Factors associated with radiation therapy incidents in a large academic institution

This study evaluated factors associated with radiation therapy (RT) planning and delivery incidents at a large academic institution.

The RT incidents (including near-misses) were recorded using an electronic incident reporting system from April 1, 2011 to April 30, 2013. Each incident's origin was categorized according to the step in the treatment process (simulation, physician prescription, treatment planning, scheduling, treatment delivery, and other) in which it occurred. The incident database was linked to the RT delivery (record and verify) database to evaluate the effect of various factors on the rate of RT incidents.

There were 189 reported RT incidents (including near-misses) among 326,448 fractions, of which there were 70 (37%) treatment planning incidents and 56 (30%) treatment delivery incidents. The rates of total incidents, planning incidents, and delivery incidents were 136.0, 50.4, and 40.3 per 10,000 patients, respectively. Logistic multivariate analysis showed that fewer work days from plan approval to treatment start, fewer fractions, higher number of prescription items, and longer beam duration were significantly associated with radiation planning incidents. Multivariate analysis also showed that first day of treatment, fewer fractions, higher number of prescription items, and longer beam duration were significantly associated with treatment delivery incidents; intensity modulated radiation therapy was associated with a lower rate of treatment delivery incidents.

More complicated radiation plans, fewer fractions, first day of treatment, and rushed processes were associated with higher risk of RT incidents. We hope that a national incident reporting database will lead to greater understanding of factors influencing the rate of RT incidents.