RESUMO
OBJECTIVE: The objective of our study is to detail our experience relating to ECMO implantations for post-cardiotomy refractory shock, by analyzing the pre-ECMO factors (history, type of surgery, LVEF), factors relating to ECMO (implantation time, duration) and post-ECMO factors (weaning, complications) in order to highlight those possibly associated with high mortality. METHODS: This is a univariate and multivariate retrospective study of ECMO data implemented between 2011 and 2019 at the Grenoble Alpes University Hospital Center following cardiac surgery. The time to implantation of ECMO was less than 3hours (intraoperative) between 3 and 24hours (early postoperative) and between 24 and 48hours after aortic unclamping (late postoperative). Preoperative or postoperative intra-aortic balloon counterpulsation (CPBIA) could be associated. RESULTS: 114 veino-arterial ECMOs were implanted for refractory cardiogenic shock after 5702 cardiac surgeries (1.9%) with a survival rate of 30.7%. The mean age of the patients was 68.6+- 10.5 years. The implantation of ECMO was performed intraoperatively in 71 patients (62.2%), early postoperatively in 22 patients (19.2%) and late postoperatively in 21 patients (18.4%). The duration of assistance was less than 48hours in 27 patients (23.6%), between 48hours and one week in 58 patients (50.9%) and more than one week in 29 patients (25.5%). Univariate analysis revealed a statistically significant association between mortality rate and male sex (P=0.002), association absent with other preoperative characteristics, delay in implantation of ECMO, installation of CPBIA, post-operative characteristics and resuscitation suites. Multivariate analysis of the entire study population demonstrated that the use of ECMO for cardio-respiratory arrest was the only independent risk factor for mortality (OR=7.57 [1.41-40, 62]). After multivariate reanalysis excluding patients with ECMO placement for cardio respiratory arrest, age, preoperative renal failure, type of procedure and EuroSCORE II were risk factors for mortality. CONCLUSION: In this study, male gender, type of intervention, occurrence of cardiac arrest were significantly associated with the death rate. A study of greater power, multicentric, and with a larger sample, will have to be carried out to reach significance.