This federal Project Grant award of $707,410.00, provided by the National Heart Lung and Blood Institute (CFDA 93.837 Cardiovascular Diseases Research), supports a multi-health system study by Duke University to develop machine learning models that can mitigate the impact of hidden hypoxemia (HH) caused by pulse oximetry inaccuracies in patients with darker skin tones. The project aims to leverage electronic health record data and informative patient sampling to build predictive models that can identify high-risk patients for HH and determine which patients may benefit from additional skin tone measurements to improve pulse oximetry accuracy. This work will lay the foundation for future clinical implementation studies to test the effectiveness of these models in reducing pulse oximetry inequities. The award commenced on February 1, 2025 and is scheduled for completion by November 30, 2029.
Generated 3/4/25, 2:19 AM