Table 2. Results of the multiple regression analysis.

Potential confounder		Mean deviation of VEGF-A results (estimate) [pg/ml] (±SD)	p- values
	(Intercept)	68.1 (±7.1)	<0.001
Analyzing center	Center A vs. central laboratory	-5.5 (±6.5)	0.396
	Center B vs. central laboratory	-21.3 (±6.6)	0.001
	Center C vs. central laboratory	-26.1 (±8.5)	0.002
	Center D vs. central laboratory	-32.2 (±8.5)	<0.001
	Center E vs. central laboratory	35.2 (±8.0)	<0.001
	Center F vs. central laboratory	32.9 (±8.0)	<0.001
Anticoagulant	PECT vs. EDTA	-33.8 (±4.8)	<0.001
	CTAD vs. EDTA	-73.4 (±7.7)	<0.001
Cannula	neonatal vs. butterfly	5.0 (±5.0)	0.322
Type of centrifuge	swing-out rotor vs. fixed-angle rotor	-23.2 (±5.7)	<0.001
Time before centrifugation	Time to centrifugation: 2h vs. 30min	-8.3 (±6.3)	0.188
Time after centrifugation	Time after centrifugation: 3-4h vs. immediately	-0.04 (±6.9)	0.995
	Time after centrifugation: 6h vs. immediately	7.7 (±7.3)	0.293
Intrapersonal variation over time	week 2 vs. week 1	-2.9 (±3.6)	0.423
Sex	Female vs. male participant	23.6 (±3.7)	<0.001
Filling level	half filled vs. completely filled	7.1 (±4.0)	0.0741
Measuring method	Multiplex bead array vs. ELISA	-38.1 (±5.1)	<0.001

Results of the multiple regression analysis; the first two columns show the investigated parameters, whereby the bold printed one was used as reference for this parameter. The column “Mean deviation of VEGF-A results (estimate)” shows the strength of the influence of this parameter in the regression model (± standard deviation)