Radiation oncology as a complex adaptive system

Radiation oncology (RO) is increasingly recognized as a complex system characterized by the intricate interplay of professionals, patients, technology, and environmental factors. However, systematic methods to assess the interconnectivity in RO remain limited, leading to potential oversimplifications of this multifaceted field. This study quantitatively assesses evolving complexity and nonlinearity in radiation oncology using system-thinking tools. The goal is to create a measurable framework to guide adaptive management in clinical practice.

Using Shannon Entropy, we analyze the evolving complexity in RO. Dynamic system simulations, including adapted predator-prey models and the SimPy discrete-event simulation library, are utilized to study nonlinearity and interactions within RO. Process mining with event log data assesses the conformance of RO processes, while social network analysis explores self-organization among RO actors.

Our findings reveal a marked increase in the complexity and nonlinearity of RO. Simulations and process mining demonstrate emergent behavior, self-organization, and adaptability within the system.

The inherent nonlinearity, complexity, emergence, adaptability, and self-organization in RO systems validate the view of RO as a complex system. This insight calls for a shift towards 'system thinking' in managing, leading, and operating within the RO field. This approach will better accommodate the dynamic, interconnected nature of RO, ensuring more effective and adaptable healthcare outcomes.