78. PERFORMANCE UNDER PRESSURE: THE USE OF EPAS TO STANDARDIZE PEDIATRIC MOCK CODE FEEDBACK

Abstract

Background Pediatric mock code simulation plays a critical role in the development of a trainee's decision making ability for high-stakes, low frequency situations. To ensure trainees receive maximal benefit from these simulations, it is important that they receive optimal, timely feedback. Objective The aim of this study is to characterize the implementation and pilot use of the American Board of Pediatrics’ (ABP) Entrustable Professional Activities (EPAs) for mock code feedback. Methods We designed a simulation-based education intervention study of pediatric trainees at a large pediatric residency program. To ensure consistency and breadth of experience, a standardized mock code curriculum of 13 cases was created. An online assessment tool was created based on the ABP's EPA 10 (resuscitation and initial stabilization) incorporating objective clinical behaviors that the learner must display during the simulation. EPAs and communication were rated based on narrative anchors with seven and five possible levels, respectively. The tool was used to facilitate and record feedback between the instructor and trainee. In addition, other mock code participants evaluated the trainee using the tool in real time. Results 6 months since implementation, 21 residents (45% of PGY-2, PGY-3 classes) have received one-to-one verbal and written feedback using our tool following a mock code. 90 evaluations were completed by faculty and observers. The average EPA rating and communication score of residents was 5.4 (SD 1.27) and 3.66 (SD 0.75), respectively. There were no significant differences observed in the EPA or communication scores between the observers and faculty (P = 0.46 and 0.8, respectively). In the post-simulation survey, 93% (13/14) of trainees indicated the feedback process would improve their future performance. Conclusions An EPA-based, online assessment tool was feasible and easily incorporated into mock code simulations. We found that observers with varying levels of expertise gave similar ratings, suggesting future possibilities of using peers for feedback in mock code settings.