across the device lifecycle |
| Performance Evaluation Report (PER) | Assesses the complete performance evaluation and documents the resulting clinical evidence | During conformity assessment and throughout the device lifecycle |
The CPSP should be designed to generate evidence that can be reported transparently in the CPSR and incorporated into the wider PER.
Start with the intended purpose and performance claims
The intended purpose determines what the study must demonstrate.
Before selecting sites or calculating sample size, the sponsor and manufacturer should define:
- the analyte or marker;
- the clinical condition or physiological state;
- the intended population;
- the intended user;
- the specimen type;
- the testing environment;
- the role of the result in patient management;
- the relevant clinical performance parameters; and
- the claims that the final evidence must support.
The primary endpoint should map directly to the intended purpose. A study designed around a broader population, different specimen type or different clinical decision may produce technically valid data that does not support the proposed claim.
Determine the IVDR regulatory route
Device class alone does not determine whether a performance study requires authorisation or notification.
The sponsor should assess the study under Articles 57, 58 and 70 of the IVDR and consult the decision pathway in MDCG 2025-5.
| Study situation | Principal IVDR consideration |
|---|---|
| All performance studies | Article 57 and the applicable Annex XIII requirements |
| Non-CE-marked device, or CE-marked device used outside its intended purpose, where the study meets one or more Article 58(1) criteria | Application for authorisation to the relevant competent authority and the additional requirements of Articles 58–77 and Annex XIV |
| Investigational companion diagnostic | Article 58(2) applies. The specific route depends on the study design; studies using left-over samples only are subject to notification under the current rules |
| CE-marked device used within its intended purpose where subjects undergo additional invasive or burdensome procedures | Article 70 notification may apply |
| Other studies that do not meet the Article 58 or 70 criteria | Article 57 continues to apply, together with applicable national ethics, data-protection and study requirements |
An interventional clinical performance study is one in which test results may influence patient-management decisions or guide treatment. Surgically invasive sample-taking undertaken specifically for the study, or other additional risks to study subjects, can also trigger Article 58(1).
Applications and notifications are directed to competent authorities. Ethics committee requirements depend partly on national provisions. A notified body may assess the resulting clinical evidence during conformity assessment, but it is not the authority that generally authorises the performance study before it begins.
For companion diagnostic-specific authorisation planning, see MDx CRO’s IVDR Annex XIV performance study guide.
What must an IVDR Clinical Performance Study Plan contain?
Annex XIII, Part A, Section 2.3.2 requires the CPSP to define the rationale, objectives, design, proposed analysis, methodology, monitoring, conduct and record-keeping of the study.
The plan should cover the following groups of information.
Study governance and identification
- study identification number;
- sponsor and EU legal representative, where applicable;
- investigators and study sites;
- planned start date and duration;
- roles, responsibilities and communication pathways.
Device and clinical claim
- device identification and description;
- intended purpose;
- analytes or markers;
- manufacturer;
- metrological traceability;
- technical and functional features covered by the study.
Design and evidence strategy
- study rationale and hypotheses;
- observational or interventional design;
- current state of the art;
- anticipated benefits and risks;
- comparator or reference method;
- clinical performance parameters and endpoints;
- justification of the design’s scientific validity.
Population and specimens
- target population and representativeness;
- inclusion and exclusion criteria;
- vulnerable populations, where applicable;
- specimen types;
- collection, handling, transport, storage and testing procedures.
Statistics and control of bias
- sample-size justification;
- statistical methods;
- success criteria;
- management of missing, invalid and indeterminate results;
- relevant subgroups;
- measures to control bias and confounding.
Study conduct and data quality
- monitoring approach;
- data collection and record-keeping;
- device accountability;
- management of deviations and amendments;
- quality controls;
- criteria for suspension or early termination.
Subject protection and reporting
- ethical principles;
- informed consent process, where applicable;
- privacy and data-protection arrangements;
- safety recording and reporting;
- subject follow-up;
- communication of test results;
- preparation and publication of the Clinical Performance Study Report.
If an Annex XIII element is not applicable because of the selected design, the CPSP should document the justification rather than simply omit it.
How should sample size be determined?
The IVDR does not establish a universal sample size based solely on device class.
