Avoid, Trap, and Mitigate: Development of an Evidence-Based Crisis Management Framework in Surgery.

Abstract

Increasingly, surgeons are adopting broader roles in emergency response, on both clinical and executive levels. These have highlighted the need to develop healthcare-specific crisis management systems. Cross-professional research between safety-critical industries is a valuable method for learning crisis control. Commercial aviation, in particular, has been used to drive innovation in surgical safety. This study aimed to identify, adapt, and operationalize a surgical crisis management framework based on current practice in commercial aviation. A multimethod qualitative study interrogated safety experts in commercial aviation and healthcare. Stage I used immersive observational fieldwork in commercial aviation practice. Stage II performed semi-structured interviews with senior airline pilots. "Snowball" sampling targeted professional networks, recruiting 17 pilots from 4 airlines. Thematic analysis was used to derive a model of crisis management. Stage III undertook 3 focus groups with 5 pilots and 5 healthcare safety specialists. Expert consensus methods were used to adapt the model to clinical practice. Interview data provided 2,698 verbatim quotes on crisis management from aviation experts with a combined flying experience of 188,000 hours. Aviation crisis management was structured in 3 phases: avoid, trap, and mitigate. Adapted to clinical practice, these translated to crisis preparedness, recovery, and containment interventions. Additionally, the study identified 7 types of implementation tools and 9 crisis management skills that could be used to operationalize this framework in surgical practice. Surgical crisis management can follow the avoid, trap, and mitigate framework used in commercial aviation. Implementation relies on the combined use of crisis skills and performance tools. Crisis management should be delivered as part of a systems-based approach that relies on well-integrated failure management models. Simulation and in-situ validation of this framework is needed.