Does Entrustment Match Performance? A Multi-institutional Simulation Study to Evaluate Validity of Entrustable Professional Activity-based Assessment for Neurocritical Care Advanced Practice Providers
Daniel Harrison1, Christa O'hana Nobleza2, Matthew Bevers3, Sahar Zafar4, Ariel Nowicki3, Elizabeth Woods4, Kelly Peronti5, Erica Perets6, Erika Sigman5, Rashid Ahmed7, Casey Albin5
1Boston Medical Center, 2Baptist Memorial Health Care, 3Brigham and Women's Hospital, 4Massachusetts General Hospital, 5Emory Healthcare, 6Grady Memorial Hospital, 7Upstate University Hospital
Objective:

To assess the relationship between checklist-based performance in simulated neurologic emergencies and supervisor- and self-assessed entrustment among neurocritical care (NCC) advanced practice providers (APPs).

Background:

APPs are key contributors to NCC teams. Six core and 47 nested entrustable professional activities (EPAs) for NCC APP training have been developed. These could be leveraged for EPA-based assessment, however, are not supported by validity evidence in this context.

Design/Methods:

Two simulation scenarios were created to assess performance related to six core and 15 nested-EPAs. The simulations’ critical action checklists were iteratively developed based upon guideline recommendations and consensus opinion from five collaborating institutions. Prior to participation, NCC APPs self-assessed their entrustment on the core and nested EPAs (the “EPA Assessment Tool”). APP supervisors completed the same assessment. After completing a rater training exercise, study staff completed the critical action checklists for each participant.

Results:

22 NCC APPs from 4 institutions participated in the study. There was a moderate positive association between physician supervisor assessed entrustment and performance on the critical action checklist (r= 0.47). There was a very strong positive association between physician supervisor assessed entrustment and participant NCC experience (r= 0.84). There was fair agreement in 393 physician and self-assessed EPAs (kw= 0.39). Overall entrustment was similar when assessed by physician supervisors (4.13 +/- 0.72 [5-point entrustment scale mean +/- standard deviation]) and participants (4.11 +/- 0.60, p= 0.83). For the subset of participants who were additionally assessed by an APP supervisor, overall entrustment was also similar between self- and supervisor-assessment (4.04+/-0.59 vs 4.20+/-0.73, p= 0.39).

Conclusions:
There was moderate correlation between physician supervisor EPA-assessment and performance in the simulation. Physician supervisor assessments agreed fairly well with APP self-assessments. These results provide validity evidence for EPA-based assessment of NCC APPs. EPA self-assessment may be a reasonable strategy for identifying strengths and learning gaps in NCC APPs.
10.1212/WNL.0000000000212324
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.