Accurately diagnosing a patient’s condition is a foundational step in delivering high quality care. An error in diagnosis can significantly impact patient outcomes. There is a critical need to strengthen the quality and accuracy of the diagnostic process, and to do so will require reliable measurement of diagnostic excellence. Measurement in this area, however, has proved an intractable challenge.  

Data critical to diagnostic excellence measurement, such as symptoms, are challenging to capture in structured fields. Artificial intelligence (AI) methods, such as natural language processing (NLP) and machine learning (ML) can help to address this challenge by pulling data from free-text fields, such as clinical notes and diagnostic test reports. While not widespread, these AI methods are becoming more common in quality measurement. However, guidance for reviewing the use of AI methods in quality measures have not been established. To help address this challenge and others related to diagnostic excellence measurement, the Gordon and Betty Moore Foundation is funding the National Quality Forum (NQF) for a three-year project to identify actionable solutions to address real-world challenges facing the field of diagnostic excellence measurement. One aspect of the project will be to forge consensus on a framework and criteria for assessing the use of AI methods in quality measures.  

To facilitate this work, NQF has convened a multistakeholder technical expert panel (TEP) that will drive insights and consensus on guidance that sets standards for reviewing the use of AI methods in national quality measures. The draft guidance will be available for public comment in summer 2025 and NQF plans to publish the final guidance in winter 2025.

For additional information about the larger Advancing Measurement of Diagnostic Excellence for Better Healthcare Initiative, see this page.

Contact Information

For more information, please contact the project team at DxExcellence@qualityforum.org.