Brain Cade, PhD, Brigham and Women’s Hospital, Boston, MA, discusses the development of a phenotyping algorithm to improve the precision of electronic health record data in the characterization of obstructive sleep apnea (OSA). The algorithm uses natural language processing (NLP) to identify associations between sleep apnea and comorbidities in a large clinical biobank in order to investigate the relationship between polysomnography statistics and comorbid disease. International Classification of Sleep Disorders criteria was applied to clinical chart reviews on 300 participants diagnosed with sleep apnea to classify true cases. NLP phenotyping strategy improved model precision compared to the use of one diagnosis code. This interview was recorded at the World Sleep Congress 2022 in Rome, Italy.