Black Swans in Radiation Therapy
Presentations
PO-GePV-T-213 (Sunday, 7/10/2022) [Eastern Time (GMT-4)]
ePoster Forums
Purpose: A Black Swan as a highly improbable event that carries extreme impact, and hindsight, should have been obvious as to why it occurred. In everyday life, Black Swans have presented themselves as events such as 9/11, the success of Google, and the financial crisis of 2008. The purpose of this work is to analyze Black Swan events that have occurred in radiation therapy and ultimately make our current processes more robust by applying lessons from the past to increase focus on what we don’t know going forward.
Methods: The international incident learning and reporting system SAFRON, Safety in Radiation Oncology, database was used to analyze radiation therapy incident reports from 1986-2022. The entire database (1767 reports) was filtered to 65 incident reports using the SAFRON severity criteria of major and critical events for each of three modalities (external beam, brachytherapy, and radionuclide therapy). The 65 remaining reports were ranked based on criteria of a Black Swan event that, by definition, is highly improbable, carries extreme impact, and is retrospectively obvious after it occurs.
Results: Each Black Swan characteristic (probability, impact, and retrospective predictability) was scored on a scale from 1 to 5. The product of these three scores was determined to be the events’ Black Swan score. Incidents with a Black Swan score greater than or equal to 84 were considered to be true Black Swan events. There were a total of 4 Black Swan events out of the 65 remaining reports once scoring was complete.
Conclusion: Risk analysis based on current knowledge of a process has been instrumental in facilitating patient safety in recent years. Moving forward, it is important to develop methods and analysis techniques that will increase the overall robustness of the radiation therapy process without waiting for unexpected failure modes to appear.
Keywords
Risk, Radiation Therapy, Quality Control
Taxonomy
IM/TH- Formal Quality Management Tools: General (most aspects)
Contact Email