PROJECT SUMMARY/ABSTRACT Healthcare and research organizations, including the American Society for Clinical Oncology, acknowledge the common use of stigmatizing language in cancer care. However, stigmatizing language has not been comprehensively defined, and drivers and outcomes of this language use have not been investigated in cancer care. One way that discrimination and stigma manifest in the healthcare system is through the electronic health record (EHR), which can reflect the author’s implicit and explicit bias. Reading stigmatizing language alters readers’ clinical decisions in vignette studies. Thus, there is an opportunity to evaluate EHR stigmatizing language as a modifiable mechanism of decreasing discrimination and improving outcomes in cancer care. The long-term goal is to evaluate sociocultural and healthcare system influences of EHR stigmatizing language and resulting changes in clinical decision-making in cancer care. The central hypothesis is that stigmatizing language is present in the cancer care EHR, inequitably experienced, and associated with altered clinical decision-making. The hypothesis will be tested via the following specific aims: (1) Determine the categories of stigmatizing language used by oncology clinicians when documenting patient encounters in the EHR; (2) Compare stigmatizing language in the cancer care EHR by patient and clinician characteristics; (3) Determine if stigmatizing language use is associated with disparities in medication prescription for cancer-related pain. This study will leverage EHR data from a large academic and community healthcare system in Minnesota. Aim 1 utilizes a qualitative study design to perform a directed content analysis of the unstructured, free text section of outpatient oncology EHR notes. Aims 2 & 3 use natural language processing to abstract stigmatizing terms and phrases for each individual with outpatient cancer visits in 2022. Multilevel models will be utilized to account for nested visit-level data within patient-level data within clinician-level data. This project is innovative in its characterization and modeling of stigmatizing language in cancer care and will have significant impact by informing interventions to reduce stigma, health disparities, and resulting negative health outcomes. By focusing on EHR stigmatizing language, this project addresses emerging opportunities and challenges as medical records become accessible to patients and persist throughout their lives. The applicant is an MD/PhD and Epidemiology PhD student at the University of Minnesota Schools of Medicine and Public Health. By expanding content knowledge in cancer survivorship and disparities, rigorous qualitative and quantitative analysis skills, tailored clinical training, and focused professional development opportunities, this fellowship will enhance her ability to excel in the MD/PhD program and become an independent academic physician- epidemiologist.