Background: Extracorporeal membrane oxygenation (ECMO) provides support to patients with severe but reversible cardiac or pulmonary failure. Vascular complications of ECMO are well recognized.
Methods: We performed a retrospective review of 70 patients (mean age 48 years; 15-85) who received peripheral veno-arterial ECMO from 2004 to 2010 in a single centre. For statistical analysis, chi-squared test and multivariate binary logistic regression analysis were used to assess for association between response variables (i.e. limb ischaemia, ECMO site bleeding and deep vein thrombosis (DVT)) and possible predictive variables.
Results: There were 14 (20%) cases of acute limb ischaemia with no statistically significant relationship between acute limb ischaemia and independent variables. Thirty-three patients received distal limb cannulas (47%). There was no statistically significant association between limb ischaemia and presence of distal limb cannula (P = 0.8). Multivariate binary logistic regression analysis identified insertion by cutdown as a predictor of lower probability of insertion site bleeding (n = 12, odds ratio 0.24, P = 0.04). Seven cases of DVT were identified; multivariate binary logistic regression analysis identified insertion by cutdown (odds ratio 0.08, P = 0.03) and days of ECMO less than five (odds ratio 0.08, P = 0.04) as predictive factors for reduced rates of DVT.
Conclusion: Ischaemic complications of ECMO are common and occur despite the presence of a distal limb-perfusing cannula; however in our study the distal limb cannula was a limb-salvaging intervention in six patients. Prolonged time on ECMO is a risk factor for DVT, and a high index of suspicion must be maintained. Percutaneous insertion was associated with higher rates of bleeding and DVT.
Keywords: cardiothoracic surgery; emergency medicine; extracorporeal membrane oxygenation; intensive care; other categories; vascular surgery.
© 2016 Royal Australasian College of Surgeons.