Abstract C003: A comparison of surveillance, epidemiology, end results (seer) and medicare claims-derived radiation and chemotherapy treatment among Black and White patients with uterine cancer

Abstract Introduction: Racial disparities in uterine cancer treatment and survival are well-documented. Despite the growing use of the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database for assessment of treatment-related uterine cancer survival effects, concordance between the two sources of radiation and chemotherapy information (medical record abstraction in SEER and billing claims in Medicare) has not been evaluated among uterine cancer patients. We aimed to assess agreement between these sources and examine potential racial/ethnic differences in concordance. Methods: Non-Hispanic Black (NHB) and non-Hispanic White (NHW) patients diagnosed with uterine cancer between 2000 and 2019 were identified using the SEER-Medicare linked database. In SEER, chemotherapy is categorized as no/unknown, yes, or missing and radiation type can be categorized as none, external beam radiation therapy (EBRT), vaginal brachytherapy (VBT), EBRT plus VBT, or missing. We identified radiation type and chemotherapy receipt using various claim types [e.g., Healthcare Common Procedure Coding System (HCPCS)] across multiple datasets [e.g., National Claims History (NCH)]. Administration of radiation or chemotherapy was established when we identified at least one claim 6 months before or 9 months after the SEER diagnosis date. We then calculated the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and Kappa statistics for EBRT, VBT, and chemotherapy using Medicare claims as the presumed gold standard. We stratified the analysis by race/ethnicity to evaluate potential variation by race/ethnicity. Results: Sample sizes varied by treatment. In the radiation analyses, we excluded 23,923 uterine cancer patients with unknown radiation receipt in SEER, leaving 29,035 patients (8.3% NHB, 91.7% NHW) for analysis. Sensitivity was high for both EBRT and VBT (>88%), with little variation by race; however, specificity for EBRT was low among NHB (20.0%) and NHW (24.4%) women, leading to poor agreement between the two sources (EBRT Kappa: 0.23 for NHB and 0.30 for NHW). PPV and NPV were moderate for EBRT and VBT (>62.0%). In the chemotherapy analysis, we excluded 20,923 uterine cancer patients with missing SEER chemotherapy data, leaving 32,035 uterine cancer patients (8.4% NHB, 91.6% NHW). Sensitivity was high for NHB (95.9%) and NHW (97.2%) patients; however, specificity (≤36%) and agreement (Kappa: ≤0.42) were low. PPV and NPV of chemotherapy were high (>68.0%). Conclusions: Agreement between SEER and Medicare varied by treatment type; however, we did not observe meaningful differences in performance characteristics by race. Future analyses will examine treatment claims data among the sizable subset of patients with missing SEER treatment. In addition, we will compare coarse SEER treatment variables (i.e., no vs yes) with more detailed, claims-defined treatment scenarios (i.e., receipt of guideline-recommended 4-6 cycles of chemotherapy) to assess the accuracy and quality of treatment data captured by SEER. Citation Format: Ashley S. Felix, Jiahao Ping, Caitlin E. Meade, Simran A. Kanal, Tasleem J. Padamsee, Macarius M. Donneyong, Kemi M. Doll, Jennifer A. Sinnott. A comparison of surveillance, epidemiology, end results (seer) and medicare claims-derived radiation and chemotherapy treatment among Black and White patients with uterine cancer [abstract]. In: Proceedings of the 18th AACR Conference on the Science of Cancer Health Disparities; 2025 Sep 18-21; Baltimore, MD. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2025;34(9 Suppl):Abstract nr C003.