Table 4. Risk analysis for time to thromboembolic event.

		Covariate factor effect		Treatment effect adjusted by covariate
Covariate	n	Hazard ratio (95% CI)	P-valuea	Hazard ratio (95% CI)	P-valuea
Treatment effect (unadjusted)	2297	—	—	1.62 (1.13; 2.31)	0.0081
Gender	2297	1.09 (0.76; 1.57)	0.637	1.62 (1.14; 2.32)	0.0078
Previous thromboembolic event	2297	2.23 (1.13; 4.38)	0.021b	1.64 (1.15; 2.35)	0.0065
Coronary artery disease	2297	1.99 (0.88; 4.52)	0.098b	1.62 (1.14; 2.32)	0.0079
Hypertension	2297	1.57 (1.08; 2.27)	0.018b	1.62 (1.13; 2.31)	0.0082
Diabetes	2297	1.16 (0.62; 2.14)	0.643	1.62 (1.13; 2.32)	0.0080
Dyslipidemia	2297	1.96 (0.86; 4.44)	0.108b	1.64 (1.14; 2.34)	0.0069
Previous MI, angina, unstable angina	2297	1.16 (0.54; 2.48)	0.704	1.62 (1.13; 2.31)	0.0081
Obesity	2297	0.76 (0.46; 1.25)	0.286	1.62 (1.14; 2.32)	0.0077
Age >65 years at baseline	2297	1.35 (0.96; 1.91)	0.085b	1.62 (1.13; 2.31)	0.0084

	Control			Epoetin-β
Subgroup	n	Number of events	KM 90 day event rate	n	Number of events	KM 90 day event rate	Hazard ratio (95% CI)
All patients	1053	46	0.96	1244	88	0.94	1.62 (1.13; 2.31)
No risk factors	555	20	0.96	646	36	0.96	1.54 (0.89; 2.66)
One risk factor	348	22	0.93	431	31	0.93	1.09 (0.63; 1.88)
Two risk factors	125	2	0.99	136	13	0.90	6.35 (1.43; 28.13)
Three or more risk factors	25	2	0.92	31	8	0.77	3.33 (0.71; 15.69)

^aHR P-value (Wald test).

^bCovariates significant at the 15% level in univariate analysis.