Support time-dependent outcome analysis for veno-venous extracorporeal membrane oxygenation

Abstract

Objective: The majority of patients suffering from pulmonary failure refractory to mechanical ventilation require extracorporeal membrane oxygenation (ECMO) support between 1 and 2 weeks. This study was designed to evaluate differences in outcome depending on ECMO duration.

Methods: A retrospective analysis on n = 127 patients requiring veno-venous (VV) ECMO support at our institution between April 2006 and March 2010 was applied. The patient population was divided into three groups according to the support duration (A: 0-10 days), 75 patients; B: 11-20 days, 32 patients; C: >21 days, (max. 67 days), 19 patients). Statistical comparisons between groups were calculated.

Results: Mean age of all patients (♀ = 42 patients ♂=85 patients) was 48 ± 16 years (range 15-78 years). Bilateral pneumonia due to bacterial infection (n = 45 patients) or due to aspiration (n = 19 patients) was the main cause for pulmonary failure, other causes were extrapulmonary sepsis (n = 27 patients), major surgery (n = 17 patients), and severe trauma (n = 12 patients). Mean lung injury score (LIS) according to Murray was 3.4 ± 0.4, and mean sequential organ failure assessment (SOFA) score was 12.6 ± 3.7. Statistical comparisons revealed no significant difference in demographic parameters between groups. VV ECMO support immediately improved oxygenation, within 2h the PaO(2)/FiO(2) ratio rose from 80 ± 42 mm Hg to 129 ± 72 mm Hg (p = 0.001). Overall survival to discharge was 51.2%. There was a statistical difference in survival between groups (A = 59%, B = 31%, C = 52%; p = 0.029). Multivariate logistic regression analysis revealed renal failure (odds ratio (OR) 12.1; confidence interval (CI) 3.9-30.0; p < 0.001) and the use of NO (OR=5.8; CI=1.9-24.9; p=0.002) as risk factors for mortality. Complications consisted of cannula-related complications (14%), bleeding issues (13%), partial vein thrombosis of the cannulated vessels (9.5%), and pumphead failure (1.5%).

Conclusion: VV ECMO in patients suffering from severe lung failure is effective in improving gas exchange with an overall survival of higher than 50%. Prolonged need of ECMO support does not have an impact on survival.