Completeness
Covering the relevant aspects of model development, validation and application.
← HOMETHE CLAIRE INITIATIVE / 02 · THE FRAMEWORKPublished foundation
02 / CLAIRE FRAMEWORK
Six principles.
One clearer standard.
The published framework organizes transparent AI reporting around six complementary principles, operationalized through a structured checklist and supporting editorial instruments.
Learnability
Describing how the model is structured and trained.
Applicability
Addressing clinical context, practical use, generalizability and limitations.
Interpretability
Supporting clinician-oriented understanding of model outputs.
Reproducibility
Promoting adequate documentation of methods, data, code and computational procedures.
Evaluation
Ensuring results are reported in sufficient detail for critical appraisal and comparison.
15
Reporting checklist
Fifteen items structured according to conventional scientific reporting.
↗
Editorial reference guide
A consistent reference for authors, peer reviewers and editors.
%
Standardized scoring
A summary of fulfilled applicable items, with four reporting-completeness categories.
An important distinction: CLAIRE assesses reporting completeness and transparency. It does not replace guidance on study design, methodological quality assessment or risk-of-bias evaluation. Completely reported studies may still have methodological limitations.