Impact of the perioperative inotropic support in grown-up congenital heart patients undergoing cardiac surgery: a propensity score adjusted analysis

Annals of Intensive Care

Table 3 Outcome variables after inverse probability of treatment weighting

Variable	Levosimendan	Milrinone	Average treatment effect 95% CI	P value
Requiring epinephrine (proportions)	0.42	0.26	0.16 [− 0.03 to 0.34]	0.10
Epinephrine support (days)	1.00	1.82	− 0.82 [− 3.35 to 0.95]	0.25
VIS on postoperative day 1	8.60	13.86	− 5.26 [− 12.70 to − 1.36]	0.02
VIS on postoperative day 2	4.46	10.11	− 5.65 [− 13.68 to − 0.42]	< 0.001
VIS on postoperative day 3	2.57	6.84	− 4.28 [− 11.76 to − 1.13]	< 0.001
VIS on postoperative day 4	1.38	5.63	− 4.25 [− 11.89 to − 1.08]	< 0.001
Mechanical ventilation (h)	19.38	56.97	− 37.59 [− 138.85 to − 19.13]	0.01
CPIS within 48 h of admission	1.64	1.43	0.21 [− 0.41 to 0.81]	0.50
Renal replacement therapy (proportions)	0.03	0.08	− 0.05 [− 0.20 to 0.04]	0.26
Intensive care unit stay (days)	4.79	7.90	− 3.11 − 10.03 to − 1.48]	0.009
Hospital stay (days)	13.84	17.65	− 3.81 [− 15.58 to 3.88]	0.26

The 95% CI of the average treatment effect was estimated using bootstrapping with 500 re-sampling. Statistically significant results are shown in italics
CPIS Clinical Pulmonary Infection Score, VIS vasoactive inotropic score