Basic structure of the Markov model. In any given year, patients can remain alive, die from breast cancer, or die from a non–breast cancer cause. In the model, receipt of treatment—endocrine therapy or chemotherapy—affects race-specific probabilities of breast cancer and non–breast cancer death. In the right panel, the “Alive AND treated” circle is larger, and the “Alive NOT treated” circle is smaller, representing the intervention effect (ie, the increase in the proportion of patients who start the model in the treated vs untreated state). We used trial-derived, race-stratified effect estimates to reflect the differential impact of inequity-reducing interventions on adjuvant therapy uptake across racial subgroups.