1. IAEA E-learning Program
Safety and Quality in Radiotherapy

2. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Sections:
9.1 Purpose and principles
9.2 Treatment planning
9.3 Information transfer
9.4 Calibration
We will first look at the purpose and principles of Fault Tree Analysis and then perform demonstration Fault tree Analyses for treatment planning, Information transfer and calibration failures.

3. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Sections:
9.1 Purpose and principles
9.2 Treatment planning
9.3 Information transfer
9.4 Calibration
In section 3 of this module we’ll develop a possible Fault Tree around an information transfer error. Information transfer was at least part of the problem leading up to the Epinal incident which we discussed in detail in Module 2 of this e-learning program. It’s worth remembering that we are exploring a hypothetical situation. A Fault Tree developed for information transfer errors in any specific clinic could look quite different.

4. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
OBJECTIVES
To develop a fault tree describing a potential error in the process of information transfer.
To suggest a few systemic, organizational issues which, if addressed, could minimize the probability of such errors.
To reinforce the need for quality control as a component of a critical process.
The section objectives are: to develop a fault tree describing a potential error in the process of information transfer, to suggest a few systemic , organizational issues which, if addressed, could minimize the probability of such errors, and to reinforce the need for quality control as a component of a critical process.

5. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
A High Level Process Map1
For  argument sake we will associate the information transfer error with Treatment Delivery although we recognise that such an error relates to the link between any 2 or more steps in the process.

6. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Modes and Effects Analysis: Treatment Planning
Here’s the Failure Modes and Effects Analysis that we developed around information transfer in Section 8.3.

7. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
A Fault Tree
Event/situation/circumstance
AND OR
And this is the Fault Tree we want to complete for one Failure Mode and one Failure Pathway.

8. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Mode
A dynamic wedge is used for treatment when the planned monitor units have been calculated for a hard wedge.
Failure Pathways
Staff had not been informed of the new practice with wedges.
Planning staff did not fully understand how to use the treatment planning system.
Therapists on the unit did not fully understand how to interpret the parameters to be set on the machine.
These are the Failure Pathways we identified in the previous Module. Let’s pick the first one for our example. Staff had not been informed of the new practice with wedges.

9. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Calgary Basic Cause Table2
As before, we’ll fill the right most boxes with causes/contributory factors from this table.

10. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Event/situation/circumstance
AND OR
Wrong wedge factor used
First, we’ll enter in the left most box the actual Failure we are simulating. This is, of course, based on the Epinal incident, but we need to think broadly about what might lead to a wrong wedge factor being used in any clinic.

11. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Event/situation/circumstance
AND OR
Wrong wedge factor used
Plan with hard wedge
The next step is to speculate on what events could lead up to such a Failure. Possible events are shown here. The plan might be generated incorporating a hard wedge, hence prescribing monitor units associated with such a hard wedge, and yet be delivered with a soft or dynamic wedge.
Tx with soft wedge

12. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Plan with hard wedge
Tx with soft wedge
Event/situation/circumstance
AND OR
Wrong wedge factor used
Few trained staff
Instructions unintelligible
Plan with hard wedge
Following the top branch, it could be that there were few staff trained in planning with dynamic wedges on the treatment planning system. Another possibility is that the instruction manual is hard to understand. Either of these could have led to a plan being inadvertently created with a hard wedge when the policy was to use dynamic wedges. As either of these situations could lead to such an error we should join
 them with an “OR” gate.
Tx with soft wedge

13. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Plan with hard wedge
Tx with soft wedge
Event/situation/circumstance
AND OR
Wrong wedge factor used
Few trained staff
Instructions unintelligible
6.1 Inadequate training
1.1 Procedure not developed
Plan with hard wedge
We can straight away suggest some Basic Causes that could precipitate such an error. If there were too few trained staff then the training program has to be viewed as inadequate. We can describe the circumstance of unintelligible instructions as being equivalent to not having a procedure at all. Both these Causes are actionable.
Tx with soft wedge

14. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Plan with hard wedge
Tx with soft wedge
Event/situation/circumstance
AND OR
Wrong wedge factor used
Few trained staff
Instructions unintelligible
6.1 Inadequate training
1.1 Procedure not developed
Plan with hard wedge
Looking now at the lower branch, we can suggest 2 causes for the circumstance of treating with a dynamic wedge when the plan is based on a hard wedge calculation.
4.1 Inadequate work planning
Tx with soft wedge
4.2 Inadequate change management

15. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Event/situation/circumstance
AND OR
Plan with hard wedge
Tx with soft wedge
Few trained staff
Instructions unintelligible
6.1 Inadequate training
1.1 Procedure not developed
4.1 Inadequate work planning
4.2 Inadequate change management
Wrong wedge factor used
Another way of looking at it is to say that none of the people involved were adequately trained and inadequate change management was present throughout the process. In which case we might prefer to develop the Fault Tree like this. At the end of the day we have the same Basic Causes and would likely adopt similar preventive actions.

16. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Event/situation/circumstance
AND OR
Plan with hard wedge
Tx with soft wedge
Few trained staff
Instructions unintelligible
6.1 Inadequate training
1.1 Procedure not developed
4.1 Inadequate work planning
4.2 Inadequate change management
Wrong wedge factor used
Whichever representation we choose will guide us to system changes, such as better documentation and more training, which are expected to make the whole process safer. However, even with these organizational improvements there is still no guarantee that errors will never occur. Any improvements implemented are expected to reduce the probability of error but not to eliminate it. So what quality control measures might we think about introducing?

17. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Event/situation/circumstance
AND OR
Plan with hard wedge
Tx with soft wedge
Few trained staff
Instructions unintelligible
6.1 Inadequate training
1.1 Procedure not developed
4.1 Inadequate work planning
4.2 Inadequate change management
Wrong wedge factor used
One option likely to be effective in a situation like this is in vivo dosimetry. A major dosimetric error is likely to be flagged by using diodes, which can be read in real time, or thermoluminescence dosimetry (TLD). However, general implementation of in vivo dosimetry does negatively impact the throughput of patients on a machine.
QC: in vivo dosimetry

18. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
Failure Pathway: “Staff had not been informed of the new practice with wedges”
Event/situation/circumstance
AND OR
Plan with hard wedge
Tx with soft wedge
Few trained staff
Instructions unintelligible
6.1 Inadequate training
1.1 Procedure not developed
4.1 Inadequate work planning
4.2 Inadequate change management
Wrong wedge factor used
QC: planner
check
Another option would be to have whoever generated the plan be present during the first set up to ensure that all instructions are clear, accurate and understood. This concludes our analysis of this possible Failure Pathway. As we’ve mentioned several times both FMEAs and FTAs are clinic dependent so we can’t, in general, take an analysis developed in a different situation and simply import it into our situation.
QC: in vivo dosimetry

19. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
A few final notes
Basic causes identified in the right most boxes of a fault tree do not, in general, make failure inevitable.
To be useful, basic causes should be actionable.
Addressing the basic causes identified with a fault tree should reduce the occurrence value in the accompanying failure modes and effects analysis.
Incorporating quality control into a fault tree should reduce the detectability value in the accompanying failure modes and effects analysis.
Always be aware that changing processes, including QC, has the potential to introduce other sources of error.
A quick reminder of some of the key points regarding Fault Trees. We’ve seen this before.

20. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
summary
We have:
Developed a fault tree describing a potential error in the process of information transfer.
Suggested a few systemic, organizational issues which, if addressed, could minimize the probability of such errors.
Reinforced the need for quality control as a component of a critical process.
To summarize, we have
Developed a Fault Tree describing a potential error in the process of information transfer.
 Suggested a few systemic, organizational issues which, if addressed, could minimize the probability of such errors
Reinforced the need for quality control as a component of a critical process.

21. Safety and Quality in Radiotherapy
MODULE 9: fault tree analysis
Safety and Quality in Radiotherapy
Section 3: information transfer
References and additional resources
E Ford, L Fong de los Santos, T Pawlicki, et al.. Consensus recommendations for incident learning database structures in radiation oncology. Medical Physics 39 (2012)
Cooke DL, Dubetz M, Rahim H, et al
We’ve seen these references before.