194. Retrospective analysis of daily Tomotherapy output check performed by means of statistical process control: A helpful way for preventive maintenance interventions?

Purpose Statistical process control (SPC) is a tool widely used in industrial engineering to monitor and improve processes, allowing early detection and prevention of problems. We retrospectively applied SPC methods on daily Tomotherapy output check in order to assess the possibility to use it as predictive maintenance tool. Methods and materials From January 2014 to September 2017 a total of 1916 daily output check have been performed in static (958) and dynamic (958) modality, using the “cheese” phantom and a ionization chamber inserted into it. In the first case, a static field of 5 cm × 40 cm has been used, while an optimized IMRT plan was considered in the second case. All collected data have been organized per year, and for each year in subgroup of 30 data, representing approximately a month of output checks. Individual (I) and exponentially weighted moving average (EWMA) charts have been used to monitoring the long and short term data variability, respectively. The process performance was evaluated by means of capability indices using the Chan method. Results For each year considered in the study, the I- and EWMA-charts have shown the presence of out of control situation referable to specialist interventions. In particular, as reported in the figure below, the EWMA-charts have shown the presence of out of control situations before the interventions. Even if the tolerance suggested by AAPM TG 148 have been respected in the considered period of time, the process performance were in agreement with the scenario considered: the higher the interventions, the worst the results. Conclusion SPC methods are able to effectively detect variation in Tomotherapy output checks. Further investigation will consider on-line daily results.