Identifying and resolving dose errors in radiation oncology Radiation therapy is the standard of care for the treatment of many cancers. To avoid unnecessary recurrences or toxicity, the correct dose must be delivered to the tumor within ±5% of what is prescribed. The radiation dose to the patient is determined with a treatment planning computer calculation, the accuracy of which depends strongly on the input parameters used by the individual clinic to describe their radiation fields. Audits of centers participating in national radiation therapy clinical trials have found that that 18% of audited cases in the United States do not deliver the radiation dose within ±7% of the intended dose. One possible cause for these errors is the inaccurate characterization of the radiation field in the institution’s computational model. This project will develop tools that can identify when this important cause of dose errors is relevant, identify which parameters are most important for the accurate calculation of radiation dose, and then directly interact with those hospitals with identified problems to resolve those errors. This advances our long term goal of improving survival and decreasing normal tissue toxicity in radiation oncology by ensuring accurate dose delivery. Our hypothesis is that a linear accelerator model-specific computational system can identify and resolve 50% of treatment errors in radiation oncology (those showing >5% error), which would be a substantial step towards improved quality. Specific Aim 1 will develop the infrastructure to identify computational errors in radiation therapy dose calculations using reference radiation beam characteristics. It will then test the performance of hundreds of institutions in calculating radiotherapy doses. Specific Aim 2 will determine which basic radiotherapy parameters are most important for accurate dose calculation under clinical conditions. Specific Aim 3 will develop infrastructure to interact with institutions where computational errors have been identified. We will work with the institution to improve their dose calculation models, and verify that their new models are more accurate at calculating dose. Because the calculation model is used to determine the dose to all patients, this improvement will benefit all patients treated at that institution.