Click here to open the filter panel
I
Inspection
I
Institutional Review Board (IRB)
I
Interactive response technology (IRT)
I
Interactive web response system (IWRS)
I
International data transfer
I
Investigational Medicinal Product Dossier (IMPD)
I
Investigational medicinal product (IMP)
I
Investigational New Drug (IND) application
I
Investigator
I
Investigator's brochure (IB)
I
In Vitro Diagnostic Regulation (IVDR)
I
ISO/IEC 27001 (information security management)
I
ISO/IEC 42001 (AI management system)
J
Joint controllers
K
k-anonymity
K
Key-coded data (coded data)
K
Key (cryptographic key and pseudonymisation key)
K
Kinetics (pharmacokinetics and pharmacodynamics)
L
Large language model (LLM)
L
Lawfulness
Click a letter to view all its definitions
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Term of the Day

Natural history study

A natural history study is a preplanned observational study intended to track the course of a disease over time, identifying demographic, genetic, environmental and other variables that correlate with its development and outcomes in the absence of intervention, or under standard of care. Designs may be retrospective (chart review of existing records) or prospective (longitudinal follow-up of a cohort or registry).

Natural history data is particularly important in rare and paediatric diseases, where randomised placebo-controlled trials may be infeasible or unethical. The FDA (guidance on rare disease natural history studies, 2019) and the EMA accept well-designed natural history studies to define endpoints and biomarkers, identify patient subgroups, estimate sample sizes and, in some cases, serve as external or historical control arms for single-arm trials supporting orphan products.

Because they are non-interventional, natural history studies fall outside the CTR and are governed by national law (for example France's MR-003 or MR-004 reference methodologies) and by the GDPR. They typically involve secondary use of medical records, long-term follow-up, genetic data and small populations in which anonymisation is rarely achievable, so pseudonymisation, a DPIA and a robust research legal basis under Art. 9(2)(j) are essential. Registries maintained by patient organisations or academic consortia raise additional questions of joint controllership and data access governance.

X

XAI (explainable artificial intelligence)

XAI, or explainable artificial intelligence, is the set of methods, tools and design practices that make the behaviour and outputs of AI systems understandable to humans: which inputs drove a prediction, how confident the model is, what would have changed the outcome, and where the model is likely to fail. Approaches range from inherently interpretable models (decision trees, rule lists, generalised additive models) to post-hoc techniques applied to complex models such as deep neural networks: feature attribution (SHAP, LIME, integrated gradients), saliency maps for imaging, counterfactual explanations, prototype-based explanations and natural-language rationales generated by large language models. Explanations may be global (how the model behaves overall) or local (why this output for this patient), and they must be validated: a plausible-looking explanation can itself be wrong.

In healthcare and life sciences, explainability is both a regulatory and a clinical requirement. Clinicians exercising human oversight of a diagnostic or triage tool need to understand its reasoning to accept or override it; notified bodies assessing AI-enabled software as a medical device under the MDR expect evidence of performance across subgroups and of failure modes; and the FDA's guiding principles on transparency for machine learning-enabled devices (2024) and good machine learning practice call for information that users can understand.

Legally, XAI supports several obligations without being named in any of them. The GDPR requires controllers to provide meaningful information about the logic involved in automated decision-making and profiling (Art. 13(2)(f), 14(2)(g), 15(1)(h)) and to guarantee the right to contest a solely automated decision (Art. 22(3)), which the Court of Justice in Dun & Bradstreet Austria (C-203/22) read as requiring an explanation of the procedure and principles actually applied, in a form the person can understand. The EU AI Act requires high-risk systems to be sufficiently transparent for deployers to interpret outputs (Art. 13), to be designed for effective human oversight (Art. 14), and, under Art. 86, gives affected persons a right to clear and meaningful explanations of the role of the AI system in decisions with legal or significant effects. Documenting the explainability approach, its validation and its limits is therefore part of the technical documentation, the FRIA and the DPIA; iliomad's AI compliance services cover this documentation.