Study Design

Unmatched and matched cohort designs for race were conducted. For the unmatched analysis, all eligible patients were included. For the matched analysis, the coarsened exact matching procedure with variable ratio matching was performed.^11,12 The black and white study arms were matched by region (Canada, Northern Europe, United States), year of TKI initiation (2003 to 2007, 2008 to 2012, 2013 to 2016), age (< 50, 50 to 59, 60 to 69, ≥ 70 years), and sex to eliminate any imbalances in these factors between the 2 racial categories. Weights were assigned to the matched white study arm, accounting for variable ratios (number of white v black patients) across strata from the matching procedure.^11,12