Presentation (CTI)

OBJECTIVES: Care for breast cancer increasingly spans academic and community healthcare organizations, yet little is known about migration between settings from diagnosis to treatment. We quantified migration between academic and community care settings and evaluated the robustness of these findings to alternative healthcare organization definitions and provider attribution strategies.
METHODS: Incident breast cancer patients were identified from NorstellaLinQ linked US EHR and open claims (≥2 C50.x diagnoses on distinct dates; index 2022-2024; 12-month treatment-free lookback; ≥1 systemic therapy post-index). Diagnosing and treating HCOs were classified as academic or community using a multi-signal hierarchical classifier incorporating NCI/NCCN designation, academic health system affiliation, teaching status, and oncology network membership. Outcomes were same-setting, community-to-academic, and academic-to-community migration, plus same-HCO treatment. Sensitivity analyses applied a stricter academic definition and an alternative treating-provider attribution based on each patient’s dominant longitudinal provider.
RESULTS: Among 264,743 patients, approximately three-quarters received diagnosis and first treatment within the same care setting, yet only one-third remained within the same healthcare organization. Migration between settings was asymmetric: 85.6% of patients diagnosed in academic settings remained in academic care, whereas 57.3% of patients diagnosed in community settings transitioned to academic care for first treatment. Using each patient's dominant longitudinal provider rather than the first treating provider produced nearly identical estimates of same-setting treatment (74.1%), despite differing provider attribution for 20.8% of patients.
CONCLUSIONS: Breast cancer care pathways are characterized by substantial migration between healthcare organizations despite relative stability in academic versus community care settings. Community-diagnosed patients frequently transition to academic care for treatment, and these directional migration patterns were robust across provider-attribution strategies. Studies evaluating site of care should report healthcare organization classification methods transparently and account for organizational fragmentation that is obscured by setting-level analyses. Future work should incorporate place-of-service information and adjust for disease stage and clinical complexity.