APPENDIX 1.

Multivariable logistic regression to model a dichotomous outcome of minimally invasive or open repair: Overall analysis

Variable	Coefficient β	OR*	95% CI	P value
Age	0.007	1.01	(0.99-1.02)	.4
BMI (vs > 30 kg/m–2)
Underweight (<18.5 kg/m–2)	0.446	2.21	(0.54-9.01)	.4
Normal (18.5-25 kg/m–2)	–0.189	1.18	(0.69-1.99)	.4
Overweight (>25 kg/m–2)	0.088	1.55	(0.92-2.61)	.7
Female gender	0.081	1.18	(0.79-1.74)	.4
NYHA functional class I	0.207	1.51	(1.01-2.27)	.05
Degenerative cause	–0.033	0.94	(0.53-1.64)	.8
Chronic lung disease	0.129	1.29	(0.69-2.42)	.4
Severe MR	0.525	2.86	(1.16-7.06)	.02
Greater than mild tricuspid regurgitation	–0.694	0.25	(0.16-0.39)	< .001
Ejection fraction (%)	0.009	1.01	(0.99-1.03)	.3
Previous cardiac intervention	0.021	1.04	(0.54-2.00)	.9
Type II MR	0.353	2.02	(0.93-4.40)	.07
Intercept	–1.989

OR, Odds ratio; CI, confidence interval; BMI, body mass index; NYHA, New York Heart Association; MR, mitral regurgitation.

^*OR represents minimally invasive sternotomy.

c-statistic 0.71; median difference in propensity scores of matched pairs 0.0015 (interquartile range, 0.0005-0.008).