Background Annually, close to 290,000 adult patients experience in-hospital cardiac arrest in the United States. Internal medicine residents in our program felt the need for code blue training and simulation sessions to run in-hospital cardiac arrest. We conducted a quality improvement project to enhance residents’ knowledge, readiness, and confidence in leading code blue.

Methods Residents had mock code simulation sessions every three months over 9 months (July/2023-April/2024) at Northwestern Medicine Huntley Hospital's simulation lab. In each session, residents were divided into two teams. These sessions aimed to be interactive, one team performed a scenario during each session while the other team observed, followed by a debriefing session to discuss what went well and what could be improved. The simulations focused on the roles of code blue leaders, rhythm recognition, and ACLS protocols. Pre-mock code session surveys were emailed to residents to assess their overall knowledge of ACLS protocol (medications, doses, ability to do other tasks like chest compressions, recording...) and confidence in leading code blue. Surveys included questions with a rating scale from (1-5). Following the mock code sessions, we sent post-session surveys that mirrored the initial ones, to reevaluate residents’ knowledge after the intervention. We used McNemar’s test to determine the significance of change between the 2 surveys before and after intervention.

Results Of the 39 residents in our internal medicine residency program, 86.8% felt that code blue simulations significantly helped them. A significant increase was observed in residents' knowledge of code blue leader roles with 42.1% feeling "Very knowledgeable" compared to 3% before (p = 0.05). Confidence in leading code blue also rose significantly with 44.7% feeling "Confident" compared to 18.1% before the intervention (p = 0.008). Knowledge of the ACLS algorithm improved significantly with 44.7% feeling "Very Knowledgeable" compared to 18.1% before (p = 0.025). Knowledge of ACLS medications and doses also increased significantly with 29% feeling "Very Knowledgeable" compared to none before (p = 0.011). While there was no significant change in interpreting rhythm strips during a code blue, more residents reported running code more frequently post-intervention (42% vs 30.3%).

Conclusion Most residents believed that code blue simulation sessions helped them run code blue confidently. Residents' confidence in leading code blue and knowledge of the ACLS algorithm increased after mock code simulation sessions. We anticipate that this project will positively impact patients’ outcomes following in-hospital cardiac arrest. Therefore, we plan to continue conducting these sessions annually for incoming interns.