Table 4

 Analysis of the hospital and country level effects on 30 day death/(re)MI and one year mortality

VariableHospital level effectsCountry level effectsIntrahospital correlation (%)Intracountry correlation (%)
Variance (SE, p)% ExplainedVariance (SE, p)% Explained
*Adjusted for age, prior myocardial infarction, prior transient ischaemic attacks, prior stroke, prior coronary artery bypass graft surgery, prior use of calcium channel blockers and β blockers, ST segment depression, troponins T, creatinine clearance, and time to randomisation. †Adjusted for country level coronary artery bypass graft surgery (CABG) rate.
30 day death/(re)MI
Two level null model0.03611.09
(0.024;
p = 0.004)
Three level models
Model 1: null0.08620.027922.72.530.82
(0.046;(0.023;
p = 0.003)p = 0.072)
Model 2: age only0.071816.70.012256.32.130.36
(0.044;(0.017;
p = 0.010)p = 0.456)
Model 3: all baseline factors*0.041951.40.001196.11.260.03
(0.041;(0.012;
p = 0.014)p>0.500)
Model 4: Baseline+country level CABG-rate†0.046246.40.000199.61.380.00
0.039;(0.001;
(p = 0.032)p>0.500)
One year death
Two level null model0.03691.11
(0.024;
p = 0.005)
Three level models
Model 1: null0.1790.017246.65.130.49
(0.058;(0.022;
p = 0.001)p = 0.312)
Model 2: age only0.1704.90.006860.54.910.20
(0.059;(0.018;
p = 0.002)p>0.500)
Model 3: all baseline factors*0.12430.60.0000299.73.640.00
(0.054;(0.015;
p = 0.014)p>0.500)
Model 4: Baseline+country level CABG-rate†0.10342.50.0002399.43.030.01
(0.050;(0.001;
p = 0.008)p>0.500)