Table 4.

Associations of increased myocardial marker levels above cut-offs with in-hospital mortality of COVID-19.

		Crude			Model 1			Model 2
		HR (95% CI)	P-value		HR (95% CI)	P-value		HR (95% CI)	P-value
Early-stage levels	Hs-TnI
	≤ cut-off	Ref			Ref			Ref
	>cut-off	7.85 (4.46-13.81)	<0.001		5.34 (2.89-9.88)	<0.001		2.31 (1.20-4.48)	0.013
	CK-MB
	≤ cut-off	Ref			Ref			Ref
	>cut-off	5.98 (3.60-9.92)	<0.001		3.73 (2.18-6.37)	<0.001		1.82 (1.01-3.31)	0.048
	MYO
	≤ cut-off	Ref			Ref			Ref
	>cut-off	23.88 (12.18-46.82)	<0.001		19.36 (9.65-38.84)	<0.001		8.31 (3.94-17.52)	<0.001
Late-stage levels	Hs-TnI
	≤ cut-off	Ref			Ref			Ref
	>cut-off	67.15 (32.05-140.68)	<0.001		63.54 (28.93-139.57)	<0.001		18.19 (7.87-42.05)	<0.001
	CK-MB
	≤ cut-off	Ref			Ref			Ref
	>cut-off	40.99 (21.91-76.69)	<0.001		31.46 (16.53-59.87)	<0.001		8.22 (3.81-17.71)	<0.001
	MYO
	≤ cut-off	Ref			Ref			Ref
	>cut-off	116.72 (53.14-256.36)	<0.001		108.64 (47.23-249.88)	<0.001		34.41 (14.15-83.68)	<0.001

The variables were categorized into two groups according to the cut-off of each biomarker. The biomarkers were included as dichotomous variables in the univariable COX regression analysis.

Model 1: Adjusted for age, sex, and co-existing diseases (hypertension, diabetes, coronary heart disease, chronic obstructive pulmonary disease, chronic liver disease, stroke history, chronic kidney disease, and cancer history).

Model 2: Model 1 plus variables related to disease severity (neutrophil count, D-Dimer, and high sensitivity C-reactive protein).

HR, hazard ratio; CI, confidence interval; Hs-TnI, high sensitivity troponin-I; CK-MB, creatine kinase-MB; MYO, myoglobin.