Methods of creatine kinase-MB analysis to predict mortality in patients with myocardial infarction treated with reperfusion therapy

Table 5 Univariable associations of infarct size determined by various methods with 90-day composite of mortality, heart failure, shock and stroke

Method	COMMA trial		COMPLY trial
Method	Hazard ratio (95% CI)	Wald χ ²(P value)	Hazard ratio (95% CI)	Wald χ ²(P value)
Total CK-MB AUC 0 to 72 hours from randomization estimated by curve modeling^a
per 1000 units	1.14	53.08	1.08	48.31
per 1000 units	(1.10 to 1.18)	(<0.001)	(1.06 to 1.11)	(<0.001)
per 1 SD	1.74		1.51
per 1 SD	(1.50 to 2.02)		(1.35 to 1.70)
Estimated peak^a
per 100 units	1.03	15.34	1.11	29.87
per 100 units	(1.02 to 1.05)	(<0.001)	(1.07 to 1.15)	(<0.001)
per 1 SD	1.15		1.36
per 1 SD	(1.07 to 1.24)		(1.22 to 1.53)
Observed trapezoidal CK-MB AUC
per 1000 units	1.15	55.76	1.09	50.26
per 1000 units	(1.11 to 1.19)	(<0.001)	(1.06 to 1.11)	(<0.001)
per 1 SD	1.79		1.53
per 1 SD	(1.53 to 2.08)		(1.36 to 1.73)
Observed peak
per 100 units	1.26	45.96	1.11	30.05
per 100 units	(1.18 to 1.34)	(<0.001)	(1.07 to 1.16)	(<0.001)
per 1 SD	1.72		1.36
per 1 SD	(1.47 to 2.01)		(1.22 to 1.52)

^aEstimated from the log-normal model. AUC, area under the curve; CI, confidence interval; CK, creatine kinase; COMMA, COMplement inhibition in Myocardial infarction treated with Angioplasty; COMPLY, COMPlement inhibition in myocardial infarction treated with thromboLYtics; SD, standard deviation.

ISSN: 1745-6215