Hemolysis during cardiac extracorporeal membrane oxygenation: A case-control comparison of roller pumps and centrifugal pumps in a pediatric population

Extracorporeal membrane oxygenation (ECMO) is a lifesaving therapy, which has been used for the support of children with a broad range of diseases. Two pumps of differing mechanisms have been used to generate the extracorporeal flow: roller-head pumps and centrifugal pumps. Seven patients supported during ECMO with Levitronix Centrimag (Centrimag group [CG]) were matched to 14 patients supported with Stockert-Shiley SIII (Stockert-Shiley group [SSG]) at a single institution from July 2007 to July 2009. We hypothesized that hemolysis as measured by plasma-free hemoglobin (PFH) is elevated in the SSG versus the CG during cardiac ECMO. Categorical data were analyzed using Fisher's exact test. Plasma-free hemoglobin differences between groups were analyzed using both Wilcoxon rank sum and beta regression. Overall, SSG patients had two times the odds of having a higher PFH than CG patients adjusting for repeated measures (odds ratio [OR] = 1.96, 95% confidence interval [CI]: [1.15-3.34], p < 0.014). Differences between circuit failure in the first 168 hours did not reach statistical significance (1/7 CG vs. 7/14 SSG; p = 0.174). In this population of cardiac patients requiring ECMO support, more hemolysis occurred in the SSG, a roller-head pump supported group, when compared with the CG, a centrifugal pump supported group. Differences in circuit life did not reach statistical significance. This pilot study contrasts with past studies, which have demonstrated more hemolysis occurring with centrifugal pumps when compared with roller-head pumps.