Equity analysis of clinical oncology treatment pathways usage in an integrated health care delivery system.

e13760 Background: Oncology treatment has been rapidly evolving and becoming increasingly complex. At Kaiser Permanente (KP), we created a decision support tool to provide standardized, evidence-based, and cost-effective treatment options for our clinicians. Here we examined the equitable use of our oncology pathways tool across patient socio-demographic groups within a large diverse national cohort unique to KP. Methods: Analysis included the earliest eligible drug treatment plan for 41,505 adult cancer patients between 2022 and 2024. Patients were excluded if they were missing data or part of a group with low numbers (3.2%). The model outcome was whether the treatment plan was ordered via the pathway tool. A logistic regression model analyzed the relationship between pathway tool use and patient race/ethnicity groups as well as insurance type, after adjusting for age, sex, cancer type, and Charlson comorbidities to account for differences that may affect the necessity for care customization. Model-based direct standardization was used to calculate the adjusted probability of tool use and average marginal effects (AME) of the differences in pathway tool use across race/ethnicity groups and insurance types (Table 1). Results: We found high usage (> 80%) of the oncology pathway tool across all patient race/ethnicity groups and insurance types (Commercial, Medicare, and Medicaid). Although the adjusted usage rates for Asian/Pacific Islander and Hispanic were statistically significantly higher than White [AME 2.2 95% CI (1.1, 3.3) and AME 3.0 95% CI (2.0, 3.9), respectively)] and adjusted rates for Medicare were statistically significantly lower than Commercial [AME -1.5 95%CI (-2.9, -0.3)], the magnitude of the differences was relatively small. Absolute differences in adjusted usage rates did not exceed 3% across all populations, thus statistical significance was likely due to large sample size and not meaningful differences in care delivery. Conclusions: This analysis suggests the application of oncology pathways has been equitable across patient race/ethnicity and insurance groups at KP. Although statistically significant differences exist in the use of oncology pathways, the absolute effect size is not robust enough to imply clinically impactful differences and is likely an artifact of large sample size. The broader implication of this analysis is that the use of oncology pathways may facilitate equitable and standardized care. Patient Group N (%) Adjusted Probability of Using Pathway Tool (%) AME vs Reference (%) 95% CI for AME (%) Asian/Pacific Islander 6,401 (15.4) 82.7 2.2 (1.1, 3.3) Black 5,195 (12.5) 81.1 0.6 (-0.6, 1.7) Hispanic 7,884 (19.0) 83.5 3.0 (2.0, 4.0) Multiracial 671 (1.6) 82.0 1.5 (-1.6, 4.3) White (Reference) 21,354 (51.4) 80.5 Medicare 21,022 (50.6) 80.8 -1.4 (-2.7, -0.1) Medicaid 1,541 (3.7) 82.1 -0.1 (-2.2, 1.8) Commercial (Reference) 18,942 (45.6) 82.2