Figure 1 Number of transcatheter mitral valve repair

Figure 1: Number of transcatheter mitral valve repair procedures performed (red/navy) and 6 -month complication rates (orange/light blue) per 6 -month period for patients of white and non-white race/ethnicity, respectively. Cochran-Armitage trend test demonstrated no statistically significant difference in complication rates over the time between white and non-white race/ethnicity (P=0. 68). 450 Number of procedures 60 421 415 Ptrend = 0. 68 [VALUE]% 400 TMVR procedures for non-white patients Complication rate for non-white patients 50 [VALUE]% 350 314 [VALUE]%289 300 [VALUE]% 250 200 [VALUE]% 221 100 50 [VALUE]% 183 150 [VALUE]% 102 64 62 13 40 28 74 [VALUE]% 30 [VALUE]% 20% [VALUE] 110 74 73 38 0 10 0 2014 2015 2016 2017 (Jan - Jun) (July - Dec) (Jan - June) (July - Dec) (Jan - Jun) (July - Dec) In-hospital major adverse event rate (%) TMVR procedures for white patients Complication rate for white patients

Figure 2: Burden of comorbidities expressed as mean Charlson Comorbidity Index (CCI, red) and Elixhauser Comorbidity Score (ECS, navy) in relationship with total complication rates (light blue) Mean CCI scores Total complication rate 16 14 Mean ECS scores 14, 55 13, 95 14, 37 13, 2 [VALUE]% 13, 11 12, 89 12, 71 12, 99 12 10 [VALUE]% 8 [VALUE]% 6 4 [VALUE]% 2, 62 2, 83 2, 67 2, 81 2, 82 [VALUE]% 2, 82 3, 18 2 0 2014 (Jan - Jun) 2014 (July - Dec) 2015 (Jan - Jun) 2015 (July - Dec) 2016 (Jan - June) 2016 (July - Dec) 2017 (Jan - Jun) 2017 (July - Dec)

Figure 3: Forest Plot showing multilevel multivariable regression analysis for in-hospital major adverse events adjusted by age, sex, relevant comorbidities, and racial/ethnic and neighbourhood income quartiles. A) Stratified by racial/ethnic groups and adjusted by ZIP code median household income quartiles. B) Stratified by ZIP code median household income quartiles and adjusted by racial/ethnic groups. OR: odds ratio; CI: confidence interval; TIA: transient ischemic attack.

Figure 4: Forest Plots showing a sensitivity analysis to determine the effect that adjustment by racial/ethnic and neighbourhood income quartiles had on their corresponding model for in-hospital major adverse events. Multilevel multivariable regression analysis adjusted by age, sex, and relevant comorbidities. A) Stratified by racial/ethnic groups. B) Stratified by ZIP code median household income quartiles. OR: odds ratio; CI: confidence interval; TIA: transient ischemic attack.

Figure 5: Forest plot showing multilevel multivariable regression analyses for comorbidities associated with in-hospital major adverse events. A) Adjusted by racial/ethnic groups. B) Adjusted by median household income quartiles. OR: odds ratio; CI: confidence interval