Benchmarking failure mode and effects analysis of electronic brachytherapy with data from incident learning systems

Area of Science:

• Medical Physics
• Radiation Oncology
• Risk Management

Background:

• Failure Modes and Effects Analysis (FMEA) is a critical prospective risk assessment tool.
• It is used to identify potential equipment and process failures, informing quality control system design.
• Electronic brachytherapy (eBT) is an advanced radiation therapy technique requiring robust safety protocols.

Purpose of the Study:

• To benchmark the predictive performance of FMEAs for electronic brachytherapy (eBT) of the skin and breast.
• To compare predicted failure modes from FMEAs with actual failure modes reported in incident learning systems (ILS).

Main Methods:

• Two public and one internal ILS were queried for Xoft Axxent eBT-related events over a 9-year period.
• FMEA failure modes and Risk Priority Numbers (RPNs) from prior analyses of skin eBT and breast intraoperative radiation therapy (IORT) were utilized.
• Each reported event's treatment site and primary failure mode were compared against the corresponding FMEA data.

Main Results:

• A total of 49 Xoft eBT events were identified: 31 (63.3%) for breast IORT and 18 (36.7%) for skin.
• The primary failure mode for 87.7% of all events ranked within the top 10 by RPN in the FMEA.
• For skin and IORT events specifically, failure modes ranked in the top 10 by RPN or severity in 83.3% and 90.3% of cases, respectively.

Conclusions:

• FMEA is effective in predicting failure modes for eBT, though its accuracy can be influenced by user experience.
• Incident learning systems (ILS) data are valuable for enhancing FMEA by identifying potential failure modes.
• ILS data can also inform more realistic estimations of occurrence, detectability, and severity values within FMEA.