Table 2

Univariate and multivariate Cox regression analyses for all-cause mortality and composite events in the derivation cohort (SAKURA AF registry)

VariableUnivariate analysisMultivariate analysis (model 1)Multivariate analysis (model 2)
HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P value
All-cause mortality
 Cluster 1 (ref)
 Cluster 28.4 (2.0 to 35.0)0.0035.4 (1.2 to 22.8)0.025.7 (1.3 to 24.0)0.01
 Cluster 317.4 (4.2 to 72.0)<0.00113.1 (3.1 to 54.8)0.0211.7 (2.8 to 49.4)<0.001
 Cluster 422.0 (5.1 to 90.2)<0.0019.2 (2.0 to 40.9)0.00321.7 (4.9 to 95.0)<0.001
 Cluster 518.2 (4.4 to 75.7)<0.0016.5 (1.4 to 30.3)0.0110.1 (2.4 to 43.2)0.001
Composite events
 Cluster 1 (ref)
 Cluster 24.2 (2.2 to 7.8)<0.0013.2 (1.7 to 6.2)<0.0013.4 (1.8 to 6.5)<0.001
 Cluster 35.2 (2.7 to 9.9)<0.0014.4 (2.3 to 8.5)<0.0014.3 (2.2 to 8.3)<0.001
 Cluster 46.3 (3.4 to 11.8)<0.0013.9 (1.9 to 7.9)<0.0017.3 (3.6 to 14.7)<0.001
 Cluster 57.9 (4.2 to 14.7)<0.0014.4 (2.0 to 9.4)<0.0015.9 (3.1 to 1.4)<0.001
  • Model 1: adjusted for CHA2DS2-VASCc score; model 2: adjusted for age ≥75 years, sex and type of atrial fibrillation.

  • CHA2DS2-VASCc, congestive heart failure, hypertension, age ≥75 years, diabetes, stroke, vascular disease, age 65–74 years and male; ref, reference; SAKURA AF, Real World Survey of Atrial Fibrillation Patients Treated with Warfarin and Non-vitamin K Antagonist Oral Anticoagulants.