According to a study published in JAMIA, researchers may be able to examine clinician EHR behavior at scale by annotating EHR audit files using machine learning. Researchers can examine physician behaviors at the microscale by using raw audit logs, which record a trail of clinician activities within the EHR. However, clinical activity context is missing from raw audit records. Although clinical professionals might manually annotate audit logs to offer context, the researchers warned that such efforts are labor-intensive, possibly error-prone, and not easily scalable. The researchers created a unique method to annotate EHR events using action embeddings, a scalable unsupervised machine learning approach, utilizing a dataset of 15 million raw audit log actions conducted by 88 intern physicians across inpatient and outpatient settings. In the report, researchers stated the following:
“Our embeddings approach provides insights—complementary to clinician expert audit log annotations—into the context in which individual EHR actions occur. By situating specific actions within the broader context of their cooccurring activities, our method provides opportunities to generate meaningful hypotheses regarding clinical work activities, clinical workflows, and user interactions with the EHR.”
The two different EHR information-gathering techniques that were employed during the chart biopsy procedures were discovered by the researchers. These techniques may have been applied differently by different individuals or in other settings. The researchers specifically mentioned the possible opportunity to examine workflow effectiveness and how it affects consequences like burnout. The method also discovered that tasks including ordering orders, making notes, and sending emails had a tendency to happen in a variety of contexts, possibly reflecting the fragmented nature of clinical job.
“Because our method treats audit log activities as tokens, it can be applied across any EHR raw audit log source, regardless of vendor or institution,” the study authors pointed out…Recent research has utilized raw audit logs to measure physician workload, workflow, teamwork, cognitive burden, and burnout, illustrating the considerable potential of raw audit log data to capture physician behaviors at the macroscale,”
