Factors impacting genomic testing rates among epithelial ovarian cancer patients across a large community-based healthcare system (595)

Objectives: To determine the rates of germline and somatic testing for epithelial ovarian cancer (EOC) patients and identify factors that impact testing rates across a large community-based healthcare system operating in seven states across WA, OR, CA, AK, MT, NM, and TX. The system is comprised of over 100,000 caregivers, 51 hospitals, and 829 physician clinics. The aim was to identify fixed barriers to testing, including region, hospital type, insurance status, racial/ ethnic disparities, and stage of diagnosis. Methods: Clinical, pathologic, demographic, and genomic testing information was obtained from the diverse dataset within the Providence St. Joseph Health Electronic Medical Records and the system-wide cancer registry for all patients with an EOC diagnosis (ICD C56.x) between January 2015 and January 2020. Structured genomic data was sourced from laboratory information systems and manual abstraction of molecular sequencing reports. This dataset encompasses patient population data among diverse hospital settings and urban/rural environments. Hospital types were broken down into academic settings, which contain residency programs (Academic), Commission on Cancer (CoC) accredited programs or smaller community sites (Community) without CoC accreditation. Descriptive statistics and logical regression were utilized to summarize key findings. Results: Within this EOC cohort (3,007 patients), 34% (n=1,027 patients) completed some type of genomic testing (GT). The percentage of patients tested increased from 31% in 2015 to 46% in 2019, reflecting the uptake of testing guidelines. The increase in GT rates was largely attributable to an increase in somatic tumor testing (14-39%), while germline testing rates were stable across the interval (25-33%). Patients were more likely to receive testing if they received care at an academic or CoC institution versus a community institution (p=0.0001). Logistic regression analysis demonstrated the following factors impacted tested rates: institution type, insurance, and stage at diagnosis (p=0.001, p= 0.0019 and p \u3c 0.0001, respectively). Race/ethnicity did not contribute significantly to the model but did have a significant effect when analyzed independently. Conclusions: This study is the first to analyze practice patterns in GT for EOC across a broad community-based healthcare system servicing seven states. The data highlight discrepancies in GT heavily influenced by practice setting, insurance status, and stage of diagnosis (likely reflecting payer coverage/ increased need for information in advanced-stage disease). There is a need for a universally defined testing approach to provide equitable access to evidence-based cancer care