The Impact of High-Frequency and Low-Fidelity Cardiac Arrest Simulation on Medical Residents' Confidence and Simulated Performance.

Purpose: Advanced cardiovascular life support (ACLS) certification is currently required for physicians every 2 years. Data suggests this alone is insufficient to maintain high-quality resuscitation performance; however, the optimal approach to develop cardiac arrest skills remains unknown.1 During residency training, simulation is increasingly used to develop cardiac arrest leadership skills. These simulations exist on a spectrum from low fidelity (low cost, low level of realism) to high fidelity (high cost, high level of realism).2 Some research suggests that high-fidelity simulation results in higher learner satisfaction and confidence scores when compared with low-fidelity simulation; however, there does not seem to be a difference in knowledge or practical skill acquisition between these simulation strategies.3 High-fidelity simulation is both time- and cost-intensive, and is often performed outside of the clinical environment, posing an additional barrier to implementation.2 The evidence for similar efficacy, lower resource burden, and opportunity for more frequent simulation experiences has sparked increased interest in low-fidelity simulations. We aimed to evaluate the impact of high-frequency, low-fidelity cardiac arrest simulation training on resident confidence and skill retention. Method: We developed and piloted a high-frequency, low-fidelity cardiac arrest simulation curriculum for internal medicine residents rotating in the intensive care unit. During the monthly, 1-hour session, residents complete 3 simulations. Each simulation is followed by a structured debrief and a short didactic emphasizing interventions that improve patient outcomes, including high-quality CPR and early defibrillation. Additional topics include leadership, communication, and how to best discontinue resuscitation efforts. Immediately before and upon completion of the 3 simulations, self-assessments of resident confidence were obtained using a 5-point Likert scale questionnaire. During the first and last simulation, resident performance of technical ACLS skills (including time to first shock, CPR quality, and appropriate implementation of ACLS algorithms) as well as 6 leadership and communication skills were quantitatively assessed using a standardized rubric. Results: A total of 9 simulations involving 72 residents have been completed to date. Following low-fidelity simulation, more residents reported feeling “somewhat confident” or “very confident” (termed “high confidence”) in various domains: code leadership skills (31%–72%), ability to provide feedback on CPR quality (41%–69%), appropriate ACLS medication administration (38%–62%), initiation of post-ROSC care (34%–76%), and making the decision to discontinue resuscitative efforts (14%–69%). Additionally, 100% of residents reported the simulation improved their understanding of cardiac arrest resuscitation. Quantitative assessment of resident performance also demonstrated marked intrasimulation improvement with decreased time to first shock (114 to 50 seconds, P = .0054), increased compression fraction, defined as seconds of chest compressions divided by total resuscitation time (82% vs 93%, P = .0011), and improved leadership and communication score (3.2 to 5.8 out of 6, P = .0043). Discussion: Our data suggests a 1-hour, low-fidelity cardiac arrest simulation not only increased resident confidence in leading cardiac resuscitation but also resulted in a statistically significant improvement in resident performance during cardiac arrest simulations. This pilot demonstrates the feasibility of implementation of low-fidelity simulation in the clinical environment, resulting in cost and time savings still with proven benefit on confidence and skill acquisition. Future research is needed to evaluate the short- and long-term retention of performance gains and optimal simulation frequency to prevent skill decay, in addition to understanding the impact of simulation on in-hospital cardiac arrest outcomes. Significance: These findings highlight the utility of low-fidelity simulations in improving resident confidence and simulated performance in leading cardiac resuscitation.