Background: Pharmacists provide an important role in oncology practice, especially as the use of oral anticancer agents (OAAs) has increased. A program supporting the integration of clinical oncology pharmacists into community practices, Pharmacists Optimizing Oncology Care Excellence in Michigan (POEM), is examining their roles and impact. The purpose of this study was to identify patient characteristics that were associated with medication modification interventions from clinical oncology pharmacists. These results will provide insight into how pharmacists support patient care as well as how to focus pharmacists’ care when their time may be limited. Methods: This multicenter, retrospective analysis was conducted with data from October 2020 – April 2023 documenting clinical care provided to patients taking OAA(s) and receiving a patient-focused intervention. POEM pharmacists received support in the development of collaborative practice agreements, billing, and clinical oncology education. Data included patient demographics, and encounter and intervention characteristics. First, we predicted medication modification interventions including each OAA in machine learning Regression Tree analyses. Second, we used zero-inflated Poisson regression to quantify predictors of receipt of medication modification interventions. We examined cross-tabulations for age and rurality and medication modifications. Results: In these analyses, 1836 met study inclusion, and 1029 medication modifications were made for 623 patients. In Tree analyses, Group 1 was patients receiving trametinib (n=1 patient) and cabozantinib (n=2), and Group 2 was patients receiving alpelisib (n=28 patients) and osimertinib/palbociclib (n=1). Groups 1 and 2 were associated with statistically significantly more medication modifications (predicted value was 6.7 and 3.8, respectively). Age (65-69 and >80) and rurality (rural/small town and micropolitan) were statistically significant predictors of receiving a medication modification as well as the number of medication modifications. For example, individuals residing in rural/small towns compared to metropolitan were less likely to have no medication modifications (-0.668 SE 0.22, p=0.003). They were also more likely to have greater numbers of medication modifications (0.545 SE 0.10, p<0.001). Age (Χ²=10.2, 3df, 0=0.017) and rurality (X²=78.81, 6df, p<0.001) were statistically significantly associated with more medication modifications. Conclusions: Individuals who took alpelisib, were older, and living in more rural communities were more likely to receive medication modifications from clinical oncology pharmacists. Developing sustainable models for interprofessional pharmacist care in community oncology practices appear to be an important strategy to address disparities in care, particularly for patients from rural areas.