Click here to


Are you sure ?

Yes, do it No, cancel

Can We Predict Radiotherapy Errors Based On Past Events?

D.Gilley, IAEA Vienna, R.Lopez, Mexico, Comision Nacional De Sequridad Nucleary Salvaquardias, Madrid


(Sunday, 7/14/2019) 4:00 PM - 5:00 PM

Room: 301

Purpose: To identify improvements in radiotherapy safety by combining the results of a retrospective analysis of events in an incident learning system and the use of a probabilistic risk assessment approach for radiotherapy.

Methods: To forecast errors in nuclear applications, an international group of regulatory authorities developed a self-evaluation tool to analyse accidents and events by determining the initiating event that lead to the report: review the sequence of events to determine what failed in the sequence leading to the event and to identify what are the possible safety measures and the consequences and use a risk matrix to reduce the frequency of both the event and the consequences; and test the self-evaluation tool on an independent radiotherapy incident learning system. The mapping of radiotherapy events was performed by radiotherapy and nuclear engineering professionals, to test the risk matrix self-evaluation tool. Each radiotherapy event was scored by each reviewer based on the frequency reducer, safety barrier and consequences. Any event that was not supported by the review team was discarded.

Results: 1,183 events were analysed by medical physicists, nuclear engineers and one physician; 886 were identified as potential for mapping to the risk matrix. Further screening and review of 886 of these indicated that 583 could, with good correlation, be mapped into the risk matrix. The first 549 events have been coded in the incident learning system and radiotherapy professional can use this link to assess the adequacy of their safety systems to prevent a certain type of error.

Conclusion: This is the first known attempt to use an incident learning system to predict the potential for errors in a radiotherapy. The results demonstrated that for some events the reviewer can predict the potential errors and identify safety barriers that can eliminate or reduce the consequences.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email