Abstract
Electronic health record (EHR) systems are essential for modern healthcare but contribute to a significant documentation burden, affecting physician workflow and well-being. While previous studies have identified differences in EHR usage across demographics, systematic methods for identifying high-burden physician groups remain limited. This study applies cluster analysis to uncover distinct EHR usage profiles and provide a framework to inform the development of targeted interventions.This study investigated two research questions: (1) Can cluster analysis effectively identify distinct physician EHR usage profiles? (2) How do these profiles vary across physician demographics and practice characteristics? We hypothesized that (1) EHR usage clusters would emerge based on workload intensity, after-hours documentation, and In Basket management patterns, and (2) would be significantly associated with physician experience, sex, and specialty.We analyzed outpatient EHR usage data from 323 physicians at an academic health system using Epic Signal, an analytical tool for Epic EHR. Using k-means clustering, we examined six metrics representing EHR workload (after-hours and extended-day activities) and In Basket efficiency (message handling and management patterns). We assessed cluster differences and conducted subgroup analyses by physician sex and specialty.Two distinct physician clusters emerged: one high-burden cluster, predominantly comprising experienced primary care physicians, and another lower-burden cluster, consisting mostly of younger specialists. Physicians in the high-burden cluster spent nearly three times as much time on after-hours documentation and In Basket management. While message response times remained similar, subgroup analyses revealed significant sex and specialty-based differences, particularly in the lower-burden cluster.Cluster analysis effectively identified distinct EHR usage patterns, highlighting disparities in workload by experience, sex, and specialty. This approach provides a scalable, data-driven method for health systems to identify at-risk groups and design targeted interventions to mitigate documentation burden and enhance EHR efficiency.