Development and Establishment of Initial Validity Evidence for a Novel Tool for Assessing Trainee Admission Notes

Documentation is a key component of practice, yet few curricula have been published to teach trainees proper note construction. Additionally, a gold standard for assessing note quality does not exist, and no documentation assessment tools integrate with established competency-based frameworks.

To develop and establish initial validity evidence for a novel tool that assesses key components of trainee admission notes and maps to the Accreditation Council for Graduate Medical Education (ACGME) milestone framework.

Using an iterative, consensus building process we developed the Admission Note Assessment Tool (ANAT). Pilot testing was performed with both the supervising attending and study team raters not involved in care of the patients. The finalized tool was piloted with attendings from other institutions.

Local experts participated in tool development and pilot testing. Additional attending physicians participated in pilot testing.

Content, response process, and internal structure validity evidence was gathered using Messick's framework. Inter-rater reliability was assessed using percent agreement.

The final tool consists of 16 checklist items and two global assessment items. Pilot testing demonstrated rater agreement of 72% to 100% for checklist items and 63% to 70% for global assessment items. Note assessment required an average of 12.3 min (SD 3.7). The study generated validity evidence in the domains of content, response process, and internal structure for use of the tool in rating admission notes.

The ANAT assesses individual components of a note, incorporates billing criteria, targets note "bloat," allows for narrative feedback, and provides global assessments mapped to the ACGME milestone framework. The ANAT can be used to assess admission notes by any attending and at any time after note completion with minimal rater training. The ANAT allows programs to implement routine note assessment for multiple functions with the use of a single tool.