The calculation should reflect:
- the primary clinical performance endpoint;
- the expected result and required precision;
- disease prevalence or the planned number of positive and negative specimens;
- the comparator or reference method;
- intended population and relevant subgroups;
- anticipated invalid, missing or unevaluable results;
- potential withdrawals or specimen losses; and
- any planned interim or multiple analyses.
Statistical planning should begin while the intended purpose and endpoints are being defined. Calculating sample size after sites or available specimens have already been selected risks allowing operational constraints to determine the scientific question.
Select a defensible comparator and state of the art
The CPSP should explain why the selected comparator or reference method is appropriate for the intended purpose.
Depending on the device and clinical question, the comparator may be:
- an established reference method;
- a recognised clinical diagnosis;
- a composite reference standard;
- an existing CE-marked device; or
- another scientifically justified method.
The plan should define how discordant, indeterminate and missing results will be handled. It should also explain how the study reflects the current state of the art in diagnosis or medicine.
Control pre-analytical and operational variability
A statistically sound design can still fail to generate transferable evidence if specimen handling and site procedures do not represent the intended-use environment.
The CPSP and supporting operational documents should address:
- specimen eligibility and traceability;
- collection materials and procedures;
- stability and storage conditions;
- transport time and temperature;
- freeze-thaw limits;
- operator training;
- device and software versions;
- site equipment and quality controls;
- blinding and result adjudication;
- data capture and audit trails.
These controls should be consistent across the CPSP, Investigator’s Brochure, instructions for use, case report forms, Statistical Analysis Plan and monitoring plan.
How EN ISO 20916:2024 supports the CPSP
EN ISO 20916:2024 is the European adoption of ISO 20916:2019 and defines good study practice for clinical performance studies using specimens from human subjects.
The standard covers planning, design, conduct, recording and reporting. Its application can strengthen areas such as sponsor and investigator responsibilities, risk management, monitoring, specimen control, data integrity and subject protection.
EN ISO 20916:2024 was cited as a harmonised standard under the IVDR through Commission Implementing Decision (EU) 2024/2625 on 9 October 2024. Applying the clauses mapped in Annex ZA supports a presumption of conformity with the corresponding IVDR requirements.
The standard remains voluntary and does not replace the IVDR, MDCG guidance or applicable national requirements.
Final CPSP planning checks
Before finalising the plan, confirm that:
- the intended purpose, endpoints and performance claims are aligned;
- the study population represents the intended population;
- the comparator and state-of-the-art assessment are justified;
- the sample size supports the primary endpoint;
- bias and confounding have been addressed;
- the regulatory route has been assessed under Articles 57, 58 and 70;
- national ethics and submission requirements have been checked;
- specimen and data workflows are operationally feasible;
- related study documents use consistent device versions, terminology and procedures; and
- amendments, safety reporting and early termination are controlled.
Frequently asked questions
In the IVDR context, the Clinical Performance Study Plan is commonly treated as the study protocol. CPSP is the formal term used in Annex XIII.
Not as a general rule. Depending on the applicable IVDR route, the sponsor may need authorisation from or notification to the competent authority and must comply with national ethics requirements. The notified body assesses clinical evidence as part of the device’s conformity assessment and may already be involved in the wider programme.
Harmonised standards remain voluntary. Applying the clauses covered by Annex ZA can provide a presumption of conformity with the corresponding IVDR requirements, but alternative solutions may be used if the manufacturer demonstrates compliance with the Regulation.
Yes, but using left-over samples does not automatically remove every IVDR obligation. The route depends on the device, intended purpose, study design, additional procedures or risks and whether the device is a companion diagnostic. MDCG 2025-5 should be consulted for the current decision pathway.
The methods, results, conclusions and negative findings are documented in the Clinical Performance Study Report. The resulting evidence is then evaluated alongside scientific validity and analytical performance in the Performance Evaluation Report.
Support with IVDR clinical performance studies
A defensible CPSP must work as both a scientific protocol and an operational document. MDx CRO supports manufacturers and sponsors with study strategy, protocol development, biostatistics, site selection and qualification, submissions, monitoring, data management and final reporting.
Explore our IVD clinical performance study services or contact MDx CRO to discuss your study.
