Moving from IMRT QA measurements toward independent computer calculations using control charts

Background and purpose: In this study, we investigated IMRT QA using Statistical Process Control for the purpose of comparing the processes of patient-specific measurements and the corresponding independent computer calculations. Materials and methods: Point dose data from the treatment planning system (TPS), independent computer calculations, and physical measurements for prostate and head and neck cases were studied. Control charts were used to analyze the IMRT QA processes from several institutions in the academic and community setting. Control charts are a method to describe the performance of a process. The width of the control chart limits (or action limits) describes the process' ability to meet clinical specifications of ±5%. In all, 24 process comparisons were made (12 measurement QA and 12 independent computer calculation QA). Results: For head and neck IMRT QA, the average process ability for the measurement QA was ±6.9% compared to ±7.2% for the independent computer calculation QA. For prostate IMRT QA, the average process ability was 4.4% for both measurement QA and independent computer calculation QA. It was found that 11 of the 24 processes were in control. At none of the institutions were the processes of measurements and independent computer calculations both in control and performing within clinical specifications. Conclusion: There is room to improve the processes of IMRT QA measurements and independent computer calculations. In situations where the improvement of the processes is such that each is in control and well within clinical specifications, it may be appropriate to suspend patient-specific IMRT QA measurements for every patient in the place of independent computer calculations