Evaluating the Portability of Rheumatoid Arthritis Phenotyping Algorithms: case study on French EHRs at the Patient and Encounter Level

Anonymous

Abstract: High-throughput phenotyping can accelerate the development of statistical analysis from cohorts of Electronic Health Records. Previous work has successfully used machine learning and natural language processing for the phenotyping of Rheumatoid Arthritis (RA) patients in hospitals within the United States and France. Our goal is to evaluate the adaptability of RA phenotyping algorithms to a new hospital, both at the patient and encounter levels. Two algorithms are adapted to the context of the new hospital and evaluated with a newly developed RA gold standard corpus, including annotations at the encounter level.The adapted algorithms offer comparable performance for patient-level phenotyping on the new corpus (F1 0.71 to 0.79), performance is lower for encounter-level phenotyping (F1 0.54 to 0.57), illustrating adaptation feasibility and cost. The first algorithm incurred a heavier adaptation burden because it required manual feature engineering. However, it is less computationally intensive than the second, semi-supervised, algorithm.

Paper Type: long