What is the clinical evidence from this study?

Study design: Editorial. Population: Cardiovascular disease. Intervention: Artificial intelligence-based prediction models.

What question did this study set out to answer?

The aim is to establish minimal quality criteria for developing and validating AI prediction models in cardiovascular health.

April 22, 2026Open Access

View Full Paper

Five critical quality criteria for artificial intelligence-based prediction models

FRFlorien S van RoyenUniversity Medical Center Utrecht FAFolkert W. AsselbergsHeart Failure & Transplant FAFernándo AlfonsoInterventional / Structural Cardiology

Key Result

The European Heart Journal editors propose five minimal quality criteria for AI-based prediction models: complete reporting, clear intended use, rigorous validation, adequate sample size, and open code.

Key Points

The aim is to establish minimal quality criteria for developing and validating AI prediction models in cardiovascular health.
Proposed five criteria: complete reporting, defined intended use, rigorous validation, adequate sample size, and openness of code.
Focus on improving the impact and relevance of AI in clinical settings.
Emphasis on complete reporting to ensure transparency.
Rigorous validation is crucial for credibility and effectiveness of models.
Large sample sizes are necessary to enhance the reliability of the predictions.

PICO

Population

Cardiovascular disease

Intervention / Comparator

Artificial intelligence-based prediction models

Limitations

Commercial interests and privacy regulations may limit the ability to fully share proprietary AI model code and data.
Explainable AI methods like SHAP values are not a sufficient replacement for sharing actual model code to allow external validation.

Abstract

To raise the quality of clinical artificial intelligence (AI) prediction modelling studies in the cardiovascular health domain and thereby improve their impact and relevancy, the editors for digital health, innovation, and quality standards of the European Heart Journal propose five minimal quality criteria for AI-based prediction model development and validation studies: complete reporting, carefully defined intended use of the model, rigorous validation, large enough sample size, and openness of code and software.

KI fragen

Bookmark

View Full Paper

KI fragen

Bookmark

View Full Paper

Five critical quality criteria for artificial intelligence-based prediction models

Key Result

Key Points

PICO

Limitations

Abstract

Cite This Study

Also Consider

Also Consider