Criteria to Assess Cardiovascular Disease Predictive Utility

Got story updates? Submit your updates here. ›

The American Heart Association has published new criteria to help assess the predictive and clinical utility of tools, including AI systems, that aim to forecast cardiovascular disease risk and outcomes. The guidelines outline key factors to consider when evaluating the accuracy, reliability, and real-world applicability of such predictive technologies.

Why it matters

As AI and other advanced analytics become more prevalent in healthcare, it's crucial to have standardized methods for vetting their effectiveness and ensuring they provide meaningful insights to improve patient care and outcomes. These new criteria from the AHA will help healthcare providers and researchers rigorously evaluate emerging cardiovascular disease prediction tools.

The details

The criteria cover areas such as the quality and diversity of training data, model performance metrics, clinical validation, and assessment of potential biases. Researchers are encouraged to test predictive tools across diverse populations to ensure equitable and unbiased results. The guidelines also stress the importance of demonstrating how the use of these technologies can lead to tangible improvements in patient management and health outcomes.

• The new criteria were published on February 11, 2026.

What’s next

The AHA plans to host a session on the new criteria at its EPI|Lifestyle conference in Boston, MA from March 17-20, 2026.