Best Practice Advisories (BPA) are frequently ignored by clinicians. As a proof of concept, we used machine learning algorithms to develop personalized user activity profiles to suppress herpes zoster vaccination BPAs unlikely to have an interaction. In a sample of nearly 13000 primary care visits on average through 5-fold cross validation, we report a highly effective model with average AUROC of 0.919 and average AUPRC of 0.562.
Describe the new knowledge and additional skills the participant will gain after attending your presentation.: Attendees will be introduced to the concept of dynamic customization of clinical decision support to predict response to clinical alerts from user profiling.
Ji Chen (Presenter)
New York University School of Medicine
Yindalon Aphinyanaphongs, New York University School of Medicine
Devin Mann, New York University School of Medicine
Eduardo Iturrate, NYU Langone Health
Sara Chokshi, New York University School of Medicine
Roshini Hegde, New York University School of Medicine