Objectives: The purpose of this study is to apply failure mode and effects analysis (FMEA) to intracranial stereotactic radiosurgery (SRS) to develop processes which reduce the risk of the procedure. By thoroughly examining each step of the different types of SRS (single or multiple isocenter, single or multiple fraction, cone or MLC collimated) we will generate potential failure modes which will lead to process improvements. These process improvements will reduce the risk of each type of SRS treatment.

Methods: In our clinic SRS treatments are performed on a linear accelerator and are collimated by either cones or multileaf collimators. We treat patients both with single- and multi-fraction regimens and deliver both single- and multi-isocenter treatments. The SRS team at our clinic analyzed our program using FMEA techniques. The group broke the entire process into 13 primary processes consisting of approximately 150 secondary steps. Each of these steps was then analyzed to determine possible failure modes. Each failure mode was assigned a risk probability number (RPN) which is the product of three factors: severity, frequency and detectability. We then binned all the RPNs into 'High', 'Medium' or 'Low' risk groups to determine how to reduce the riskiest parts of the procedure.

Results: Our intracranial SRS program was broken down into 13 individual steps which were then divided further into approximately 150 discrete processes. Based on their severity, frequency and detectability, an RPN was calculated for each of these individual processes. The processes with the highest RPN were further analyzed to determine how to reduce the risk to acceptable levels. We saw notable distinctions in RPN between cone-based and MLC-based collimation treatments, between single-isocenter and multiple-isocenter treatments and between single-fraction and multiple-fraction regimens. Risk was highest for cone-based single-fraction multiple-isocenter treatments and new processes were developed to ensure that all treatment parameters were confirmed before treatment was delivered. Noting the distinctions between the highest risk parts of an SRS treatment process and other lower risk parts, we developed new processes to reduce the risks to manageable levels.

Conclusions: Failure mode and effects analysis was used to analyze our intracranial stereotactic radiosurgery program. The analysis showed portions of our processes that had higher than acceptable risk of deviation from expectation. New processes were developed to address the processes with the highest risk to ensure that each patient is treated accurately and safely.