Abstract Background: There have been renewed calls for greater inclusivity of populations most prominently affected by cancer health inequities across the cancer research enterprise, particularly Black and African American populations. In cancer health services research there are myriad methodologies and data sources to study cancer treatment outcomes including medical records, cancer registries, and primary data collection. These different methodologies have varying levels of resource intensity that may affect the recruitment of marginalized populations. Little is known about how much concordance between data sources may exist among marginalized populations with cancer, which has substantial implications for cancer inequities research. Objective: To evaluate the data concordance between medical records, cancer registries, and self-reported survey data among an underrepresented population in cancer population sciences research. Methods: This investigation used a population-based sample of self-identified Black women recruited through the Tennessee and Florida Cancer registries who were diagnosed with invasive breast cancer at or below the age of 50. Participants completed medical records release forms and questionnaires about their cancer treatment trajectories. Comparing data contained in the cancer registries, medical records, and self-reported data, we quantitatively assessed the concordance between data sources for receipt of surgery, chemotherapy, and radiation, using medical records as the gold standard. Results: In total, 558 Black women at or below 50 with invasive breast cancer had available registry data and also completed a medical records release. In terms of missing data for receipt of surgery, chemotherapy, and radiation, self-report was the most complete, followed by the registry and then the medical records. Among those with complete data, comparing the medical records to the cancer registry, accuracy was high for surgery (94.6% ; 95% CI 92.2, 96.4) and chemotherapy (89.7%; 95% CI: 85.7, 92.8) , but low for radiation (55.4%, 95% CI: 48.5, 62.2) with the cancer registry missing the receipt of radiation that was noted in the medical records. Comparing medical records to self-report data, accuracy was >90% for surgery, chemotherapy, and radiation. Comparing the cancer registry to self-report, accuracy was high for surgery (94.5%; 95% CI: 92.0, 96.4), moderate for chemotherapy (88.2; 95% CI: 85.1,90.9), and low for radiation (67.2%; 95% CI 62.9, 71.3). Conclusion: For receipt of surgery and chemotherapy, there was high concordance between medical records, cancer registries, and self-reported data. The receipt of radiation was inconsistently reported in the cancer registry and medical records. When considering inclusivity in the study of cancer services delivery among marginalized populations, tradeoffs of data validity and resource intensity must be weighed. Citation Format: Mya L. Roberson, Lindsay Venton, Anne Weidner, Sonya Reid, Tuya Pal. Evaluating treatment data concordance among young Black women diagnosed with breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1939.
Read full abstract