O2053: The Autonomy Toolbox: a Multi-center Collaborative to Improve Resident Autonomy

Purpose/Objectives: Clinical autonomy is necessary for the professional development and well-being of resident trainees. An increased focus on patient safety in recent years has led to increased supervision of residents in hospital wards and a decrease in their autonomy. Despite multiple studies describing its importance, there are few validated measures to improve resident autonomy.

Design/Methods: In this quality improvement project, we developed a bundled intervention approach to improve senior resident (SR) autonomy on the Pediatric Hospital Medicine (PHM) service at 5 academic children’s hospitals. A needs assessment survey regarding the level of autonomy on PHM service was sent to SR and PHM faculty to define areas of improvement. We chose questions in which PHM faculty and SR had discordant perceptions of SR level of autonomy as targets for our interventions. SR answers to these questions were converted to a 5-point Likert scale and scores were combined to form the Resident Autonomy Score (RAS). Our goal was to increase the RAS from 3.67 to 4.5 within 1 year and sustain for 6 months. We instituted resident and faculty development, formalized expectation-setting huddles, and established SR independent rounding as our interventions (Figure 1). We built in flexibility to our approach to allow each site to modify interventions to fit their needs.

Results: 232 of 427 SR and 129 of 220 PHM faculty completed at least part of the needs assessment survey. PHM faculty perceived greater SR autonomy compared to SR on 6 questions. Question topics included: 1) how seriously attending takes SR input, 2) if SR are encouraged to make straightforward medical decisions, 3) if attendings allow SR to choose the plan even if they prefer an equivalent alternative 4) if SR receive constructive feedback 5) if SR are the leader of the medical team, and 6) if attending oversight is “just right”. The baseline RAS of 3.67 increased to 4.36 (19% increase) 1 month after the start of this QI project (Figure 2). This occurred simultaneously with SR and faculty development. While further increases were not seen with expectation-setting and independent rounding interventions, this initial increase was maintained throughout the study period.

Conclusion/Discussion: SR and faculty perceive discordant levels of SR autonomy. Making faculty aware they are hindering autonomy can lead to improvements in the level of autonomy granted to SR. We created a highly-adaptable bundled intervention approach to increase and maintain SR autonomy.