SHM Converge Scientific Abstract ePoster Gallery-Learning Behaviors of Hospital Medicine Resident Physicians: Insights From Ehr Audit Logs

Pdf Summary

This study explored how hospital medicine residents use the electronic health record (EHR) and whether audit log data can reveal meaningful learning patterns. Because EHR proficiency is a key competency in modern residency training, the authors aimed to create scalable methods to characterize resident behavior and examine how it changes over time.<br /><br />Using six months of 2022 EHR audit log metadata, the researchers collected data on all residents scheduled for hospital medicine shifts during the study period. They applied a two-step machine learning approach that used clustering and sequencing techniques to identify common “phenotypes” of EHR behavior. They then tested whether these patterns were associated with resident characteristics and schedule factors.<br /><br />The algorithm identified 6 distinct activity states and 4 shift-level work patterns. The main difference between residents was when they completed their notes. More senior residents were more likely to fall into an “early” note-writing cluster, with a statistically significant association (p = 0.004). Over time, residents’ EHR behaviors became more similar, suggesting co-learning and convergence in work patterns across training levels.<br /><br />The findings suggest that residents differ in note-writing timing and overall EHR efficiency, but their behaviors tend to align as training progresses. The authors propose that EHR audit logs could support precision education by helping identify learner needs and tailoring feedback or coaching. Future work will incorporate resident wellness data to better understand how EHR behaviors relate to overall training experience and well-being.

Asset Subtitle

Josh Weiner

Meta Tag

Author List Andrew P.J Olson, Dori A. Cross, Josh Weiner

Category Research

Concept Electronic Health Record

Concept Resident Physician

Concept Behavior Pattern

Concept Audit Log

Concept Activity State

Distinguished Non-Finalist

Presenter Organization University of Minnesota School of Public Health

Presenting Author Josh Weiner

Track Translating Research into Practice

Keywords

electronic health record

EHR audit logs

hospital medicine residents

resident behavior

machine learning

clustering

sequence analysis

note-writing timing

precision education

resident training

Electronic Health Record

Resident Physician

Behavior Pattern

Audit Log

Activity State