3618 - Evaluating Risk in Treating Prostate Cancer with Radiation Using Failure Mode and Effects Analysis
Presenter(s)
E. Byrd1, M. Foster2, E. Kwong3, A. Karunaker4, S. Sud5, S. A. Saraiya6, G. H. Goldin7, A. Wijetunga8, M. C. Repka5, B. M. Anderson5, S. K. Das9, L. B. Marks5, L. Mazur5, and R. McGurk5; 1UNC Chapel Hill, Chapel Hill, NC, United States, 2Division of Healthcare Engineering, UNC School of Medicine, Chapel Hill, NC, United States, 3University of North Carolina at Chapel Hill, Chapel Hill, NC, United States, 4University of North Carolina, Chapel Hill, Chapel Hill, NC, 5Department of Radiation Oncology, University of North Carolina, Chapel Hill, NC, 6Virginia Commonwealth University, Richmond, VA, 7University of North Carolina Hospitals, Chapel Hill, NC, United States, 8Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, 9University of North Carolina, Chapel Hill, NC
Purpose/Objective(s): Increasingly complex radiation therapy (RT) planning and delivery processes challenge conventional prescriptive quality management (QM) approaches, for example simple comparison of plan parameters against established tolerances. A more effective approach may be to analyze errors originating from deficiencies in workflows and processes. Failure mode and effect analysis (FMEA) is one common approach used to map processes, evaluate risk, and the effectiveness of mitigation strategies at each step in order to focus QI initiatives. We herein perform a FMEA for external beam radiation therapy of prostate cancer with the goals of a) identifying the process steps with the highest perceived risk and b) measuring the level of consensus between physician-experts.
Materials/Methods: Five radiation oncology physicians with expertise in prostate cancer at a US academic medical institution completed an FMEA of the workflow in place from March to November 2024 that was required to irradiate a patient with prostate cancer. The process steps and most likely failure mode for each step were initially pre-defined by a physicist using information extracted from an internal incident reporting system. Physicians assigned scores on a scale of 1-10 for the occurrence, severity, and detectability of each failure mode, and the product of these three scores was taken as the Risk Priority Number (RPN) for each step (on a scale of 1 to 1,000). The mean RPN scores across the five physicians were used to rank failure modes. The level of consensus between individual physician-raters was analyzed using the intra-class correlation coefficient (ICC) with a two-way mixed effects model evaluating for consistency.
Results: There were 173 process steps including 173 failure modes quantitatively scored by five physicians (865 assessments in total). RPN scores ranged from 2.6-253.2. Seven of the top ten RPN values were associated with treatment planning. RPN scores from the five physicians produced an ICC of 0.626 (95% CI: 0.505-0.725, p < 0.001), indicating moderate inter-rater reliability.
Conclusion: The FMEA suggests that the greatest risks occur during the treatment planning steps surrounding image segmentation (i.e. contouring), pre-treatment planning peer review, and plan review. These steps appear to represent the most fruitful targets for future QI initiatives.
Abstract 3618 - Table 1: Summary of top 3 process steps with the highest mean RPN values| Description | Occurrence | Severity | Detectability | RPN |
| Suboptimal review of image segmentation by MD (of targets or Organs at Risk [OARs]) | 5.6 | 6.2 | 5.6 | 253.2 |
| Comments from our pre-treatment planning Peer Review meeting not passed on to the responsible MD | 4.6 | 6.2 | 6.4 | 199.2 |
| Suboptimal review of treatment plan by MD | 4.8 | 6.2 | 5.0 | 161.4 |