A performance study is a clinical performance study used to demonstrate the clinical performance of an in vitro diagnostic medical device (IVD) for its intended purpose. It provides the clinical evidence that the device produces reliable results in the target population and can thus make a relevant contribution to patient care.
Objective, Differentiation, and Typical Endpoints
In EU law, the performance study is closely linked to the In Vitro Diagnostic Regulation (IVDR). The term is frequently used in the context of clinical evidence, performance evaluation, and the demonstration of clinical performance. Unlike pharmaceuticals, the focus is not on the therapeutic efficacy of an active substance, but on the diagnostic validity of a test. A performance study can be prospective or retrospective in design, depending on the research question, availability of specimen material, and ethical requirements.
It is important to differentiate these from analytical performance studies: analytical performance refers to laboratory and measurement parameters such as precision, accuracy, limit of detection, or interferences. Clinical performance, on the other hand, addresses metrics such as diagnostic sensitivity and specificity, positive and negative predictive values, or agreement with an established reference procedure. In practice, both levels are merged into a consistent performance evaluation and documented summarily in a Performance Evaluation Report.
- Endpoints: Sensitivity, specificity, AUC/ROC, agreement (e.g., Kappa), clinical relevance of cut-offs.
- Comparison: Reference standard (gold standard) or clinical standard-of-care test.
- Population: Target population according to the intended purpose, including relevant subgroups (e.g., age, comorbidities).
A common pitfall is a non-representative study population, for example, if samples originate only from specialized centers or if the prevalence in the study collective deviates significantly from real-world care. This particularly affects predictive values and can lead to misleading performance claims in subsequent labeling.
Study Design and Practical Implementation
The study design depends heavily on the type of product (e.g., screening test, companion diagnostic, point-of-care test) and the question of which clinical statement is to be substantiated. Common designs include case-control studies, cohort studies, or studies with consecutive sample collection. For rare diseases, the use of biobanks or existing samples (where legally and ethically permissible) can be decisive.
Operationally, roles and responsibilities must be clearly separated: sponsor responsibility, investigator responsibility, laboratory responsibility, and, if applicable, the tasks of a Contract Research Organization. Interfaces often arise between the clinical study, laboratory processes, and regulatory documentation, necessitating clean versioning and a consistent Trial Master File or Study File. Depending on the complexity, Electronic Data Capture and a suitable electronic Case Report Form are used for electronic data collection.
Typical documents include the Performance Study Plan, informed consent forms, laboratory manual, sample logistics documentation, statistical analysis plan, and the clinical performance study report. In multicenter setups, training records, delegation logs, and procedures for handling deviations (protocol deviations) must also be established.
In practice, it is worthwhile to establish a clear definition of the reference standard and to harmonize pre-analytics before the start of the study. Differences in sampling, transport times, storage, or freeze-thaw cycles can influence the measured performance more significantly than the test itself. Risk-based monitoring and clear data reconciliation between the laboratory information system and the clinical database reduce subsequent query loops and support a rapid, database-lock-like finalization of the datasets.
Regulatory Integration under IVDR
Under the IVDR, the performance study is part of the overall performance evaluation of an IVD. The results are incorporated into the technical documentation and the Performance Evaluation Report. Requirements for clinical evidence vary depending on the risk class and product type, with companion diagnostics and high-risk IVDs regularly facing increased requirements.
Furthermore, data protection (e.g., GDPR), biosample regulations, and national requirements must be observed, particularly if samples are collected or processed in Germany. For manufacturers, it is crucial that the study objective, intended purpose, and claims in the labeling are consistent. Discrepancies between the clinical question, statistical analysis, and subsequent product claims frequently lead to requests for additional information during audits or from Notified Bodies. Clear traceability of raw data, an audit trail in IT systems, and a plausible benefit-risk profile are central components of a robust submission.
The handling of deviations is also relevant for the approval process: every protocol deviation should be evaluated and documented to ensure the validity of the analysis population. Especially in retrospective designs, consent and data protection concepts must be clarified early to avoid the subsequent non-usability of data.
Relevance for clinical trials
Performance studies combine clinical methodology with regulatory requirements from the medical device world. For sponsors and manufacturers, they are an essential building block for achieving market access and CE marking while simultaneously underpinning the clinical validity of the test. In CRO practice, the effort often lies less in pure patient recruitment and more in clean sample logistics, the standardization of pre-analytics, and the harmonization of multiple data sources (clinical data, laboratory results, device logs).
Full-service CROs like mediconomics typically provide support with study design, site management, monitoring, data management, and the preparation of study-related documentation. Early coordination with Regulatory Affairs and, if necessary, with a Notified Body is particularly valuable to avoid subsequent iteration loops during the performance evaluation.
Frequently Asked Questions (FAQ)
When is a performance study required under the IVDR?
It is required if the existing clinical evidence is insufficient for the intended purpose or if new claims, new populations, or a new reference procedure need to be substantiated. The scope depends on the risk class, state of the art, and available data.
What is the difference between clinical performance and analytical performance?
Analytical performance describes the measurement and laboratory parameters of the test, while clinical performance describes the diagnostic validity in the target population. Both levels are complementary and are merged in the performance evaluation.
What typical errors lead to regulatory inquiries?
These often include unclear intended purpose formulations, non-representative populations, lack of justification for the reference standard, inconsistent cut-offs, and incomplete evidence of data integrity and governance.
Regulatory References
- Regulation (EU) 2017/746 (IVDR): Requirements for performance evaluation and performance studies for in vitro diagnostic medical devices.
- Regulation (EU) 2017/745 (MDR): Framework for the clinical evaluation of medical devices, relevant for interfaces and general concepts.
- ICH E6(R3) Good Clinical Practice: Principles on data integrity, roles, and quality systems in clinical investigations.
- GDPR (EU) 2016/679: Data protection requirements for the processing of personal health data.