Table 4.

Results of multivariate analyses with the optimal cut-offs from table 3

	Coeff		p value
Transfusion requirement, ml RBCs/kg: multiple linear regression, cut-off ≤ 2 days
Body weight, kg	9.06		<0.001
Previous sternotomy	0.21		0.86
RACHS score	0.61		0.68
RBCs older than 3 days	9.06	<0.001
	OR	CI
FFP requirement: multiple logistic regression, cut-off ≤ 4 days
Body weight, kg	1.07	0.89–1.30	0.47
Previous sternotomy	1.15	0.35–3.80	0.81
RACHS score	1.73	0.97–3.09	0.06
RBCs, ml/kg	1.04	0.99–1.08	0.06
RBCs older than 4 days	5.13	1.92–13.72	0.001
	HR	CI
Length of ventilation, days: Cox regression proportional hazards for delayed extubation, cut-off ≤ 6 days
Body weight, kg	0.80	0.72–0.88	<0.001
Previous sternotomy	1.19	0.71–2.00	0.50
RACHS score	1.64	1.25–2.17	<0.001
RBCs, ml/kg	1.03	1.01–1.05	<0.001
RBCs older than 6 days	1.72	1.15–2.63	0.009
Length of ICU, days: Cox regression proportional hazards for delayed release from ICU, cut-off ≤ 8 days
	HR	CI
Body weight, kg	0.93	0.85–1.02	0.12
Previous sternotomy	0.88	0.53–1.47	0.64
RACHS score	1.59	1.23–2.08	<0.001
RBCs, ml/kg	1.02	1.00–1.03	0.06
RBCs older than 8 days	1.43	0.99–2.08	0.054
	Coeff
Post-OP-CRP peak, mg/dl: backward stepwise regression, cut-off ≤6 days
Body weight, kg	0.007		0.97
Previous sternotomy	4.9		<0.001
RACHS score	−0.22		0.68
RBCs, ml/kg	−0.04		0.27
RBCs older than 6 days	2.27		0.009

For each outcome parameter, the results for the optimal cut-off with best separation between fresh and old RBCs as given in table 3 are listed. In addition to the significance levels P, regression coefficients for linear regression, ORs with confidence intervals (CI) for use of FFP calculated by logistic regression and HRs with CI for delayed extubation or delayed release from the ICU calculated by Cox regression are listed.