Table 3: Predicting risk of 1-year angina model (The reduced model).

This table contains the estimated logistic regression coefficients for the significant variables that are entered into the 1-year angina model.

Effect	Regression Coefficient	Standard Error	P value
Intercept	−0.1099	0.8079	0.8918
Revascularization strategy (PCI)	−0.07646	0.2149	0.7220
Age	−0.01316	0.007276	0.0707
Male	−0.2904	0.1458	0.0466
Presentation as acute coronary syndrome	−0.2818	0.1442	0.0509
Hemoglobin (+1 g/dL)	−0.06585	0.04219	0.1188
Daily/Weekly angina at baseline*	1.1275	0.1972	<.0001
Monthly angina at baseline*	0.6244	0.2009	0.0019
SYNTAX SCORE (>22)	−0.3790	0.1963	0.0536
Revascularization with PCI*SYNTAX (>22)	0.5900	0.2700	0.0290

^*Reference group for angina is (no angina)

The 1-year individualized angina predicted risk can be calculated as follows (substitute 1 or 0 for presence or absence of any categorical variable: presenting as ACS, history of MI, history of peripheral vascular disease, male and angina at baseline. For the variable “treatment” substitute 1 for PCI and 0 for CABG. For all continuous variables (BMI, eGFR, LVEF, hemoglobin and age) plug the actual value in the equation)= 1/(1+exp(− (−1.1504–0.02040* treatment*BMI + 0.01612*BMI - 0.00727* treatment*eGFR + 0.002129*eGFR + 0.2870* treatment *presenting as ACS - 0.2912* treatment *history of MI + 0.05326* history of MI - 0.5213* treatment *history of peripheral vascular disease + 0.08552* history of peripheral vascular disease + 0.006867*LVEF - 0.07286*hemoglobin - 0.01575*age - 0.4273* presenting as ACS + 1.4594*treatment - 0.2648*Male + 0.8988*angina at baseline)))