Table 2 Open in new tab Risk estimates...

Model

95% CI

P-value

Mortality

Unadjusted logistic regression

1.18

1.13–1.23

<.001

Multivariable logistic regression^a

1.59

1.49–1.71

<.001

*IV logistic regression^b

1.67

1.54–1.81

<.001

Myocardial infarction

Unadjusted logistic regression

1.45

1.37–1.54

<.001

Multivariable logistic regression^a

1.38

1.29–1.48

<.001

IV logistic regression^b

1.51

1.41–1.62

<.001

Stroke

Unadjusted logistic regression

.94

.87–1.12

.140

Multivariable logistic regression^a

.96

.87–1.06

.401

IV logistic regression^b

.93

.85–1.03

.215

Revascularization

Unadjusted logistic regression

3.23

3.04–3.44

<.001

Multivariable logistic regression^a

3.35

3.08–3.65

<.001

IV logistic regression^b

3.01

2.57–3.51

<.001

Heart failure

Unadjusted logistic regression

1.10

1.05–1.16

<.001

Multivariable logistic regression^a

1.19

1.12–1.28

<.001

IV logistic regression^b

1.15

1.07–1.25

<.001

Table 2

Open in new tab

Risk estimates in different statistical models

Model	OR	95% CI	P-value
Mortality
Unadjusted logistic regression	1.18	1.13–1.23	<.001
Multivariable logistic regression^a	1.59	1.49–1.71	<.001
*IV logistic regression^b	1.67	1.54–1.81	<.001
Myocardial infarction
Unadjusted logistic regression	1.45	1.37–1.54	<.001
Multivariable logistic regression^a	1.38	1.29–1.48	<.001
IV logistic regression^b	1.51	1.41–1.62	<.001
Stroke
Unadjusted logistic regression	.94	.87–1.12	.140
Multivariable logistic regression^a	.96	.87–1.06	.401
IV logistic regression^b	.93	.85–1.03	.215
Revascularization
Unadjusted logistic regression	3.23	3.04–3.44	<.001
Multivariable logistic regression^a	3.35	3.08–3.65	<.001
IV logistic regression^b	3.01	2.57–3.51	<.001
Heart failure
Unadjusted logistic regression	1.10	1.05–1.16	<.001
Multivariable logistic regression^a	1.19	1.12–1.28	<.001
IV logistic regression^b	1.15	1.07–1.25	<.001

Model	OR	95% CI	P-value
Mortality
Unadjusted logistic regression	1.18	1.13–1.23	<.001
Multivariable logistic regression^a	1.59	1.49–1.71	<.001
*IV logistic regression^b	1.67	1.54–1.81	<.001
Myocardial infarction
Unadjusted logistic regression	1.45	1.37–1.54	<.001
Multivariable logistic regression^a	1.38	1.29–1.48	<.001
IV logistic regression^b	1.51	1.41–1.62	<.001
Stroke
Unadjusted logistic regression	.94	.87–1.12	.140
Multivariable logistic regression^a	.96	.87–1.06	.401
IV logistic regression^b	.93	.85–1.03	.215
Revascularization
Unadjusted logistic regression	3.23	3.04–3.44	<.001
Multivariable logistic regression^a	3.35	3.08–3.65	<.001
IV logistic regression^b	3.01	2.57–3.51	<.001
Heart failure
Unadjusted logistic regression	1.10	1.05–1.16	<.001
Multivariable logistic regression^a	1.19	1.12–1.28	<.001
IV logistic regression^b	1.15	1.07–1.25	<.001

The reference group in all models is PCI.

ECG, electrocardiogram, IV, instrumental variable.

^aMultivariable logistic regression with follow-up time as a log-transformed offset variable with the following covariates: age at admission, sex, smoking status, history of chronic obstructive pulmonary disease, heart failure, history of myocardial infarction, diabetes mellitus, peripheral arterial disease, hypertension, coronary artery bypass grafting, percutaneous coronary intervention, body mass index, Killip ≥ 2, ST-T changes, ECG rhythm.

^bInstrumental variable analysis with treatment-preference instrument based on the quintiles of the hospital preference for using percutaneous coronary intervention. The IV model also used all the covariates from the multivariable logistic regression. *Primary analysis.

This Feature Is Available To Subscribers Only