Background: Studies using healthcare claims data to investigate the burden of skeletal-related events employ various approaches to identify radiation to the bone (RttB) because billing codes available in claims data do not distinguish RttB from radiation to the prostate gland. We investigated the use of Eventflow data visualization software to identify components of a claims-based algorithm for RttB. Methods: We analyzed data for PCa cases identified in the Surveillance, Epidemiology, and End Results (SEER) registry data linked with Medicare claims. We identified two cohorts of individuals diagnosed between 2005 and 2009 and receiving radiation therapy, C1: diagnosed with incident bone metastasis (BM) according to SEER data; C2: diagnosed with incident stage IV M0 PCa. We defined radiation episodes of care from claims for external beam radiation therapy, radiopharmaceutical therapy, intensity modulated radiotherapy and stereotactic radiosurgery. Eventflow was used to compare cohorts and identify criteria for identifying RttB using claims data. Results: Application of inclusion criteria resulted in 1,491 individuals: 999 in C1 and 492 in C2. Median follow up was 596 days in C1 and 882 days in C2. Average age was 77 years in C1 and 73 years in C2. The median time to radiation therapy was 133 days in C1 compared to 171 days in C2. When requiring a BM diagnosis code on the radiation claim, the median time was 160 days in C1 compared to 514 days in C2. The median time to a fracture was 107 days in C1 compared to 369 days in C2. The median time to bone surgery was 183 days in C1 compared to 447 days in C2. The median time to spinal cord compression was 154 days in C1 compared to 375 days in C2. A BM diagnosis code concurrent with the radiation episode and radiation episodes less than or equal to 4 weeks in length were more common among C1 compared to C2. Conclusions: Analysis of data visualization output indicates that incorporating a bone metastasis code on claims concurrent with a radiation episode of care or the information regarding the length of the radiation episode will be useful for identifying receipt of palliative radiation using claims data.