Council on Medical Student Education in Pediatrics


Search This Site

COMSEP Meeting in Ottawa, ON

Poster Presentation:

Decoding the EMR for Medical Students

Kirstin . Nackers,University of Wisconsin School of Medicine and Public Health,Madison,Wisconsin,Andrea . Carberry,University of Wisconsin School of Medicine and Public Health,Madison,WI,Michael . Wilhelm,University of Wisconsin School of Medicine and Public Health,Madison,WI

Background: Electronic medical records (EMRs) are ubiquitous in academic medical centers.  Evidence suggests they may negatively impact medical education, and some centers limit student documentation in the EMR. Our institution has allowed documentation in the EMR since its implementation in 2007. To provide more consistent feedback, clerkship directors (authors KN & MW) began grading all submitted admission notes (“H&Ps”) this year.

Objective: Identify common themes for errors in medical student notes and determine which may be caused (or encouraged) by using EMRs.

Methods: Our institution deteremined this research does not qualify as human subjects research and is therefore exempt.  During the clerkship, students submit two H&Ps for grading; the standardized rubrics and narrative feedback examine data gathering (HP1) and clinical reasoning (HP2). A random sample of HP2 rubric and narrative feedback was coded independently by KN and MW using open coding and constant comparative techniques. The coding structure was validated by a third investigator (AC).  Once refined and theoretical saturation achieved, all remaining transcripts were coded independently; 15% were double-coded to ensure interrater reliability. When uncertainty arose, individual HP2s were randomly assigned for independent, blinded analysis and discrepancies resolved via consensus.  

Results: Eleven themes were identified and used for coding the remaining HP2s; the majority had errors in at least three themes. The four most common were “missing key information”, “important abnormal findings identified but not addressed”, “missing major components of the plan”, and “rationale not provided/insufficient for plan”. Each was present in over half. A number of other errors were classified as “errors induced by the use of a template” and three other themes were considered potentially related to the use of an EMR template (i.e., automated highlighting of abnormal labs that were then not addressed).

Discussion: While clinical reasoning errors are expected for third-year students, a surprisingly large number of notes were missing key information, had abnormal results not addressed, or lacked key components of the plan. Despite instructions not to use EMR templates for the assignment, many errors mapped to the use of templates. Future work will determine if early structured feedback impacts notes written by these students as fourth-year acting interns.