Learning from errors boosts QA in radiation therapy

Establishing a successful radiotherapy quality assurance (QA) program and a department-wide culture of safety is challenging. Sustaining it can be achieved with an incident learning system, as explained in an article published online in Practical Radiation Oncology.

The Ottawa Hospital in Ottawa, Ontario, has deployed an incident learning program since 2007. Over a five-year period from January 2007 through December 2011, its radiation oncology department performed a total of 345,792 individual treatments and had an incident rate of 0.7% (Pract Radiat Oncol, September 7, 2012).

While the number of nonminor incidents has declined, medical physicist Brenda Clark, PhD, who oversees the program with radiation therapist Robert Brown, does not expect them to disappear. She thinks this would indicate complacency rather than perfect quality assurance performance.

"There are so many new treatment techniques, software updates, and new technology introduced in the radiation oncology department that incidents are bound to occur," she told AuntMinnie.com. "The only way to keep a quality assurance program abreast with all the changes is to identify mistakes, analyze why they occurred, and make changes to prevent them."

The elements of a successful incident learning system include senior management support, a nonpunitive culture, diligent incident reporting, investigation of incidents to determine their basic causes, and identification of strategies to prevent reoccurrence and to facilitate learning. All elements require communication. Regular feedback is also provided to the program staff, emphasizing improvements that have been made as a result of following up an incident report. This feedback is publicized within the department, reinforcing the idea that safety is important because it generates positive change.

All members of the department are encouraged to report any incident. Most incidents are caused by manual activities, and a consistent 70% to 80% of incidents relate to standard procedures -- specifically, a lack of or inadequate procedures, or failure to follow them precisely. Issues relating to communication account for 7% to 9% of incidents.

Work-planning issues have risen fourfold, from 2.9% in 2007 to 11.6%. This could be due to instructions for patient setup and placement being manually transcribed into the information system from the treatment plan, according to Clark and colleagues. In fact, 35% of the highest-severity incidents were caused by incorrect patient setup.

The introduction of new software programs or updates or new equipment typically generates a spike in incidents that declines as individuals become experienced with them, or as the QA team identifies and rectifies factors that generate incidents. Some treatment units are more error prone, such as those that treat a variety of sites, or units with a staff that rotates regularly. Treatments in the tomotherapy units at the Ottawa center produce the fewest errors.

Approximately eight to 12 incidents are reported each week, and whatever their level of severity, all are investigated. Investigated incidents are discussed at a weekly multidisciplinary meeting. The causes of incidents are analyzed, and steps are taken to make positive changes.

"Medical physicists have intuitive skills to solve problems," Clark said. "Our jobs require constant vigilance and attention to detail. A single error made on the part of a medical physicist has the potential to affect a large number of patients. I think our profession is ideally suited to champion implementation of incident learning programs which provide valuable feedback to a radiation oncology department's quality assurance activities."

Page 1 of 461
Next Page