Table 2.

Univariate and multivariate linear regression analysis for clinical variables associated with initial urinary TIMP-1 level

	Univariate regressiona			Multivariate regressionb
	B	SE	P value	B	SE	P value value
Age, months	0.100	0.073	0.17	N/A
Body weight, kg	0.139	0.161	0.39	N/A
Sex	− 0.116	0.100	0.25	N/A
PRISM III score	0.052	0.005	< 0.001	0.031	0.006	< 0.001
AKI stage	0.489	0.041	< 0.001	0.377	0.049	< 0.001
MV	0.494	0.103	< 0.001	0.180	0.104	0.09
Sepsis	0.487	0.116	< 0.001	0.277	0.102	0.007
MODS	1.099	0.159	< 0.001	0.015	0.183	0.93
Shock/DIC	0.769	0.167	< 0.001	− 0.325	0.175	0.06
Antibiotic	0.116	0.121	0.34	N/A
Inotrope	0.583	0.151	< 0.001	− 0.251	0.151	0.10
Furosemide	0.411	0.106	< 0.001	− 0.034	0.102	0.74
Steroid	− 0.239	0.095	0.01	− 0.110	0.081	0.17
Hemofiltration	1.021	0.182	< 0.001	0.306	0.171	0.07

AKI acute kidney injury, DIC disseminated intravascular coagulation, MODS multi-organ dysfunction syndrome, MV mechanical ventilation, N/A not applicable, PRISM III pediatric risk of mortality III. Continuous variables were log-transformed in the linear regression analyses

^aAll variables in Table 1 were analyzed in the univariate linear analysis

^bVariables with P < 0.05 were entered into the multivariate analysis after checking the multicollinearity by variance inflation factor and tolerance values