Presenter(s)
P. A. Papanikolaou1, I. C. Liu1, M. B. Roumeliotis2, W. T. Hrinivich1, R. Ger1, B. Floreza1, D. Miles1, H. Li1, C. Deville Jr1, and A. N. Halthore1; 1Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, 2Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins Medicine, Baltimore, MD
Purpose/Objective(s): Proton radiotherapy (PT) planning is complex and sensitive to setup and planning deviations. We evaluated the frequency, types, and clinical severity of errors identified during physics plan review using a standardized checklist.
Materials/Methods: Under an IRB-approved registry, we reviewed charts of patients treated with PT between April 2025 and August 2025 at a single proton center using the RayStation treatment planning system, MOSAIQ oncology information system (OIS), and a Hitachi PT delivery system. After dosimetry sign-off, qualified medical physicists reviewed plans using a standardized 49-item checklist; any identified errors were corrected prior to treatment. Errors were classified as “design/optimization errors” or “scheduling/documentation errors”. A panel consisting of a medical physicist and a radiation oncologist assigned severity scores to each error on a 1-10 scale adapted from the severity scale used in the AAPM TG-100 report (1 = no effect; 10 = catastrophic).
Results: Eighty-nine of 104 PT plans had completed checklists available for review. We identified 28 errors in 20 plans (22.5% of plans with at least 1 error). There were 12 unique errors identified from the 49-item checklist. Nineteen errors (67.9%) were scheduling/documentation errors and 9 (32.1%) were design/optimization errors. The most common error detected involved field visibility settings in the OIS (n = 6; 21.4% of errors); the second most common error pertained to inappropriate density overrides (n = 5; 17.9% of errors). The median error severity score assigned was 3 (range 2-7). The most severe error detected was an incorrect isocenter (n = 1; 3.6% of errors); 4 of the 5 most severe errors were design/optimization errors (80%).
Conclusion: A standardized checklist detected errors in approximately one quarter of reviewed proton plans. Nearly one third of these errors were design/optimization errors. Given the complexity and sensitivity of PT planning, these findings support continued checklist-based QA, improved TPS-OIS interoperability, and implementation of upstream interventions to reduce errors and enhance patient safety.