Enhancing Physician Satisfaction and Patient Safety Through an Artificial Intelligence-Driven Scheduling System in Anesthesiology.

Abstract

Background: Overcoming challenges to effective clinical practice depends on finding dynamic solutions to issues such as physician burnout and patient safety. This study evaluated the impact of an artificial intelligence (AI)-driven scheduling system on physician burnout and patient safety, using intraoperative transitions of care as an operative metric for patient safety.

Methods: In May 2019, the Department of Anesthesiology at Ochsner Health in New Orleans, Louisiana, implemented an AI-driven scheduling system called Lightning Bolt Scheduling (PerfectServe, Inc). Utilizing an idealized design framework, the department steering committee analyzed 12 months of historic operating room data and developed more than 400 scheduling rules to optimize staffing. The scheduling rules, representing the steering committee's new work model, were provided as inputs into Lightning Bolt Scheduling, which then used combinatorial optimization to recommend an ideal staff schedule. Preimplementation and postimplementation data were collected on physician satisfaction, vacation approval rates, and intraoperative transitions of care.

Results: Six months postimplementation, physician satisfaction scores and vacation approvals increased, reflecting improved work-life balance, schedule flexibility, and decreased symptoms of burnout. Additionally, 1,072 fewer handoffs occurred, equating to 71.5 fewer adverse events and a savings of approximately $335,550 in health care costs during the 21 months after implementation.

Conclusion: Our study findings underscore the potential of data-driven scheduling systems to enhance physician well-being and patient safety, thereby promoting continuous improvement and adaptability in health care operations.