ESTRO 36 Abstract Book

513 / 1082

S498 ESTRO 36 2017
_______________________________________________________________________________________________
deliveries, slightly higher standard deviation was observed
than IMRT.
Figure 1- Planar dose assessment of the centres for head
and neck and post-prostatectomy plans: a) IMRT delivery,
b) VMAT delivery
Figure 2- 3D dose assessment of the centres for head and
neck and post-prostatectomy plans and: a) IMRT delivery,
b) VMAT delivery
Conclusion
All linacs were equipped with EPIDs so a consistent
detection system was used by the centres. The method
was significantly less expensive and faster than
conventional audits due to its remote nature and use of
virtual phantoms. All measured data were analysable with
relatively high pass rates. Interactive communications
with centres was often necessary to ensure quality data
were provided.
PO-0908 Application of Failure Mode and Effects
Analysis to linac quality controls: advantages and limits
F. Bonfantini1, T. Giandini1, S. Meroni1, C. Stucchi1, M.
Carrara1, V. Mongioj1, I. Veronese2, E. Pignoli1
1 Fondazione IRCCS Istituto Nazionale dei Tumori, Medical
Physics, Milan, Italy
2 Università degli studi, Physics, Milan, Italy
Purpose or Objective
The increased complexity of the modern linac-based
radiotherapy requires more thorough quality assurance
programs to reduce the risk of errors and ensure patient
safety. However, these demands are cumbersome and the
efforts should be optimized in order to take maximum
advantage of the available resources. In this context,
prospective methods for risk analysis, such as Failure Mode
and Effects Analysis (FMEA), can be a useful tool. Aim of
this work was to evaluate advantages and limits of the
application of FMEA for the optimization of linac quality
controls (QCs).
Material and Methods
Each parameter tested by the QC was considered as a
potential failure mode (FM) and a Risk Priority Number
(RPN) was calculated from the product of three indexes:
likelihood of occurrence (O), severity of effect (S) and lack
of detectability (D). Forty tests were examined just above
the expected tolerance levels and indexes O, S, and D
were scored from 1 (lowest risk) to 10 (highest risk) using
two methods:
1) A survey was submitted to each of the medical
physicists of our institute involved in the linac QC
2) The QC data over a period of three years were analyzed
and some FMs were simulated with the treatment planning
system.
The average RPN for each test was obtained taking into
account both the methods. For each linac, the tests were
then sorted by their frequency (daily, monthly or annual)
and RPN value.Two different Varian linacs (DBX, Unique)
were considered, the first used only for conformal therapy
and the second one used essentially with volumetric
modulated arc therapy (VMAT) technique.
Results
A high variability was found in the O-D-S scores of the
survey, as shown in the box plots of figure 1 for the
dosimetric tests of the Unique linac. Nevertheless, a lower
variability was obtained for RPNs, highlighting at the same
time the more relevant tests.
Both the FM simulations and the analysis of the QC trend
allowed to reduce the subjectivity of the FMEA score.
Integration of both evaluations provided the RPN-based
ranking of tests: an example is shown in figure 2 for
monthly tests for DBX and Unique linacs.