Quantifying recirculation in extracorporeal membrane oxygenation: a new technique validated.

Rationale: The efficacy of veno-venous extracorporeal membrane oxygenation is limited by the phenomenon of recirculation, which is difficult to quantify. Existing measurement techniques using readily available equipment are unsatisfactory.

Objectives: 1) To compare the accuracy of measurements of recirculation made using equations comparing blood oxygen content or saturation alone at different points in an ex vivo circuit; 2) to validate a new step-change technique for quantifying recirculation in vivo.

Methods: anesthetized greyhound dogs cannulated for veno-arterial support were connected to a circuit that allowed the creation of a known level of recirculation ex vivo and blood oxygen content/saturation monitoring. In two dogs, the accuracy of measurements derived from oxygen content and oxygen saturation were compared. The potential of a new technique for measuring recirculation in vivo by comparing the oxygen content of blood sampled during oxygenator bypass to that following a step-change in circuit oxygenation was demonstrated in a veno-venous pilot study and validated in a three-dog veno-arterial study.

Results: Measurements made using oxygen content versus oxygen saturation showed superior correlation with true recirculation (r(2)=0.87 vs. 0.64, p<0.0001) and less proportional measurement bias (10.3% vs. 49.8%, p=0.0045). Measurements of recirculation made using a step-change in circuit oxygenation and comparing oxygen content as is required for measuring in vivo recirculation overestimated by only 18.6% (95% Cl: 3.9-33.2%) and had excellent correlation with true values (r(2)=0.89).

Conclusions: 1) Measurement of recirculation using oxygen content is superior to that using oxygen saturation alone, which demonstrates significant measurement bias; 2) the novel step-change technique is a sufficiently accurate technique for the measurement of recirculation in animal models.