Conceptualizing Machine Learning for Dynamic Information Retrieval of Electronic Health Record Notes

Abstract

The large amount of time clinicians spend sifting through patient notes and documenting in electronic health records (EHRs) is a leading cause of clinician burnout. By proactively and dynamically retrieving relevant notes during the documentation process, we can reduce the effort required to find relevant patient history. In this work, we conceptualize the use of EHR audit logs for machine learning as a source of supervision of note relevance in a specific clinical context, at a particular point in time. Our evaluation focuses on the dynamic retrieval in the emergency department, a high acuity setting with unique patterns of information retrieval and note writing. However, our framework is general and can be applied to other clinical settings and with other data modalities (e.g., labs, medications, imaging). We apply our framework to the oncology setting to demonstrate its utility to other clinical workflows. We show that our methods can achieve an AUC of 0.963 in the ED and 0.937 in oncology when predicting which notes will be read in an individual note writing session. We additionally conduct user studies with several clinicians and find that our framework can help clinicians retrieve relevant information more efficiently.